PULMONARY-ALLERGY DRUGS ADVISORY COMMITTEE MEETING

 

SEPTEMBER 6, 2002

 

 

 

CLINICAL BRIEFING DOCUMENT

 

 

 

NDA 21-395

SPIRIVAÒ

(TIOTROPIUM BROMIDE)

INHALATION POWDER

FOR COPD

 

APPLICANT:

BOEHRINGER INGELHEIM PHARMACEUTICALS, INC.

 


Table of Contents

            Cover Page................................................................................................ 1

Table of Contents............................................................................................................... 2

I.          Introductory Statement.......................................................................................... 4

II.        Overview................................................................................................................. 7

A........ Brief Overview of the Clinical Program........................................................... 7

B......... Efficacy Evaluations....................................................................................... 7

C........ Safety............................................................................................................ 9

D........ Dosing......................................................................................................... 10

E......... Special Populations...................................................................................... 10

III.       Introduction and Background.............................................................................. 11

A........ Drug Established and Proposed Trade Name, Drug Class, Sponsor’s Proposed Indication(s), Dose, Regimens, Age Groups................................................................................ 11

B......... State of Armamentarium for Indication......................................................... 11

C........ Important Milestones in Product Development.............................................. 11

D........ Other Relevant Information.......................................................................... 12

E......... Important Issues with Pharmacologically Related Agents............................... 12

IV.       Human Pharmacokinetics and Pharmacodynamics............................................ 14

A........ Pharmacokinetics......................................................................................... 14

B......... Pharmacodynamics...................................................................................... 19

V.        Description of Clinical Data and Sources........................................................... 30

A........ Overall Data................................................................................................ 30

B......... Tables Listing the Clinical Trials.................................................................... 30

C........ Postmarketing Experience............................................................................ 32

VI.       Clinical Review Methods..................................................................................... 33

A........ How the Review was Conducted................................................................. 33

B......... Overview of Materials Consulted in Review................................................. 33

C........ Overview of Methods Used to Evaluate Data Quality and Integrity............... 33

D........ Were Trials Conducted in Accordance with Accepted Ethical Standards...... 33

E......... Evaluation of Financial Disclosure................................................................. 33

VII.     Integrated Review of Efficacy............................................................................. 35

A........ Brief Statement of Conclusions..................................................................... 35

B......... General Approach to Review of the Efficacy of the Drug.............................. 35

C........ Detailed Review of Trials by Indication......................................................... 36

D........ Efficacy Conclusions.................................................................................... 47

VIII.    Integrated Review of Safety................................................................................ 50

A........ Brief Statement of Conclusions..................................................................... 50

B.  ...... Description of Patient Exposure................................................................... 50

C........ Methods and Specific Findings of Safety Review.......................................... 51

D........ Adequacy of Safety Testing......................................................................... 66

E......... Four-Month Safety Update.......................................................................... 66

IX.       Appendix: Detailed Reviews of Individual Studies............................................ 68

One-Year Placebo-Controlled Studies:.................................................................... 68

Six-Month Placebo- and Active-Controlled Studies............................................... 103

One-Year, Active-Controlled Studies.................................................................... 135




I.       Introductory Statement

The Pulmonary – Allergy Drugs Advisory Committee (PADAC) is being convened on September 6, 2002, in order to discuss the New Drug Application submitted to the FDA by Boehringer Ingelheim Pharmaceuticals, Inc. for SpirivaÒ (tiotropium bromide) Inhalation Powder (NDA #21-295).  Tiotropium is a long-acting anticholinergic agent that is proposed for use in chronic obstructive pulmonary disease (COPD).  No formulation of tiotropium has previously been approved for any use in humans in the US.  Spiriva is a dry powder formulation of tiotropium bromide, which is intended for administration by oral inhalation, using a re-usable, hand-held, breath-actuated device called the HandiHalerÒ.  The proposed dose is one (18mcg) capsule QD.  The Phase 3 clinical development program constituted six, multicenter, clinical studies of 6 to 12 months in duration.  Two of the studies were placebo-controlled, two were active- and placebo-controlled, and two were active-controlled studies.

 

The Applicant has proposed the following indication for Spiriva:

“Spiriva is indicated for the long term, once daily, maintenance treatment of bronchospasm and dyspnea associated with chronic obstructive pulmonary disease (COPD), including chronic bronchitis and emphysema.”

 

Inclusion of the word “dyspnea” in the “Indications” section of the product label would mark a departure from the language commonly used in the product labels of other medications approved in the US for COPD.  The “Indications” section of these labels commonly refer to the “treatment of bronchospasm” associated with COPD, intentionally focusing on the bronchodilator activity of the drugs, and avoiding the use of language that would imply that the drugs have been shown to treat a specific symptom of the disease, or the disease itself.  This custom is based, in part, on the recognition that, while FEV1 represents a direct measure of bronchospasm, it is only an indirect, or surrogate, measure of the overall disease that is COPD, which is characterized by a constellation of clinical signs and symptoms, physiologic processes, and histopathologic features.  The approval of drugs for COPD has been based, therefore, on the demonstration that the drug provides a clinically meaningful degree of bronchodilation for patients with COPD.  The post-treatment change in FEV1 is commonly used to demonstrate this. 

 

In general, the Agency approves drugs only if it can determine that the drug will provide a real benefit to the patient.  As stated above, FEV1 can be considered a “direct” measure of bronchospasm.  However, a drug whose sole benefit was an improvement in a physiologic parameter, without clinical benefit discernible to the patient, would generally not be approved unless the physiologic parameter was a validated surrogate for a clinical benefit discernible to the patient.  Intrinsic to the approval of COPD drugs indicated for the treatment of bronchospasm (based on an FEV1 endpoint), has been the implicit assumption that the temporary relief of bronchospasm is associated with a clinically discernible benefit.  This raises the question of whether it is appropriate to list specific symptoms of the disease, such as dyspnea, which may improve based on the stated bronchodilator activity of the drug, as “Indications” for a drug.

 

The Phase 3 clinical development program for Spiriva has attempted to support both the efficacy of the drug as a bronchodilator, and the efficacy of the drug in the treatment of the symptom of dyspnea in patients with COPD.  Each of the six “pivotal” studies submitted in support of the application have addressed the bronchodilator activity by including FEV1 as a primary or co-primary endpoint, and including other secondary endpoints that assess bronchodilation (e.g. forced vital capacity, peak expiratory flow rates, and “rescue” albuterol use).  In this application, the primary endpoint was the change from baseline in the pre-dose (or “trough”) FEV1 value, rather than a post-dose value, such as peak FEV1, as is more commonly the case in COPD clinical studies.  A benefit of using the “trough” FEV1 endpoint is that it can provide justification of the proposed dosing interval, by demonstrating continued efficacy at the end of the dosing interval.  One potential drawback is that there is less consensus regarding the minimum magnitude of effect that should be considered to be clinically meaningful at this timepoint.  The pivotal clinical studies included numerous secondary analyses of FEV1 and FVC to evaluate the bronchodilator effect in the early post-dosing period (e.g. peak values, and average values from serial post-dosing spirometry).

 

In regard to the proposed dyspnea claim, two of the six “pivotal” Phase 3 studies included an index of the symptom, the Mahler Transitional Dyspnea Index (TDI), as a co-primary endpoint (Studies 205.130 and 205.137).  This variable was also included as one of the secondary efficacy variables in the remaining four studies.  In fact, the decision to amend the statistical plan for Studies 205.130 and 205.137 to include the TDI as a co-primary endpoint was made after these studies were completed, before un-blinding, based on post-hoc analyses of the TDI data from the earlier Phase 3 studies.

 

The purpose of this PADAC meeting is to discuss the adequacy of the safety and efficacy data submitted in the NDA to support approval for marketing of Spiriva.  Given the proposal for the unique indication of dyspnea, the topics for discussions will include the development, validation, and statistical analysis of the dyspnea instrument used in these studies (the TDI), the clinical significance of the TDI findings, and a more general discussion of what type and amount of data would constitute substantial, convincing evidence of a clinically meaningful benefit with regard to the symptom of dyspnea in patients with COPD.

 

During the meeting, the Applicant will present an overview of the NDA to the PADAC.  The FDA presentation will include:

·        A discussion of the Mahler TDI instrument.

·        Salient pharmacokinetic and pharmacodynamic features of tiotropium bromide.

·        An overview of the Phase 3 clinical program, including:

-        The extent and findings of the safety database

-        The efficacy findings in regard to bronchodilator effect

-        The efficacy findings in regard to dyspnea effect

 

 

 

During the meeting, members of  the PADAC are encouraged to keep in mind the following issues, on which the Agency seeks input.

 

1)      The extent to which the data submitted provides convincing evidence of a clinically meaningful bronchodilator effect of Spiriva, when used in the chronic treatment of patients with COPD.  Any specific further data that would be needed in order to provide such evidence.

 

2)      Any specific safety concerns regarding the use of Spiriva in this patient population that might prevent approval.

 

3)      Any specific safety concerns regarding the use of Spiriva in this patient population that might merit specific attention in the product label.

 

4)      The overall adequacy of the safety database, any further safety information that should be obtained, and when such information should be obtained, in relation to approval.

 

5)      In general, the type and amount of data that would constitute substantial, convincing evidence of a clinically meaningful benefit for a drug, with regard to the symptom of dyspnea in patients with COPD.

 

6)      The extent to which the data submitted provide convincing evidence that Spiriva has a clinically meaningful effect on the symptom of dyspnea in patients with COPD.

 

7)      The appropriateness of listing symptoms of COPD, which may improve based on the bronchodilator activity of a drug, as “Indications” for drugs that are approved for the treatment of bronchospasm associated with COPD.

 


II.      Overview

The purpose of this Clinical Briefing Document is to summarize those aspects of the New Drug Application (NDA) for SpirivaÒ (tiotropium bromide) Inhalation Powder (NDA #21-395) that may be relevant to the discussions of the Pulmonary-Allergy Drugs Advisory Committee, during the meeting to be held on September 6, 2002.  These aspects include human pharmacokinetic and pharmacodynamic data, reviews of the important clinical studies, and integrated discussions of both the safety and the efficacy of the drug.  Although they play an important role in regulatory decision-making, issues related to the Chemistry, Manufacturing, and Controls and the Preclinical Toxicology aspects of the NDA are not included in this Clinical Briefing Document because they will not be a topic of discussion at the PADAC meeting. 

 

Throughout the document, data sources within the NDA submission are referenced in square brackets.  It is recognized that the members of the PADAC do not have access to the full NDA submission, from which these references are drawn.



 

A.                 Brief Overview of the Clinical Program

A total of 4,124 subjects participated in the clinical program.  This included 224 healthy volunteers, 3,411 COPD patients, 471 asthma patients, and 18 patients with renal impairment.  Of these, a total of 2,117 subjects were exposed to tiotropium by inhalation of the powder capsule formulation.  This included 57 healthy volunteers, 1,723 COPD patients, and 337 asthma patients.  A total of 1,701 subjects were exposed to the 18mcg dose of tiotropium.

 

The Phase 3 program consisted of six, multicenter, controlled “pivotal” studies in patients with COPD.  For inclusion in these “pivotal” studies, patients were required to be 40 years old or older, have a smoking history of >10 pack-years, have a clinical diagnosis of COPD, and meet certain spirometry criteria (FEV1£ 60% or 65% [depending on the study] of predicted and FEV1£ 70% of FVC).  Baseline responsiveness to bronchodilator was not tested or required.  A total of 2,663 patients with COPD were enrolled in these six studies, approximately 1,300 of whom were treated with tiotropium.  These studies were:

-        Two, 1-year, placebo-controlled studies,

-        Two, 1-year, active (ipratropium bromide MDI) controlled studies, and

-        Two, 6-month, placebo- and active (salmeterol xinafoate MDI) controlled studies.

For further details regarding the clinical development program, the reader is referred to the section of this Clinical Briefing Document entitled “Description of Clinical Data and Sources.”

 

B.                 Efficacy Evaluations

The Phase 3 clinical studies used standard spirometric variables to assess for bronchodilator efficacy.  In all six studies, the primary efficacy endpoint was the “trough FEV1 response,” defined as the mean FEV1 change from baseline at the end of the dosing interval.  Both the baseline and the trough FEV1 were calculated as the mean of two pre-treatment FEV1 readings measured in the morning prior to administration of study medication.  This primary efficacy endpoint is somewhat atypical for studies of bronchodilator drugs, which usually examine the early post-dosing bronchodilator effect (e.g. peak FEV1) or the average FEV1 (e.g. the area under the FEV1-Time curve) as the primary efficacy analysis. 

 

One benefit of using the trough FEV1 as the primary efficacy endpoint is that this variable provides insight into the drug’s efficacy at the end of the dosing interval, thus providing support for the proposed dosing interval.  One limitation with using this primary efficacy endpoint is that there is little consensus regarding what magnitude of effect constitutes a clinically important effect at the very end of the dosing interval.  Customarily, in evaluating the results of a primary efficacy analysis both statistical and clinical significance are considered.  In justifying a proposed dosing interval for a bronchodilator drug, the Agency has generally expected that some efficacy is maintained for the bulk of the dosing interval.  However, a specific effect size at the end of the dosing interval has not been required.

 

Numerous secondary efficacy endpoints, including early post-dose spirometry and supplemental “rescue” albuterol use were also employed in order to examine the bronchodilator efficacy of this product.  One finding from these secondary endpoints is interesting because it represents a unique pharmacodynamic feature of tiotropium bromide.  That feature is the delayed onset of maximal bronchodilator response.  For most orally inhaled bronchodilators, the degree of bronchodilation achieved with the first dose is not different from that of subsequent doses.  With tiotropium bromide, a degree of bronchodilation is achieved with the first dose; however, the bronchodilator effect increases with multiple dosing, reaching a maximal effect at approximately Day 8.  Additional secondary efficacy endpoints employed in these studies included occurrences of COPD exacerbations and patient-reported outcomes such as the Saint George’s Respiratory Questionnaire and the Medical Outcomes Study SF-36.

 

In two of the six “pivotal” studies, the Mahler Transitional Dyspnea Index (TDI) focal score was included as a co-primary efficacy variable in order to support a proposed indication for the treatment of dyspnea in COPD patients.  The TDI focal score is the sum of the individual scores of the three components of the TDI (the “functional impairment,” “magnitude of task,” and “magnitude of effort” components).[1]  Four of the six “pivotal” studies included TDI assessments as secondary efficacy variables.  In those studies, the mean values of the TDI focal scores were analyzed.  After reviewing the TDI data from these studies, the Applicant decided to alter the primary efficacy endpoints for the two remaining “pivotal” studies, which were completed but for which the blind had not been broken (Studies 205.130 and 205.137).  These protocols were amended to include both the trough FEV1 response and the TDI focal score as co-primary efficacy variables.  Rather than the mean value analyses used in the other studies, a “responder” analysis of the TDI focal score was specified.

 

At various stages during the clinical development of tiotropium bromide, the Agency informed the Applicant that, for inclusion anywhere in the product label, the TDI instrument and the proposed analysis of the TDI data must be supported by substantial evidence.  Specifically, the instrument itself must be validated, the proposed “responder” threshold (sometimes referred to as the “minimal clinically important change”) must be validated, and the clinical significance of any difference in rates of “response” between active and placebo must be established.  One topic for the PADAC’s discussion will be the extent to which these requirements have been met, and the extent to which the data definitively demonstrate a clinically meaningful drug effect on the symptom of dyspnea.


 


 

C.        Safety


The table below summarizes the numbers of patients exposed to tiotropium, and the duration of exposure, in the six “pivotal” Phase 3 studies.

 

Patient Exposure to Tiotropium in the Six “Pivotal” Phase 3 Studies                            [iss.pdf/p113-4]

 

Total

³101 days

³200 days

³ 330 days

One-year, placebo-controlled studies

550

501

(91%)

482

(88%)

302

(55%)

One-year, ipratropium-controlled studies

356

325

(91%)

316

(89%)

260

(73%)

Six-month, salmeterol- and placebo-controlled studies

402

353

(88%)

354

(88%)

not applicable


 

The mean age for all patients was 65 years in the one-year, placebo-controlled studies, and 64 years in the one-year, ipratropium-controlled studies and the six-month, salmeterol and placebo-controlled studies.  Nearly all patients were Caucasian, and 65% to 85% were male.  The mean baseline FEV1 ranged from 1.0 to 1.25 liters, or 38-44% of predicted.

 

In the pivotal clinical trials safety was monitored with the following assessments:

-        clinical adverse events,

-        vital signs,

-        physical examination,

-        clinical laboratory testing, and

-        electrocardiograms. 

ECGs were performed at baseline and every 90 days for the duration of the study.  However, the protocols did not specify the timing of the ECGs in relation to study drug administration and the case report forms did not capture that information.  Therefore it cannot be assumed that the ECGs were obtained at Cmax, as would be most desirable.  However, timed ECGs were performed in a Phase 2 multiple-dose, dose-ranging study in which doses up to 44mcg were examined for up to 29 days. 

 

The pivotal clinical studies did not include Holter monitoring.  Holter monitoring was included in one Phase 2 study in which a total of 81 COPD patients were treated with tiotropium 18mcg QD for six weeks. 

 

The safety findings are discussed in the section of this Clinical Briefing Document entitled “Integrated Review of Safety.”  The following comments briefly summarize the safety findings.  The incidence of death was similar in all treatment groups, and the causes of death were consistent with what might be expected in this patient population.  Two causes of death were reported in the tiotropium group but not in the comparator groups.  They were myocardial infarction (4 deaths) and arrhythmia (1 death).  In the one-year, placebo-controlled studies, five of the seven deaths among the tiotropium patients, but only one of the seven deaths in the placebo patients, were attributable to cardiac ischemia or arrhythmia. Fewer patients in the tiotropium groups reported serious adverse events, as compared with both the placebo and the active comparator groups.  The incidence of discontinuation due to adverse events was also lower in the tiotropium groups as compared to both the placebo and active comparator groups.  In the one-year, placebo-controlled studies, the most notable adverse events (AEs) were related to the gastrointestinal system (dry mouth, dyspepsia, abdominal pain, constipation, and vomiting).  The occurrence of AEs in the category of “Gastrointestinal System Disorders” was 38.5% in the tiotropium group and 29.1% in the placebo group.  Among these, by far the most common was dry mouth, with an incidence of 16% in the tiotropium group, and 2.7% in the placebo group.  The one-year, ipratropium controlled studies demonstrated that the incidence of dry mouth was greater in the tiotropium group (12.1%) than in the ipratropium group (6.1%).  Upper respiratory tract infections were also more common in the tiotropium group than in the placebo group (41.1% vs. 37.2%).  There were subtle indications that tiotropium may be associated with an increased frequency of adverse cardiac effects, specifically in the category of “heart rate and rhythm disorders.”  This is discussed in the subsection of the Integrated Review of Safety entitled “Adverse Events Related to the Pharmacologic Actions of the Drug.”

 

D.        Dosing

The proposed dose of tiotropium bromide inhalation powder is 18mcg QD.  This is the dosing regimen that was studied in the Phase 3 clinical program.  In general, there are two aspects to a proposed dosing regimen that must be established, the dose and the dosing interval.  Insight into the appropriateness of the proposed dosing interval may be taken from the results of the primary efficacy variable utilized in the Phase 3 studies, the “trough” FEV1.  The clinical development program also included single- and multiple-dose dose-ranging studies in COPD patients, using a variety of formulations and doses of tiotropium.  The relevant dose-ranging studies are summarized in the section of this Clinical Briefing Document entitled “Human Pharmacokinetics and Pharmacodynamics.


 

E.         Special Populations

As mentioned above, the majority of the patients in the pivotal studies were men, and nearly all were Caucasian.  Drug-demographic safety interactions are discussed in the section of the Integrated Review of Safety entitled “Interactions.”  In the one-year, placebo-controlled studies, the AEs “dry mouth” and “constipation” occurred with greater frequency in older patients in the tiotropium group, but not the placebo group.  In these studies, the AE “urinary tract infection” occurred with greater frequency in older patients in both treatment groups, although the apparent age effect was more marked in the tiotropium group.  The occurrence of “dry mouth” was also more common in women in the tiotropium group, but not in the placebo group.  Because very few patients in the pivotal studies were non-white, analyses for drug-race safety interactions were not informative.  However, pharmacokinetic studies in African-American and Caucasian asthma patients indicate similar urinary excretion. There were no patients on tiotropium who became pregnant during the clinical development program.  Because the Applicant is seeking an indication for COPD, a disease of older adults, the Applicant has not studied the drug in pediatric patients.



 

III.           Introduction and Background

 

A.        Drug Established and Proposed Trade Name, Drug Class, Sponsor’s Proposed Indication(s), Dose, Regimens, Age Groups

This NDA is submitted in support of SpirivaÒ (tiotropium bromide) Inhalation Powder, a long-acting anticholinergic bronchodilator intended for use in patients with COPD.  In early development, the drug was identified as Ba679.  This product consists of two discrete elements [summary.pdf/p44].  The first element is a hard gelatin capsule containing a pre-metered dose of the drug substance and lactose as a dry powder.  The second element is the HandiHalerÒ inhalation device. The HandiHaler is a reusable, hand-held, breath-actuated device used to inhale the dry powder.  The active component of Spiriva is tiotropium.  Tiotropium is a quaternary ammonium compound.

 

The proposed language for the Indication is: “for the long term, once daily, maintenance treatment of bronchospasm and dyspnea associated with chronic obstructive pulmonary disease (COPD), including chronic bronchitis and emphysema.”

 

The proposed dose is one inhalation (18mcg) QD. The Indication section of the label will not refer to specific age groups.  COPD is a disease of adults.  The pivotal clinical studies performed in support of this application appropriately contained an inclusion criterion of age ³40 years.  This will be described in the Clinical Studies section of the label.


 


 

B.        State of Armamentarium for Indication

The only currently approved category of drugs for COPD are the bronchodilators.  Currently approved bronchodilators include several short-acting beta2-adrenergic agonists (e.g. albuterol, pirbuterol, bitolterol, metaproterenol, and terbutaline), two long-acting beta2-adrenergic agonists (salmeterol and formoterol), a short-acting anti-cholinergic agent (ipratropium), and theophylline.  These drugs are available in various formulations, including solutions and metered dose inhalers for oral inhalation, as well as various formulations for oral ingestion. Other classes of agents, such as corticosteroids and mucokinetic agents, have been investigated for their utility in the pharmacologic management of COPD but none of these are approved for COPD in the US.

 

If approved, tiotropium bromide inhalation powder would represent the first once-daily oral inhalation drug indicated for COPD.  The proposal to include a claim that tiotropium bromide is indicated for the treatment of dyspnea related to COPD would also be unique.  No other drug is approved for the treatment of dyspnea, or any other specific symptom associated with COPD in the US.


 

 

C.        Important Milestones in Product Development

This drug was developed under IND 46,687, which was originally submitted to the Agency on November 30, 1994.  The indication listed at the time of the original submission was “bronchodilator for maintenance treatment of bronchospasm associated with chronic obstructive pulmonary disease, including chronic bronchitis, emphysema, and moderate to severe asthma.” [indnda.pdf/p1]  In an Annual Report dated April 29, 1999, the Applicant notified the Agency that clinical development in patients with asthma had been discontinued.  In a submission dated October 8, 2001, the Applicant stated that studies of the product in adults with asthma have failed to demonstrate effectiveness. 

 

An End-of-Phase-2 meeting was held on December 3, 1996.  In 1999, two pre-NDA meetings were held.  The first, on May 10, 1999, focused on CMC issues.  Two days later, on May 12, 1999, a General pre-NDA meeting was held to discuss issues relevant to the other review disciplines.  Finally, on July 24, 2000, the Agency met with the Applicant to discuss the Applicant’s plans regarding the pursuit of a unique indication for this drug.  Based on its review of the completed Phase 3 studies, the Applicant wished to discuss the possibility of pursuing a “dyspnea” indication.  At that time, two additional large, 6-month studies were ongoing (Studies 205.130 and 205.137).  The Applicant intended to amend the protocols for these studies in order to designate two co-primary endpoints: FEV1 and the Mahler Transitional Dyspnea Index (TDI), in hopes of justifying the dyspnea indication.  At that meeting, and in a subsequent communication (October 11, 2000) sent to the Applicant in response to an additional submission (Dated August 22, 2000) the Agency advised the Applicant that the dyspnea indication would be unique and would require substantial supportive evidence.  The Agency informed the Applicant that substantial validation would be required in regard to the use of the TDI instrument, as well as justification of the clinical significance of the proposed definition of a “responder” and the clinical significance of the differences demonstrated in the percentages of “responders” in each treatment group.  The Agency also requested that the NDA include comparisons of mean TDI scores, in addition to the planned “responder” analysis. 

 

No previous NDAs have been submitted for this product.


 

D.        Other Relevant Information

As of November 9, 2001, Spiriva (tiotropium bromide) Inhalation Powder is not marketed in any country [summary.pdf/p43].  Registration dossiers have been filed in 18 countries, and approval has been obtained in two countries, The Netherlands and New Zealand.  In Europe, the Mutual Recognition Procedure is being adopted, with Netherlands serving as the Reference Member Site.


 

E.                 Important Issues with Pharmacologically Related Agents

Tiotropium is a long-acting, anticholinergic bronchodilator.  Ipratropium bromide is a short-acting, anticholinergic bronchodilator that is manufactured by Boehringer Ingelheim and is approved for use in patients with COPD.  The drug substance is marketed as a metered dose inhaler in two formulations: as the sole active agent (Atrovent Inhalation Aerosol), and as a combination product with albuterol sulfate (Combivent Inhalation Aerosol).  Ipratropium bromide is also approved as an inhalation solution and a nasal spray.  Ipratropium bromide has proven to be relatively safe in the COPD patient population.  According to the product label for Atrovent Inhalation Aerosol, the product should be used with caution in patients with narrow-angle glaucoma, prostatic hypertrophy, or bladder neck obstruction.  These precautions are based on the potential systemic anticholinergic effects of the drug, and cases of precipitation or worsening of narrow-angle glaucoma and acute eye pain have been reported.  Cases of hypotension and allergic-type reactions have also been reported.  The most common adverse events occurring in 90-day active-controlled trials were cough (5.9%), nervousness (3.1%), nausea (2.8%), dry mouth (2.4%), gastrointestinal distress (2.4%), dizziness (2.4%), headache (2.4%), and exacerbation of symptoms (2.4%).

 


 

IV.            Human Pharmacokinetics and Pharmacodynamics

 

A.                 Pharmacokinetics

1.                  Summary

The bioavailability of tiotropium is poor after oral administration (2-3%), and somewhat greater after oral inhalation (19.5%).  The Cmax after oral inhalation occurred at 5 minutes, the time of the first sample.  The drug remains measurable in the blood for 2-4 hours after single-dose oral inhalation.  The volume of distribution of tiotropium is quite large, 32 liters/kg.  Approximately 74% of the drug is eliminated in the urine as the parent compound.  Active renal secretion is likely, based on the observation that renal clearance of the drug exceeds the creatinine clearance.  The fate of the remaining 26% of the dose has not been established, but it may be metabolized by a combination of non-enzymatic hydrolysis and cytochrome P450-mediated metabolism (predominantly CYP2D6, and to a lesser extent, 3A4).  Although much of the drug is eliminated in the urine quickly (e.g. 44% of the administered dose by 4 hours after single dose administration), the drug persists in the urine for many days, with a terminal elimination half-life of 5 to 6 days.  Despite this long half-life, daily administration for 14 days resulted in accumulation of only 2 to 3 fold.  This finding, consistent with the large volume of distribution, suggests a multi-compartment model, whereby the drug is distributed to more than one physiologic compartment, from which it is slowly released back into the circulation. Older patients and subjects with impaired renal function exhibit increased plasma concentrations of tiotropium.

 

2.                  Background

During drug development, tiotropium was quantified using two analytical methods [biosum.pdf/p15].  The radioreceptor assay, which had a limit of quantification of 400ng/mL, was used in the initial studies to quantify the tiotropium in the urine.  Subsequently, this test was replaced by a liquid chromatographic/mass spectrometric assay, which was able to measure concentrations down to 5pg/ml in human plasma and 10pg/mL in human urine.  Using this assay tiotropium was measurable in the plasma up to 2-4 hours and in the urine for many days following a single dose of 18mcg.

 

During drug development, drug doses and concentrations were initially expressed in terms of the salt (tiotropium bromide monohydrate).  Later in development, in order to comply with  a European Directive, a decision was made to label the product in terms of the active entity in the molecule (i.e. the tiotropium cation) for the Phase 3 supplies and commercial drug product.  In order to be able to use whole numbers, the actual drug content in the capsules was adjusted (+2.5%) [biosum.pdf/p30].  In addition, the dry powder inhalation capsules used during Phase 1 and 2 actually contained 10% more tiotropium bromide monohydrate than was expressed in the label claim [biosum.pdf/p32].  This was the Applicant’s practice at that time, based on its experience with other inhalation capsules, which suggested that only about 90% of the content of an inhalation capsule actually leaves the capsule and the device during inhalation (i.e. delivered dose).  Finally, it should be noted that the dry powder inhalation studies were performed with two different devices, the FO2 device (also called the Inhalator Ingelheim) and the HandiHaler device.  The Applicant states that these two devices showed identical functional properties and did not differ relevantly in their flow characteristics [biosum.pdf/p34].

 

The pharmacokinetics of tiotropium were studied in 15 clinical studies in a total of 600 subjects.  These include 142 healthy male subjects in eight Phase 1 studies, 18 subjects (3 female, 15 male) with renal impairment (mild to severe), and 434 patients with COPD or asthma in six studies [biosum.pdf/p29].  The studies involved single and multiple tiotropium doses, ranging from 4.5mcg to 282mcg for dry powder inhalation, from 2.4mcg to 14.4mcg for IV infusions, and from 8.0mcg to 64mcg for oral solutions.

 

Five of the six studies in patients with lung disease included sparse data sets with more extensive urine samplings [biosum.pdf/p16].  The sixth included single- and multiple-dose administration and frequent blood and urine collections (Study #205.133; Report #U00-3029).

 

The PK studies included the following routes of administration [biosum.pdf/p77]:

-        Intravenous: Studies 205.105 (Report U99-1315), 205.107 (Report U98-2282), and 205.134 (Report U00-1289).

-        Oral (solution): Studies 205.105 (Report U99-1315) and 205.106 (Report U97-2337)

-        Oral inhalation:

-        Piezoelectric dispersion of solution: 205.101 (Report U93-0252)

-        BINEB device (dispersion of solution, later modified to the RESPIMAT device): 205.112 (Report U97-2462)

-        Dry powder inhalation: Studies 205.102 (Report U93-0704), 205.103 (Report U93-0939), 205.104 (Report U93-0940), 205.105 (U99-1315), 205.108 (Report U96-3068), 205.117 (Report U99-3169), 205.120 (Report U94-0198), 205.127 (Report 00-0077),  205.133 (Report U00-3029), and 205.201 (Report U98-3174)

 

 

The following table summarizes the clinical studies in which pharmacokinetic assessments were made.


 

Clinical Studies with Pharmacokinetic Assessments                                                            [biosum.pdf/p39-65]

Study #

(Report #)

Design/

Duration

Diagnosis/

# of Subjects

Route

Treatments

205.101

(U93-0252)

R, SB, PC/

Single Dose

Healthy males aged 21-50 years/

N=6 per treatment group

Inhalation Solution (via piezo electric)

0.8, 4, 8, 20, 40, 80, or 160mcg, or pbo

205.102

(U93-0774)

R, SB, PC/

Single Dose

Healthy males aged 21- 50 years/

N= 6 per treatment group

Inhalation

(inhalet via FO2 device)

35.2, 70.4, 140.8, or 281.6mcg, or pbo

205.103

(U93-0939)

R, DB, PC, XO/

7 days

Healthy males aged 21-50 years/

N=12

Inhalation

(inhalet via FO2 device)

70.4 or 140.8mcg, or pbo

205.104

(U93-0940)

R, DB, PG

14 days

Healthy males aged 21-50 years/

N=15

Inhalation

(inhalet via FO2 device)

8.8, 17.6, or 35.2mcg

205.105

(U99-1315)

R, OL, PG

Single dose

Healthy males aged 21-50 years/

N=12 per treatment group

Inhalation (via HandiHaler),

Oral solution, and  Intravenous solution

108mcg inhaled;

64mcg oral soln.;

14.4mcg IV soln.

205.106

(U97-2337)

One day at each dose level

Healthy males aged 21-50 years/

N=4-6 at each dose level

Oral solution

8, 16, 32, or 64mcg, or pbo

205.107

(U98-2282)

DB, PC, increasing dose

3 days

Healthy males aged 21-50 years/

N=17

Intravenous solution

Single dose 2.4 or 14.4mcg, two subsequent daily doses of 4.8 or 9.6mcg; or pbo

205.108

(U96-3068)

R, DB, PC, PG

4 weeks

COPD patients

N=169 (33-35 per group)

Inhalation (inhalet via FO2 device)

4.4, 8.8, 17.6, or 35.2mcg, or pbo

205.112

(U97-2426)

PC, DB within group, multiple rising dose

14 days

Healthy males aged 21-50 years/

N=36 (9 per group)

Inhalation Solution (Respimat device)

8, 16, or 32mcg, or pbo

205.114/

205.117

(U99-3169)

R, DB, PC, PG

49 weeks

COPD

N=470

Inhalation (HandiHaler device)

18mcg or pbo

208.120

(U94-0198)

R, DB, PC, XO

Single dose

COPD

N=35

Inhalation (inhalet via FO2 device)

8.8, 17.6, 35.2, or 70.4mcg, or pbo

205.127

(U00-0077)

R, DB, PC, PG

3 weeks

COPD

N=202

Inhalation (inhalet via FO2 device and solution via Respimat)

Respimat: 1.25, 2.5, 5, 10, or 20mcg;

Inhalet: 18mcg; or pbo

205.133

(U00-3029)

OL

14 days

COPD

N=29

Inhalation (HandiHaler Device)

18mcg

205.134

(U00-1289)

OL

Single dose

Volunteers w/ renal impairment

N=24

Intravenous solution

4.8mcg

205.139

DB, PC, XO

Single dose

COPD

N=28

Inhalation (inhalet via HandiHaler)

9, 18, or 36mcg, or pbo

205.201

(U98-3174)

R, DB, PC, PG

21 days

Asthma

N=204

Inhalation (inhalet via HandiHaler)

4.5, 9, 18, or 36mcg, or pbo

R= randomized; SB= single blind; DB= double blind; PC= placebo controlled; PG= parallel group; OL= open label; pbo=placebo


 

3.                  Absorption

Tiotropium was administered to humans as intravenous infusion, oral solution, and by inhalation.  Inhalation was accomplished by various means including piezoelectric dispersion, dry powder inhalation capsules, and aerosolization of aqueous solution [biosum.pdf/p16].  Tiotropium was shown to be poorly absorbed after oral ingestion of a solution (absolute bioavailability of 2-3% for a 64mcg dose) (Study #205.105, Report #U99-1315).  Administration as an orally inhaled dry powder resulted in greater bioavailability (19.5% after an inhaled dose of 108mcg [3 doses of a 36mcg dry powder capsule using the HandiHaler device] in Study #205.105, Report #U99-1315) [biosum.pdf/p16].  After oral inhalation of a single dose of dry powder formulation, tiotropium may be detected in the blood at the time of the first sample (levels of 17-19pg/mL 5 minutes following inhalation of 18mcg) [biosum.pdf/p18].  Tiotropium remains measurable until 2-4 hours after oral inhalation of a single dose.  Interestingly, the second once-daily dose generates consistently higher AUC values than expected from the first dose.  The Applicant states that this is not likely due to limited assay sensitivity for the first dose, since a similar finding was observed after intravenous dosing (Study #205.107, Report #U98-2282).  The Applicant postulates that the finding may be due to incomplete saturation of binding sites (including muscarinic receptors) after the first dose, and a very slow dissociation constant of the tiotropium binding site complex.  Once all binding sites are at least near to saturation, more tiotropium can escape from the tissue and the drug appears faster in the systemic circulation [biosum.pdf/p18].

 

Tiotropium concentrations after oral inhalation differ in healthy subjects, younger COPD patients, and older COPD patients.  Five minutes after a single inhalation of 17.6mcg in these subjects, the geometric mean tiotropium concentrations were 24.6pg/mL (Study 205.104), 15.3pg/mL, and 9.63pcg/mL (Study 205.133), respectively [biosum.pdf/p83].

 

Although much of the drug is rapidly eliminated in the urine (e.g. 44% by 4 hours, 48% by 8 hours, and 54% by 24 hours), tiotropium remains present in the urine for many days, and thus has a very long elimination half-life (5-6 days) (Study #205.105, Report #U99-1315).  After multiple administration, pharmacokinetic steady state was reached after 2-3 weeks.

 

4.                  Distribution

In rats, autoradiography studies after intratracheal (Study #not given, Report #U90-0448) and intravenous (Study #PK-99011, Report #U99-0210) administration indicated that tiotropium distributes in higher amounts in the lung, liver, kidney, stomach, and gastrointestinal tract, with particularly long persistence in lung tissue after intratracheal administration [biosum.pdf/p18].  In addition, tissue sampling performed in Study #PK-99011 demonstrated notable distribution in the brown fat, pancreas, salivary gland, prostate, hypophysis, and thyroid gland [U99-0210.pdf/p15].  In three autoradiography studies in rats, distribution to the brain was not detected (Study #, Report #U90-0448), detected at low levels (Study #PK-99011, Report #U99-0210), or detected at higher levels (Study #PK-98005, Report #U99-0205) [biosum.pdf/p19]. Experiments in rats demonstrated that tiotropium crosses the placenta and is excreted in the milk of lactating rats [biosum.pdf/p19].

 

In an in vitro human plasma binding study, 72% of the drug was bound to plasma proteins.  In humans, the volume of distribution after a 14.4mcg intravenous infusion was 2665 Liters or 32 L/kg (Study #205.105, Report #U99-1315) [biosum.pdf/p78].  This large volume of distribution indicates extensive tissue binding.

 

5.                  Metabolism and Elimination

Tiotropium is an ester of the N-quaternary alcohol N-methylscopin with dithienylglycolic acid, which is cleaved in solution at physiologic pH with a half-life of up to 17 hours, and more slowly at lower pH.  There is evidence to suggest that this ester hydrolysis is non-enzymatic [biosum.pdf/p66]. 

 

Tiotropium is predominantly eliminated via renal secretion of unchanged drug.  After intravenous administration in healthy young men, 73.6% of the dose was recovered in the urine (Study #205.105, Report #U99-1315). The fate of the remaining quarter of the intravenous dose in young healthy subjects is not known.  It is expected that a portion of the drug is metabolized by hydrolysis or by the cytochrome P450 system; however, mass balance studies were not performed.  The Applicant suggests that binding of tiotropium to its binding sites may prevent cleavage.  Once it is released from its binding site and appears in the circulation, it is rapidly cleared.  Renal clearance after both intravenous and inhalation exposure exceeded calculated creatinine clearance, indicating that tiotropium is actively excreted by a transporter.  It is not known which cation transporter is responsible for the active renal secretion.  The Applicant states that in vitro studies using cyclosporine, a competitive inhibitor of p-glycoprotein, suggest the transporter is not p-glycoprotein [biosum.pdf/p20].

 

Urinary data in healthy subjects demonstrate that tiotropium was excreted with a geometric mean elimination half-life of 5.71 days after single-dose intravenous administration and 4.84 days after single-dose inhalation.  Urinary excretion indicated an accumulation by a factor of 2-3 from the first to the fourteenth inhalation [biosum.pdf/p21].  Thus, the AUC after 14 days is 2-3 times higher than after a single dose.

 

Tiotropium does not inhibit cytochrome P450 1A1, 1A2, 2B6, 2C9, 2C19, 2D6, 2E1, or 3A in human liver microsomes [biosum.pdf/p22].  However, in vitro studies showed that quinidine, a CYP 450 2D6 and 3A4 inhibitor, can inhibit the metabolism of tiotropium [biosum.pdf/p25]. The submission dated April 18, 2002 (Four-Month Safety Update), contained the following information.  Poor metabolizers of CYP 2D6 had a 33% higher tiotropium AUC0-4h after intravenous administration in comparison to extensive metabolizers [4/18/02 submission, iss.pdf/p269].

 

Pharmacokinetic studies to assess special populations indicate the following [biosum.pdf/p22-4]:

-        Gender does not significantly influence drug plasma or urinary excretion of tiotropium. 

-        Elderly COPD patients (>65 years) demonstrate decreased renal clearance of tiotropium and increased plasma concentrations.  In Study 205.133, the renal clearance was 326mL/min in younger COPD patients (mean age: 53 years), versus 163mL/min in the older patients (mean age: 74 years).  The AUC0-4h values were 18.2pg.h/mL in the younger group and 26.1pg.h/mL in the older group.

-        Patients with renal impairment demonstrate lower renal clearance and higher plasma concentrations.  Tiotropium plasma concentrations (AUC0-4h) were 39, 81, and 94% higher in mild, moderate, and severe renal impairment when compared to control subjects.

-        The effect of hepatic impairment was not studied.  The Applicant states that such studies were not performed because renal excretion dominated the elimination of tiotropium in healthy volunteers.

-        The Applicant states that the effect of chronic pulmonary disease on the absorbed fraction of the inhaled dose is not exactly known because this effect is hard to separate from the confounding effects of age and formulation on the urinary excretion.  A study in asthma patients suggested that increased severity of lung disease is associated with decreased urinary excretion.  This effect was not demonstrated in studies with COPD patients.

-        African American and Caucasian asthma patients excreted very similar amounts of tiotropium after once daily inhalations of 4.5, 9, 18, or 36mcg of tiotropium.

 

6.                  Drug-Drug Interactions

The Applicant states that tiotropium is not expected to influence the metabolism of other drugs because of “the very small dose of tiotropium and the lack of inhibition of CYP 450 isoenzymes by tiotropium.” [biosum.pdf/p25]  The Applicant also states that it is unlikely that other drugs will influence the metabolism of tiotropium, although the possibility of such interactions “cannot be completely excluded.”  It is possible that a drug that inhibited the renal cation transporter could result in increased plasma tiotropium concentrations. The submission dated April 18, 2002 (Four-Month Safety Update), included data from a pharmacokinetic study in which repeated supratherapeutic doses of cimetidine to inhibit these transporters increased the tiotropium AUC0-4h by 20%, while repeated 300mg doses of ranitidine had no effect (Study 205.222) [4/18/02 submission, iss.pdf/p269].

 

The effect of food on the oral bioavailability was not examined.

 

Factors that can increase systemic exposure are impaired renal function, concomitant cimetidine (inhibitor of transporter, 20%), and 2D6 poor metabolizers (33%) [4/18/02 submission, iss.pdf/p269].



 

B.                 Pharmacodynamics

 

1.                  Efficacy Dose-Ranging

The Applicant indicates that a total of 22 studies have been completed to evaluate the pharmacology of tiotropium [hpsum.pdf/p10].  This section of the Clinical Briefing Document will focus on the dose-ranging studies used to support the proposed dose.  The COPD dose-ranging studies are listed in the table below. 

 

COPD Dose-Ranging Studies (Inhalation Powder)                                                            [hpsum.pdf/p12 and ise.pdf/p88]

Study #

Country/ Dates

Design

Treatments

(Tiotropium)

Device

Duration

# of Subjects

Population

Primary Endpoint

205.119

Netherlands  11/91-4/92

Dose-ranging

Open label

XO

10mcg

20mcg

40mcg

80mcg

160mcg

RESPIMAT

Single Dose

6

(2F/ 4M)

COPD

FEV1

205.120

Netherlands 10/92-5/93

Dose-ranging

R, DB, PC, XO

10mcg

20mcg

40mcg

80mcg

Placebo

INHALATOR INGELHEIM (FO2)

Single Dose

35

(3F/ 32M)

COPD

FEV1

205.139

Japan   7/98-5/99

Dose-ranging

R, DB, PC, XO

11.3mcg1

22.5mcg1

45mcg1

Placebo

HANDIHALER

Single Dose

27

COPD

FEV1

205.108

US       1/95-9/95

Dose-ranging

Multicenter, R, DB, PC, PG

4.4mcg2 QD

8.8mcg2 QD

17.6mcg2 QD

35.2mcg2 QD

Placebo QD

INHALATOR INGELHEIM (FO2)

4 Weeks

169

(73F/ 96M)

COPD

FEV1


 

 

            Summaries of the COPD Dose-Ranging Studies

 

·        Study 205.119: “Pilot dose-escalation study of Ba 679 BR in chronic obstructive pulmonary disease.”  (Report #U92-0750)

-        This was an open-label, single-dose, five-period, cross-over study performed in The Netherlands between 11/91 and 4/92 [U92-0750.pdf/p16].  A total of six patients with COPD received the following doses of tiotropium inhalation solution, using the RESPIMAT device: 10mcg, 20mcg, 40mcg, 80mcg, and 160mcg.  The duration of the washout period between doses was determined based on the pharmacodynamic effect.  The washout was specified to be at least 48 hours after the last observed efficacy (defined as FEV1 ³15% above baseline).  For inclusion into the study, patients were required to demonstrate reversible airway obstruction, defined as a >15% improvement in FEV1 30 minutes after inhalation of ipratropium bromide, and to report coughing and excess mucus production on most days for at least 3 months of the year for at least 2 successive years.  The primary endpoints were the peak FEV1, the time to peak FEV1, and the area under the 24-hour FEV1 curve (divided by 24).

-        The mean peak FEV1 change from baseline showed dose ordering for doses up to 80mcg (21% for 10mcg, 30% for 20mcg, 32% for 40mcg, 47% for 80mcg, and 43% for 160mcg) [U92-0750.pdf/p18]. The mean time to peak FEV1 change from baseline, which ranged from 110 to 148 minutes, did not show dose-ordering [U92-0750.pdf/p43].  The FEV1 AUC0-24h/24 showed approximate dose-ordering (with the exception of the 40mcg dose, which was inferior to the 20mcg dose on this parameter) [U92-0750.pdf/p43]. 

-        The serial FEV1 curves demonstrate an interesting finding.  In all dose groups, the FEV1 declined gradually to a nadir at 23 hours.  However, in all dose groups the 24-hour FEV1 measurement was remarkably higher than the 23-hour measurement. Because of this finding, hourly spirometry was continued from 24 to 29 hours in the 160mcg dose cohort.  Each of these measures was notably higher than the 23-hour nadir.  Reviewer’s Note: This is an unusual finding.  However, interpretation is difficult in the absence of a placebo group.

-        This was a pilot study that demonstrated a dose-response bronchodilator effect of tiotropium .  However, it is difficult to draw conclusions relevant to this NDA based on this study because: 2) the dose escalation was not blinded; 2) the washout periods were not likely sufficiently long to allow elimination of previous doses; and 3) the formulation and delivery device differ substantially from the proposed drug product.  The study drug was administered as an inhalation solution, using the RESPIMAT device.  The significance of the unusual finding of improvements in FEV1 between the 23-hour and 24-hour measurements is not known.

 

·        205.120: “Dose-response and time-response study of Ba 679 BR in patients with chronic obstructive pulmonary disease.” (Report #U94-0198)

-        This was  a randomized, double-blind, placebo-controlled, single dose study performed in The Netherlands, between October, 1992 and May, 1993 [U94-0198.pdf/p26].  A total of 35 patients (32 male, 3 female) with COPD received the following doses of tiotropium dry powder capsule using the Inhalator Ingelheim device (also known as the FO2 device): 10mcg, 20mcg, 40mcg, and 80mcg, and placebo.  The washout period between dosing was 72 hours. For inclusion into the study, patients were required to demonstrate reversible airway obstruction, defined as a >15% improvement in FEV1 30 minutes after inhalation of ipratropium bromide.  The primary efficacy variable was FEV1, focusing on peak response, and average FEV1 over a various time periods (8, 12, 24, and 32 hours).

-        The baseline FEV1 on the first test day was significantly different from other test days (p=0.001), indicating carry-over effect.  Reviewer’s Comment: Given the pharmacokinetics of this drug, it is not surprising that carry-over effects would be demonstrated in a study using a 72-hour washout period.  In addition to performing analyses that did not attempt to adjust for carry-over effects, the Applicant performed two additional analyses in order to adjust for carry-over effects.  In one analysis, a parallel group comparison was performed based only on the test day 1 data.  In a separate analysis, comparisons were made using a data set that excluded visits following a visit in which the subject received a 20, 40, or 80mcg dose of tiotropium. 

-        As seen in Study 205.119, the FEV1 increased in the period following the 23-hour measurement. The figure below illustrates this data.  Note that the data illustrated in this figure do not reflect adjustments for carry-over effects.


 

 

 


Note that in the data set illustrated in the figure above, which does not attempt to adjust for carry-over effect, the post-23-hour increase in FEV1 is seen to a small degree in the placebo group, although the effect was much more pronounced in the drug treated groups, particularly at doses above 10mcg. The figures below, using adjustments for carry-over effects (either Test Day 1 only data, or a data set that excludes test days following test days in which doses of tiotropium greater than 10mcg were given), suggest that this phenomenon is not seen with placebo and is a drug-related finding.

 


 


 

 

 

 

 


-        The serial FEV1 data suggest a dose-response effect in the dose range of 10mcg to 40mcg.  The 80mcg dose does not seem to provide added benefit above the 40mcg dose.

-        The incidence of adverse events was comparable across the five treatment groups.  There was no evidence of systemic anticholinergic effects (dry mouth, increased heart rate).  Increases in systolic and diastolic blood pressure were noted in all treatment groups, including placebo.  However, carry-over effects could not be ruled out.

 

·        205.139: “Dose ranging study of Ba 679 BR inhalation powder following single inhalation in COPD patients.” (Report #U00-0156)

-        This was a randomized, placebo-controlled, four-period, cross-over study performed in Japan between July 27, 1998, and May 22, 1999 [U00-0156.pdf/p10].  A total of 27 patients with COPD received the following doses of tiotropium inhalation powder, using the HandiHaler device: 11.3mcg, 22.5mcg, 45mcg, or placebo.  Note: The Applicant states that the labeling method for tiotropium inhalation powder differs in Japan.  The doses labeled 11.3mcg, 22.5mcg, and 45.0mcg in Japan are equivalent to the doses labeled 9mcg, 18mcg, and 36mcg elsewhere [U00-0156.pdf/p28].  Twenty-four hour serial spirometry was performed at each dose level.  The duration of the washout period between doses was ³ 7 days.  For inclusion into the study, patients with COPD were required to demonstrate reversible airway obstruction, defined as a >10% improvement in FEV1 at 1 hour after inhalation of an anticholinergic agent (TersiganÒ Aerozol).  The primary endpoint was the peak FEV1.  Secondary endpoints included FEV1 AUC0-24h,  time to peak FEV1, time to response (defined as an increase in FEV1 of ³15%).

-        Carry-over effects were not observed [U00-0156.pdf/p86].  However, the drug was detected in some urine samples before dosing [U00-0156.pdf/p84].  Peak FEV1 was significantly higher in all active treatment groups, as compared with placebo.  A dose response effect was demonstrated for peak FEV1 and FEV1 AUC0-24hours.  Although the incremental improvement in peak FEV1 between the 22.5mcg dose and the 45mcg dose was minimal, the increment in FEV1 AUC0-24hours was more apparent [U00-0156.pdf/p68,70].  A significant dose-response effect was not seen in regard to time to response or time to peak response [U00-0156.pdf/p71].  No safety concerns were reported (adverse events, laboratory measurements, vital signs, oxygen saturation, ECG).

-       

The serial FEV1 curves in other single-dose dose-ranging studies indicated a rise in the FEV1 at 24 hours (see discussions above).  In this study a similar phenomenon was demonstrated.  This effect was seen in all groups, including placebo, suggesting that it may represent, in part, a normal circadian variation.  However, the figure below suggests that the effect was greater in the active treatment groups, suggesting an element of drug effect [U00-0156.pdf/p74].

 


 

·        205.108: “Randomized, multiple-dose, double-blind, parallel group study to determine the optimal dose of Ba 679 BR Inhaled as a dry powder in patients with chronic obstructive pulmonary disease.” (Report #U96-3068)

-        This was a multicenter, randomized, double-blind, placebo-controlled, multiple-dose, parallel group study performed in the US between January 16, 1995, and September 19, 1995 [U96-3068.pdf/p24].  A total of 169 patients with COPD received one of the following doses of tiotropium inhalation powder (expressed as the tiotropium cation), using the HandiHaler device for the four-week treatment period: 4.4mcg, 8.8mcg, 17.6mcg, or 35.2mcg, or placebo.  Note: The doses of active drug expressed in terms of tiotropium bromide monohydrate are 5.5mcg, 11mcg, 22mcg, and 44mcg.  Study medication was dosed once daily, at 12 noon.  Spirometry was conducted weekly at 8:00AM, 10:00AM, and 12 noon.  During the weekly visits during the treatment period, study drug was administered following the 12 noon spirometry, and serial spirometry was conducted hourly for six hours post-drug administration.   The primary variable was FEV1, “with emphasis on the last four hours of the dosing interval” [U96-3068.pdf/p32].  Secondary endpoints included FEV1 during the first six hours after the first dose and after multiple daily dosing at the end of each of the four weeks.

-        All doses were statistically more effective than placebo [U96-3068.pdf/p71].  No statistically significant differences were seen among doses.  The six-hour serial spirometry on the first treatment day shows evidence of a dose-response effect, however, the incremental benefit from the 17.6mcg and 35.2mcg doses is slight [U96-3068.pdf/p66].  The trough FEV1 data following multiple daily dosing indicates little consistent difference among the doses in the range of 4.4mcg to 17.6mcg [U96-3068.pdf/p67].  The trough FEV1 for the 35.2mcg dose is consistently higher than the other doses.  The Applicant fitted a maximum efficacy (Emax) model to the dose-response data including all trough FEV1 measurements from Week 2 onward [hpsum.pdf/p52].  In this model, the 8.8mcg dose provided 75%, the 17.6mcg dose provided 86%, and the 35.2mcg dose provided 92% of the maximum effect.

-        There were no dose-dependent increases in the incidence or severity of any adverse event [U96-3068.pdf/p94].  Dry mouth was the only event that appeared to be drug-related.

 

 

The four studies summarized above utilized either an inhalation solution or an inhalation powder formulation.  The following study examined dose-ranging using an inhalation solution formulation and one dose level of an inhalation powder formulation.

 

·        205.127: “Pharmacodynamic and pharmacokinetic dose ranging study of tiotropium bromide administered via Respimat device in patients with chronic obstructive pulmonary disease (COPD): A randomized, 3-week, multiple-dose, placebo-controlled, intraformulaiton double-blind, parallel group study.” (Report #U00-0077)

-        This was a multicenter, randomized, double-blind, placebo-controlled, multiple-dose, parallel group study performed in the France between 1998 and 1999 [U00-0077.pdf/p18].  A total of 202 patients with COPD received one of the following doses of tiotropium inhalation solution, using the Respimat device: 1.25mcg, 2.5mcg, 5mcg, 10mcg, or 20mcg, or tiotropium inhalation powder 18mcg using the HandiHaler device, or placebo.  The treatment period was 3 weeks.  Study medication was dosed once daily, between 8:00AM and 10:00AM.  Spirometry was conducted at each weekly visit at: 120, 60, and 5 minutes prior to dosing, immediately following dosing, and at 60, 120, 180, and 240 minutes after dosing.   The primary variable was FEV1,at Day 23,  “with emphasis on the last two hours of the dosing interval” [U00-0077.pdf/p47].  Secondary endpoints included FEV1 during the first four hours post-dose.

-        Trough FEV1 data (defined as the mean of the three pre-dosing values) from Day 7, Day 14, and Day 21 did not suggest a consistent dose-response effect for the Respimat groups [U00-0077.pdf/p62].  The trough FEV1 was consistently higher in the 18mcg HandiHaler group than in the other treatment groups.  Interestingly, the placebo response was consistently greater in the Respimat placebo as compared to the HandiHaler placebo.

-        Dry mouth appeared to be drug-related, and occurred more frequently in the higher dose groups [U00-0077.pdf/p86].

 

 

The COPD efficacy dose-ranging studies summarized above were submitted, in part, to support the proposed dose, which is 18mcg QD.  They are somewhat difficult to interpret for this purpose because of several factors.  These factors include inadequate washout periods in crossover studies, different formulations and delivery devices used, differences in the actual drug content due to changes in labeling conventions (See Section III, A above), and non-blinded dosing (in one case).  The only COPD dose-ranging study that used the proposed HandiHaler device was the single-dose study from Japan.  The only multiple-dose, dose-ranging study utilized the Inhalator Ingelheim (FO2) device, rather than the HandiHaler.   Nonetheless, these studies generally demonstrate a dose-response pharmacodynamic relationship.  The added efficacy benefit of the highest dose examined was small or non-existent.  The single-dose DPI study that used a 7-day washout, and the multiple-dose DPI study supported suggested that a dose of approximately 18mcg was superior to lower doses, and nearly as effective as a dose of approximately 36mcg.  This would support the proposed dose of 18mcg.


 

 

2.                  Tolerability Dose Ranging

Seven human pharmacology studies were performed to assess the pharmacodynamic properties and tolerability of tiotropium, in relation to dose in healthy volunteers.  These included various formulations routes of administration (inhalation powder in Studies 205.102, 205.104, and 205.104, inhalation solution in Studies 205.101 and 205.112, oral ingestion in study 205.106, and IV infusion in Study 205.107) [hpsum.pdf/p14].  Two of the five inhalation studies evaluated single dose administration and three of the five evaluated multiple dose administration.  The single-dose inhalation studies examined doses up to 281.6mcg and the multiple-dose inhalation studies used doses up to 140.8mcg.  In these studies, no effects were noted on pupil diameter, vital signs, ECG, or clinical laboratory tests [hpsum.pdf/p15].  Dose-related reports of dry mouth and reductions in salivary secretion were noted after multiple daily doses of 70.4 and 140mcg of the inhalation powder and after 32mcg of the inhalation solution from the RESPIMAT device.  Reports of dry mouth and taste perversion were dose-related.  Dry mouth was reported in 60-100% of subjects receiving multiple daily doses of 32 to 142mcg, and was reported in 0-22% of subjects receiving 8 to 17.6mcg.  Taste perversion was reported in 17-83% of subjects following single doses of ³40mcg, and was not reported at lower single doses. After multiple daily dosing, taste perversion was reported by up to 83% of subjects, in a dose-dependent fashion.  Dry mouth was not reported in the IV dosing studies.  These observations in healthy volunteers were considered in dose selection [hpsum.pdf/p54].  The excessive incidence of dry mouth at doses at and above 32mcg suggested that a lower dose would be preferable.

 

In the dose-ranging studies performed in COPD patients, no drug effects were seen in regard to vital signs, ECG, or clinical laboratory values.  With the exception of dry mouth, adverse events were comparable across all treatments, including placebo.  Dry mouth was not observed in the single-dose studies.  In the multiple-dose studies, 5.2% of patients reported dry mouth, with an onset ranging from 1 to 29 days (mean 10.6 days, median 3 days) and duration of 8 to 52 days (mean 29.7days, median 28 days) [hpsum.pdf/p16].  The time to onset and duration of this adverse effect did not appear dose-related.  Taste perversion was not reported in the COPD dose-ranging studies.

 

3.                  Pharmacologic Properties Related to Possible Safety Concerns (Pupilary Effects)

Because of possible ocular effects of this drug, the Applicant performed a randomized, placebo-controlled, double-blind, parallel group study examining the effects of topical ocular administration of tiotropium (Study 205.138) [hpsum.pdf/p56-7].  A total of 48 healthy male volunteers participated in this study.  Six subjects received one of the following single doses of tiotropium in one eye: 0.02, 0.04, 0.08, 0.16, 0.28, or 0.4μg, and twelve subjects received placebo.  The Applicant indicates that pupil diameter, pupillary reflex, intraocular pressure, accommodation, vital signs, and clinical laboratory values did not reveal any clinically relevant, drug-induced changes.

 

4.                  Onset of Pharmacodynamic Steady State

The onset of pharmacodynamic steady state was examined Study 205.129 (Report #U99-1072), which was performed in a subset of subjects in one of the one-year, double-blind, ipratropium-controlled, parallel-group studies (Study 205.122A/205.126A, reviewed in Section XI of this document) [hpsum.pdf/p57].  In this sub-study, 31 subjects (25 men, 6 women; n=20 treated with tiotropium and n=11 treated with ipratropium) underwent more frequent spirometry than was required in Study 205.122A/205.126A [U99-1072.pdf/p16].  Additional spirometry was performed on one hour prior to and just prior to dosing, and at 30, 60, 120, 180, 240, 300, and 360 minutes post-dosing on Days 1, 2, 3, 8, and 50.  After completion of the six-hour post-dose serial spirometry, the subjects inhaled 2 puffs of ipratropium or placebo and additional pulmonary function tests were conducted at 30, 60, and 120 minutes after this.  Of the 31 randomized subjects, only the 28 subjects with complete data were used in the efficacy analysis [U99-1072.pdf/p42]. 

 

As demonstrated in the table below, data for the trough, peak, and average FEV1 indicate that the maximum effect (“steady state”) was achieved on Day 8, and remained stable at Day 50.

 

Study 205.129: Mean (SE) FEV1 Trough, Peak, and Average Response (Liters) (Completers Data Set)                                             [U99-0172.pdf/p48]

Response

Test Day

Tiotropium

(N=17)

Ipratropium

(N=11)

Trough

Baseline

2

3

8

50

1.04 (0.09)

0.17 (0.03)

0.14 (0.03)

0.19 (0.02)

0.19 (0.04)

1.07 (0.12)

0.05 (0.03)

0.05 (0.06)

0.00 (0.07)

0.06 (0.08)

Peak

Baseline

2

3

8

50

0.35 (0.02)

0.40 (0.03)

0.35 (0.03)

0.37 (0.02)

0.39 (0.04)

0.33 (0.04)

0.33 (0.06)

0.36 (0.06)

0.33 (0.08)

0.34 (0.04)

Average

Baseline

2

3

8

50

0.27 (0.02)

0.30 (0.03)

0.25 (0.03)

0.29 (0.02)

0.28 (0.04)

0.20 (0.03)

0.23 (0.06)

0.22 (0.05)

0.20 (0.06)

0.22 (0.06)

 

 

Daily AM PEFR reached maximum effect (“steady state”) at Day 6.



V.               Description of Clinical Data and Sources 

 

A.                 Overall Data

The clinical data submitted in support of this NDA are derived from the studies performed as part of the Applicant’s clinical development program.  The application does not rely on reports in the medical literature or other sources of data.

 

B.        Tables Listing the Clinical Trials

The clinical program submitted in support of efficacy included six “pivotal” studies and five “supportive” studies [S8/ise.pdf/p88].  These are summarized in the two tables below.



 

Summary of Pivotal Studies

Study Number (Report #)

Study Type

Treatment Groups

Location

Duration

Design

Number of Subjects

Primary Endpoint

205.114/

205.117

(U99-3169)

Safety/ Efficacy

Tiotropium 18mcg capsule QD

Placebo Capsule QD

US

1 year

(49 weeks)

R, DB, PC, PG

470

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.115/

205.128

(U99-3170)

Safety/ Efficacy

Tiotropium 18mcg capsule QD

Placebo Capsule QD

US

1 year

(49 weeks)

R, DB, PC, PG

451

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.122A/

205.126A

(U00-3113)

Safety/ Efficacy

Tiotropium 18mcg capsule QD +           Placebo MDI QID

Placebo capsule QD +              Ipratropium MDI 40mcg QID 

Netherlands

1 year

(52 weeks)

R, DB, PG

Active comparator

288

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.122B/

205.126B

(U00-3114)

Safety/ Efficacy

Tiotropium 18mcg capsule QD +           Placebo MDI QID

Placebo capsule QD +              Ipratropium MDI 40mcg QID         

Netherlands and Belgium

1year

(52 weeks)

R, DB, PG

Active comparator

247

Trough FEV1 response at 13 weeks               (mean of values at 23 and 24 hours)

205.130

(U01-1236)

Safety/ Efficacy

Tiotropium 18mcg capsule QD +    Placebo MDI BID

Placebo capsule QD +               Salmeterol MDI BID

Placebo capsule QD +                    Placebo MDI BID

Multinational

6 months

R, DB, PC

Active comparator

623

TDI focal score (responder analysis)

AND

Trough FEV1  Response

205.137

(U01-1231)

Safety/ Efficacy

Tiotropium 18mcg capsule QD +     Placebo MDI BID

Placebo capsule QD +                Salmeterol MDI  BID

Placebo capsule QD +                    Placebo MDI BID

Multinational

6 months

R, DB, PC

Active comparator

584

TDI focal score (responder analysis)

AND

Trough FEV1  Response


Supporting Studies

Study #

Country/ Dates

Design

Treatments

(Tiotropium)

Device

Duration

# of Subjects

Population

Primary Endpoint

205.119

Netherlands  11/91-4/92

Dose-ranging

Open label

XO

10mcg

20mcg

40mcg

80mcg

160mcg

RESPIMAT

Single Dose

6

(2F/ 4M)

COPD

FEV1

205.120

Netherlands 10/92-5/93

Dose-ranging

R, DB, PC, XO

10mcg

20mcg

40mcg

80mcg

Placebo

INHALATOR INGELHEIM (FO2)

Single Dose

35

(3F/ 32M)

COPD

FEV1

205.139

Japan   7/98-5/99

Dose-ranging

R, DB, PC, XO

11.3mcg1

22.5mcg1

45mcg1

Placebo

HANDIHALER

Single Dose

27

COPD

FEV1

205.108

US       1/95-9/95

Dose-ranging

Multicenter, R, DB, PC, PG

4.4mcg2 QD

8.8mcg2 QD

17.6mcg2 QD

35.2mcg2 QD

Placebo QD

INHALATOR INGELHEIM (FO2)

4 Weeks

169

(73F/ 96M)

COPD

FEV1

205.123

UK      5/97-7/98

AM/PM Dosing

Multicenter, R, DB, PC, PG

18mcg QAM

18mcg QPM

Placebo QAM

Placebo QPM

HANDIHALER

6 Weeks

121

(46F/ 75M)

COPD

FEV1

 

 

C.        Postmarketing Experience

There are no postmarketing data available because the drug has not been marketed in any country [summary.pdf/p43].

 

 


VI.            Clinical Review Methods

 

A.        How the Review was Conducted

The six studies that were designated by the Applicant as “pivotal” studies were reviewed individually in-depth in regard to study design issues and efficacy conclusions.  These in-depth reviews may be found in the Appendix to this Clinical Briefing Document.  Safety data from the individual studies were reviewed less rigorously.  Rather, the safety assessment was primarily derived from the integrated safety data provided in the Applicant’s Integrated Summary of Safety.  Individual pharmacokinetic and pharmacodynamic studies were reviewed primarily for evidence to support the proposed dose and dosing interval. 

 

B.        Overview of Materials Consulted in Review

This Clinical Briefing Document is based on the materials submitted in the original NDA submission, the 120-Day Safety Update, and the various amendments submitted by the Applicant either on its own initiative or in response to the Division’s requests for specific information.  These amendments are listed on the first page of this Review.

 

C.        Overview of Methods Used to Evaluate Data Quality and Integrity

The Division of Pulmonary and Allergy Drug Products requested that the Agency’s Division of Scientific Investigations perform an audit of two clinical centers.  The clinical centers to be audited were chosen based on participation in Study 205.130 or 205.137 (the two studies submitted in support of the dyspnea claim), number of subjects enrolled, and the magnitude of benefit reported in regard to the TDI.  Two large US centers that reported greater benefit of study drug were selected.  DSI has concluded that one of the two study sites adhered to all pertinent federal regulations and/or good clinical investigational practices governing the conduct of clinical investigations and the protection of human subjects.  At the second study site, which enrolled 13 patients into Study 205.130, one potentially important protocol violation was noted. At this site, the TDI questionnaire was improperly administered.  Rather than having study site personnel ask questions of the patient and complete the questionnaire, the patients themselves read the questionnaire and completed the form.  This is not the validated method of administration. A review of the case report forms by the DSI Inspector indicated that this may have caused some confusion for the patients, potentially impacting the validity of the scoring.  One patient made several significant corrections to his/her answers, two patients provided divergent descriptions of their status in the TDI compared with the SGRQ.  Because this was a large, multicenter study, this finding at a single study center is unlikely to impact the conclusions of the study. However, it must be recognized that this type of protocol violation may have occurred at additional study centers, which were not audited.


 

D.        Were Trials Conducted in Accordance with Accepted Ethical Standards

The Applicant has indicated that all clinical trials were conducted in accordance with accepted ethical standards [gcp.pdf].

 

E.         Evaluation of Financial Disclosure

Section 19 of the NDA addresses the Applicant’s compliance with the Final Rule on Financial Disclosure by Clinical Investigators.  The Applicant notes that, as a privately held company, it has no equity available to investigators and does not provide compensation to investigators based on the outcome of studies conducted on its behalf.  In addition, no investigators can have or own a proprietary interest in a product, trademark, licensing agreement or patent owned by the company.  The Application contains a signed FDA Form 3454 for each of the six “pivotal” clinical studies.  These forms certify that the Applicant did not enter into financial arrangements with any investigator whereby the value of compensation could be affected by the outcome of the study, than none of the investigators disclosed a proprietary interest in the product or a significant equity interest in the Sponsor, and that no investigator received significant payments of other sorts, as defined in 21 CFR 54.2 (f).  One investigator in Study 205.130 was reported to be involved in a financial arrangement with the Applicant.  The Applicant states that because payment was made in August, 1998, prior to the FDA Regulation date February 2, 1999, no form 3455 is submitted [financial.pdf/p13].  Based on this information, as well as the multi-center nature of the pivotal clinical studies, it is unlikely that financial interests could have influenced or biased the results of these studies.


 


VII.        Integrated Review of Efficacy

 

A.                 Brief Statement of Conclusions

The evidence derived from the six pivotal clinical trials appears to establish the efficacy of tiotropium as a bronchodilator in patients with COPD.  The data regarding the effect of this drug on the symptom of dyspnea in this patient population is less convincing.  These are the subject matter for discussion at the September 6, 2002, PADAC meeting.

 

The pharmacodynamic properties of tiotropium are unusual for an orally inhaled drug. As discussed in the Human Pharmacokinetics and Pharmacodynamics section of this document, the bronchodilator effect seen after a single dose increases with multiple daily dosing, reaching “steady state” by Day 8.  The text and figures used to illustrate the pharmacodynamic properties of tiotropium in the product label should capture this feature. 

 

 

 

B.                 General Approach to Review of the Efficacy of the Drug

Conclusions regarding the efficacy of tiotropium bromide inhalation powder (18mcg QD) were developed following detailed review of the efficacy findings of each of the individual pivotal Phase 3 studies.  There were six such studies, as outlined in the table below.  These studies included two one-year placebo-controlled studies (205.114/205.117 and 205.115/205.128), two six-month placebo- and active-controlled studies (205.130 and 205.137), and two one-year active-controlled studies (205.122A/205.126A and 205.122B/205.126B).

 

Pivotal Clinical Studies

Study Number (Report #)

Study Type

Treatment Groups

Location

Duration

Design

Number of Subjects

Primary Endpoint

205.114/

205.117

(U99-3169)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD

-         Placebo Capsule QD

US

1 year

(49 weeks)

R, DB, PC, PG

470

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.115/

205.128

(U99-3170)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD

-         Placebo Capsule QD

US

1 year

(49 weeks)

R, DB, PC, PG

451

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.122A/

205.126A

(U00-3113)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD +           Placebo MDI QID

-         Placebo capsule QD + Ipratropium MDI 40mcg QID 

Netherlands

1 year

(52 weeks)

R, DB, PG

Active comparator

288

Trough FEV1 response at 13 weeks              (mean of values at 23 and 24 hours)

205.122B/

205.126B

(U00-3114)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD +           Placebo MDI QID

-         Placebo capsule QD + Ipratropium MDI 40mcg QID         

Netherlands and Belgium

1year

(52 weeks)

R, DB, PG

Active comparator

247

Trough FEV1 response at 13 weeks               (mean of values at 23 and 24 hours)

205.130

(U01-1236)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD + Placebo MDI BID

-         Placebo capsule QD +SalmeterolMDI BID

-         Placebo capsule QD + Placebo MDI BID

Multinational

6 months

R, DB, PC

Active comparator

623

TDI focal score (responder analysis)

AND

Trough FEV1  Response

205.137

(U01-1231)

Safety/ Efficacy

-         Tiotropium 18mcg capsule QD +     Placebo MDI BID

-         Placebo capsule QD + Salmeterol MDI  BID

-         Placebo capsule QD +  Placebo MDI BID

Multinational

6 months

R, DB, PC

Active comparator

584

TDI focal score (responder analysis)

AND

Trough FEV1  Response

 

Currently approved medications for COPD are indicated for the relief of bronchospasm due to COPD.  As such, the basis for approval of these drugs has been adequate and well controlled studies demonstrating bronchodilator efficacy.  Consistent with this traditional approach, all of the pivotal clinical studies in this NDA specified as the primary (or co-primary) variable an established measure of bronchodilator activity (FEV1).  In addition, numerous secondary variables supporting bronchodilator activity were employed.  The unique aspect to this NDA is that the Applicant has proposed that this drug be labeled for the treatment of dyspnea as well as bronchospasm due to COPD.  In order to support this proposal, the primary endpoints of two of the pivotal studies were changed after study completion but prior to un-blinding (Studies 205.130 and 205.137).  The co-primary variables for these studies were FEV1 and an index of subjective dyspnea, the Mahler Transitional Dyspnea Index.  This Integrated Review of Efficacy will discuss the efficacy findings of the pivotal clinical studies in regard to the bronchodilator efficacy of the drug and in regard to putative effects on subjective dyspnea.   

 

 

 

C.                 Detailed Review of Trials by Indication

 

1.                  Data Addressing Bronchodilator Efficacy

 

ONE-YEAR, PLACEBO-CONTROLLED STUDIES

Two, nearly identical, large, randomized, double-blind, placebo-controlled, parallel group studies examined the safety and efficacy of tiotropium versus placebo administered for approximately 1 year (49 weeks) (Study 205.114/205.117 and Study 205.115/205.128).  These two studies differed only in that the former included pharmacokinetic assessments, whereas the latter did not.  Detailed reviews of these studies are located in the Appendix to this Clinical Briefing Document.  In these studies, a total of 921 patients with COPD were, following a 2-week baseline period, randomized to receive tiotropium or placebo once daily in the morning.  Eligible patients had a history of COPD, a smoking history of ³10 pack-years, age ³40 years, and FEV1 £65% of predicted and £70% of FVC.  Baseline bronchodilator reversibility was not assessed.  Spirometry was performed at baseline, and after 1, 7, 13, 37, and 49 weeks of treatment.  On these test days spirometry was performed at one-hour prior to dosing, immediately prior to dosing, and at 30, 60, 120, and 180 minutes after dosing.  The pre-specified primary efficacy endpoint was the “trough FEV1 response” at the end of the first 13 weeks of treatment.  The “trough FEV1 response” was defined as the change from baseline in the mean of the two FEV1 values at the end of the dosing interval (approximately 23 and 24 hours post-dosing).  Secondary efficacy endpoints included the trough FEV1 response at other timepoints, the average and peak FEV1 response for the first 3-hours post-treatment on each test day, individual FEV1 and FVC values, weekly mean PEFR measured by the patient at home twice daily, physician’s global evaluation, COPD symptom scores (wheezing, shortness of breath, coughing, and tightness of chest), rescue albuterol use, number of nocturnal awakenings during the first 13 weeks, number and length of COPD exacerbations and hospitalizations for respiratory disease, the Saint George’s Respiratory Questionnaire (SGRQ), and pharmacoeconomic variables.

 

Most of the patients in these studies were White (91.9% and 96.7%), and the majority were men (66.7% and 66.4%).  At screening, these patients had a mean FEV1 of approximately 1 liter, and a ratio of FEV1/FVC of approximately 45%.

 

Primary Endpoint: Trough FEV1 Response (liters), Week 13 (Studies 205.114/205.117 and 205.115/205.128)

Study

Tiotropium

Placebo

p-value

205.114/205.117

0.11

-0.03

0.0001

205.115/205.128

0.13

-0.01

0.0001

 

 

Both of these studies demonstrated that tiotropium was superior to placebo on the pre-specified primary endpoint, trough FEV1 response after 13 weeks of treatment (p=0.0001).  The mean trough FEV1 response in the tiotropium group was 0.11 liters (compared with –0.03 liters in the placebo group) in Study 205.114/205.117, and 0.13 liters (compared with –0.01 liters in the placebo group) in Study 205.115/205.128.  These data indicate that tiotropium has a statistically significant bronchodilator effect at the end of the proposed dosing interval.  It should be noted that the Division has not previously taken a position regarding the magnitude of effect that would be considered to be clinically meaningful for the end-of-dosing interval FEV1.  In assessing acute bronchodilator efficacy, a threshold of at least 12% and at least 200ml is commonly used to determine a clinically meaningful bronchodilator effect.  However, it would not seem reasonable to use this threshold for the end of the dosing interval.  Thus, the analysis of the primary endpoint established that the bronchodilator effect of tiotropium remains statistically significant at the end of the dosing interval.  The magnitude of that effect is small compared to what would be expected if this measure were taken at peak effect, but is probably clinically meaningful at the end of the dosing interval.

 

Secondary spirometry endpoints included trough FEV1 response after 1, 7, 25, 37, and 49 weeks of treatment.  At each of these timepoints, tiotropium was statistically superior to placebo (p=0.0001), with effect sizes (tiotropium minus placebo) of 0.11 to 0.16 liters.  These data further support the conclusions regarding end-of-dosing interval efficacy that were drawn from the primary efficacy endpoint analysis.

 

Insight into the early post-dose bronchodilator effect of tiotropium can be drawn from the 3-hour serial spirometry performed on each test day.  In both studies, tiotropium was superior to placebo in regard to the mean average FEV1 response during the 3-hour serial spirometry, on all test days (p=0.0001).   Because this parameter is an average of several spirometry measures, interpretation of the effect size is less intuitive. 

 

Perhaps more helpful is the information derived from the analyses of the peak FEV1 data.  In both studies, tiotropium was superior to placebo in regard to the mean peak FEV1 response on all test days (p=0.0001).  However, the mean treatment effect size (i.e. tiotropium effect minus placebo effect) was small, ranging from 0.15 liters on test day 1, to 0.19-0.22 liters on subsequent test days.  It should be noted that in assessing for what is considered a clinically meaningful degree of bronchodilation (using the threshold of 12% and at least 200ml), it is not customary to consider placebo responses.  Thus, the absolute increase in FEV1, without subtraction of placebo effect, is customarily used.  In these studies, the mean peak FEV1 response was 0.24 liters on test day 1, and ranged from 0.25 to 0.31 liters on subsequent test days.  This would support the assertion that, despite the relatively small difference between tiotropium and placebo, tiotropium is associated with a clinically meaningful degree of bronchodilation on all test days. 

 

One further insight into the pharmacodynamics of tiotropium can be obtained from the peak FEV1 data.  While the mean peak FEV1 on test day 1 was 0.24 liters in the tiotropium groups, the mean peak FEV1 at each of the four individual test day 1, post-dose assessments was <0.20 liters.  This unusual circumstance is due to the fact that patients reached their personal peak FEV1 values at differing time points (see table below). 

 

Percentage of Patients Who Reached Their Peak FEV1 at Each Timepoint (Test Day 1)                                                                                   [Submission dated 7/16/02; page 8]

Timepoint

Tiotropium

Placebo

 

205.114/205.117

205.115/205.128

205.114/205.117

205.115/205.128

30 minutes

1 hour

2 hours

3 hours

14.7%

20.4%

29.7%

35.1%

18.8%

19.2%

29.2%

32.8%

26.2%

25.1%

26.7%

22.0%

30.0%

25.0%

19.4%

25.6%

 

Other measures of pulmonary function also supported the bronchodilator efficacy of  tiotropium. In both studies, tiotropium was also statistically significantly superior to placebo for the trough, average, and peak FVC responses on all test days.  The FVC data from both studies suggested that the bronchodilator efficacy increased between Day 1 and Day 8.  Daily morning and evening peak flow measurements were performed and recorded by the patients.  For the morning peak flow measurements, tiotropium was statistically superior to placebo during approximately one-half of the weeks in one study (205.114/205.117), and during nearly all of the weeks in the other, with effect sizes ranging from 8 to 31 liters/minute.  Tiotropium was statistically superior to placebo in regard to evening peak flow measurements, with effect sizes ranging from 13 to 40 liters/minute.

 

Other evidence in support of the efficacy of tiotropium as a bronchodilator includes the reported use of as-needed supplemental albuterol.  During each week of treatment, patients in the tiotropium group used statistically significantly fewer doses of as-needed albuterol.  On average, patients in the tiotropium group used approximately 5-6 fewer doses of albuterol per week, compared with patients in the placebo group.  Although in one study (205.114/205.117) patients in the tiotropium group reported statistically fewer nocturnal awakenings due to COPD symptoms during 7 of the 13 weeks this was assessed, in the second study, no effect on this variable was seen.

 

Despite the efficacy of tiotropium as a bronchodilator, in both studies there was no difference between tiotropium and placebo in regard to the number of patients with COPD exacerbations, time to COPD exacerbation, number of COPD exacerbation days, number of patients with hospitalization due to COPD, or number of hospitalizations due to COPD.

 

The studies also included two health-related quality of life assessments, the “disease-specific” St. George’s Hospital Respiratory Questionnaire (SGRQ) and the non-disease specific Medical Outcomes Study SF-36.  Differences between groups rarely reached the generally accepted threshold for a minimal clinically meaningful effect on the SGRQ, which was administered at baseline, and after 7, 13, 25, 37, and 49 weeks of treatment.  The study reports did not describe analyses of the total SF-36 scores.  Results for the “physical health” domains within the SF-36 were not consistent between studies.

 

Finally, in both studies the scores on the Physician’s Global Evaluation were statistically superior in the tiotropium group on all test days.  However, the clinical significance of the treatment effect seen (0.25 to 0.59 on a scale of 1-8) is not known.

 

ONE-YEAR, ACTIVE-CONTROLLED STUDIES (205.122A/205.126A and 205.122B/205.126B)

Two, identical, large, randomized, double-blind, active-controlled, parallel group studies examined the safety and efficacy of tiotropium (QD) versus ipratropium bromide (QID) administered for approximately 1 year (52 weeks) (Study 205.122A/205.126A and Study 205.122B/205.126B).  Detailed reviews of these studies are located in the Appendix to this Clinical Briefing Document.  In these studies, a total of 535 patients with COPD were, following a 2-week baseline period, randomized to receive either tiotropium inhalation capsules QD or ipratropium bromide MDI QID (2:1 randomization).  Eligible patients had a history of COPD, a smoking history of ³10 pack-years, age ³40 years, and FEV1 £65% of predicted and £70% of FVC.  Baseline bronchodilator reversibility was not assessed.  Spirometry was performed at baseline, and after 1, 7, 13, 26, 39, and 52 weeks of treatment.  On test days during the first 13 weeks, spirometry was performed at one-hour prior to dosing, immediately prior to dosing, and at 30, 60, 120, 180, 240, 300, and 360  minutes after dosing.  On the remaining test days, the serial spirometry ended after the 180-minute measure.  The pre-specified primary efficacy variable was the “trough FEV1 response,” defined as the change from baseline in the mean of the two FEV1 values at the end of the dosing interval (approximately 23 and 24 hours post-dosing).  The protocol did not state which specific treatment visit would serve as the primary efficacy endpoint.  Secondary efficacy endpoints included the average and peak FEV1 response for the first 6-hours post-treatment at Weeks 1, 7, and 13, and the first 3-hours post treatment on the remaining test days, individual FEV1 and FVC values, weekly mean PEFR measured by the patient at home twice daily, an Energy-Fatigue Questionnaire, rescue albuterol use, the Saint George’s Respiratory Questionnaire (SGRQ), and pharmacoeconomic variables.

 

The great majority of patients were men (81.7% and 87.3%, in each study), and all patients except one were white.  The baseline mean FEV1 for all patients was approximately 1.2 liters, with an FEV1/FVC ratio of approximately 45%.

 

Before discussing the efficacy results of these studies, two issues should be noted.  First, in these studies the primary efficacy variable (trough FEV1) was determined at a timepoint at which the active comparator, based on its known pharmacodynamic properties, would not be expected to be effective.  The active comparator, ipratropium bromide, is indicated for use four times daily.  Given the relatively long interval between the evening dose and subsequent morning dose of ipratropium, little if any bronchodilator effect is likely to be detected on morning pre-dose spirometry. The second issue is that, for US regulatory purposes, a new drug does not need to demonstrate superiority over existing drugs.  Therefore, although the primary endpoint may be intrinsically biased to favor a longer-acting drug over a shorter-acting drug, in this circumstance, for regulatory decision-making, the ipratropium treatment group may be considered analogous to placebo.  Presuming that treatment with ipratropium has no detrimental effect in terms of COPD efficacy endpoints, demonstrated superiority over ipratropium may be construed as superiority over placebo.

 

In both studies, tiotropium was statistically superior to ipratropium for the trough FEV1 response at all test days.  The difference in trough FEV1 response between groups ranged from 0.11 liters to 0.18 liters.

 

Because these studies did not include a placebo treatment group, the post-dosing serial spirometry offer little data relevant to regulatory decision-making.  This data will not be discussed further here, but is discussed for each study in the Appendix to this document.  Home morning and evening PEFR values were statistically superior in the tiotropium group during all of the weeks of one study (205.122B/205.126B), and during most of the weeks in the other.  The effect sizes for these measures were variable.

 

The use of as-needed albuterol was not different between groups in one study (205.122B/205.126B), and was statistically lower in the tiotropium group for 36 of the 52 weeks of the other study.  The effect on COPD exacerbations was not consistent. In one study (205.122A/205.126A), no difference between groups was observed in regard to the number of patients with COPD exacerbation, time to first COPD exacerbation, number of COPD exacerbations, number of COPD exacerbation days, number of patients with hospitalizations due to COPD, or number of hospitalization days for COPD.  However, in the second study, the tiotropium group had significantly fewer subjects with COPD exacerbations, fewer COPD exacerbations, and fewer COPD exacerbation days.  Also in that study, the time to first COPD exacerbation was longer in the tiotropium group.  Hospitalizations due to COPD were not different.

 

The studies also included two health-related quality of life assessments, the “disease-specific” St. George’s Hospital Respiratory Questionnaire (SGRQ) and the non-disease specific Medical Outcomes Study SF-36.  Differences between groups rarely reached the generally accepted threshold for a minimal clinically meaningful effect on the SGRQ, which was administered at baseline, and after 7, 13, 25, 37, and 49 weeks of treatment.  The SF-36 did not demonstrate statistical differences between groups.  Finally, the studies included a three-question “Energy Fatigue Questionnaire,” administered on test days 8, 50, 92, 182, 273, and 364.  There were no statistically significant differences between groups on this questionnaire.

 

 

SIX-MONTH PLACEBO- AND ACTIVE-CONTROLLED STUDIES (205.130 and 205.137)

Two, nearly identical, large, randomized, double-blind, placebo- and active (salmeterol inhalation aerosol) controlled, parallel group studies examined the safety and efficacy of tiotropium versus placebo administered for six months (Study 205.130 and Study 205.137).  These two studies differed only in that the former included post-dosing serial spirometry for 12 hours after dosing, whereas the latter included 3-hour post-dosing serial spirometry.  Detailed reviews of these studies are located in the Appendix to this Clinical Briefing Document.  In these studies, a total of 1207 patients with COPD were, following a 2-week baseline period, randomized to receive either tiotropium (18mcg QD), salmeterol xinafoate inhalation aerosol (50mcg BID), or placebo.  Eligible patients had a history of COPD, a smoking history of >10 pack-years, age ³40 years, and FEV1 £60% of predicted and £70% of FVC.  Baseline bronchodilator reversibility was not assessed.  Spirometry was performed at baseline, and after 2, 8, 16, and 24 weeks of treatment.  On these test days spirometry was performed at one-hour prior to dosing, 10 minutes prior to dosing, and at 30minutes, 1, 2, 3, 4, 6, 8, 10, and 12 hours after dosing in Study 205.130.  In Study 205.137, post-dose serial spirometry included only 3 hours after dosing.  The pre-specified primary efficacy endpoints were the “trough FEV1 response”  and the focal score of the Mahler Transitional Dyspnea Index (TDI), at the end of the 24 weeks of treatment.  The “trough FEV1 response” was defined as the change from baseline in the mean of the two FEV1 values at the end of the dosing interval (approximately 23 and 24 hours post-dosing).  Secondary efficacy endpoints included the TDI focal score on other test days, the trough FEV1 response on other test days, the average and peak FEV1 response on each test day, individual FEV1 and FVC values, weekly mean PEFR measured by the patient at home twice daily, physician’s global evaluation, COPD symptom scores (wheezing, shortness of breath, coughing, and tightness of chest), rescue albuterol use, the shuttle walking test with Borg dyspnea rating scale, number and length of COPD exacerbations and hospitalizations for respiratory disease, the number of patients with at least one COPD exacerbation, the number of patients with at least one hospitalization for respiratory disease, the Saint George’s Respiratory Questionnaire (SGRQ), patient preference, and pharmacoeconomic variables.

 

The majority of patients were men (74.6% and 77.9% in Study 205.130 and 205.137, respectively), and nearly all patients were white (99.5%).  The mean age of the patients was approximately 64 years, and the mean screening FEV1 was approximately 1.1 liters.

 

Before discussing the efficacy results of these studies, one important issue should be noted.  Although these studies included three treatment arms, the pre-specified primary comparison was that of tiotropium versus placebo.  This is appropriate because, from the regulatory perspective, it is this comparison that is most important.  Therefore, this Integrated Review of Efficacy will focus on the comparison of tiotropium versus placebo.  However, the study report also discusses the comparison of tiotropium versus salmeterol.  In considering the findings of the tiotropium versus salmeterol comparison, one must keep in mind that one of the co-primary efficacy variables (trough FEV1) was determined at a timepoint at which the active comparator, based on its known pharmacodynamic properties, would not be expected to be effective.  The active comparator, salmeterol, is indicated for use twice daily.  Given the relatively long interval between the evening dose and subsequent morning dose of salmeterol, little if any bronchodilator effect is likely to be detected on morning pre-dose spirometry.

 

In both studies, tiotropium was statistically superior to placebo for the trough FEV1 response after 24 weeks of treatment.  The mean trough FEV1 response in the tiotropium group was 0.14 liters in Study 205.130, and 0.11 liters in Study 205.137.  These data indicate that tiotropium has a statistically significant bronchodilator effect at the end of the proposed dosing interval.  It should be noted that the Division has not previously taken a position regarding the magnitude of effect that would be considered to be clinically meaningful for the end-of-dosing interval FEV1.  In assessing acute bronchodilator efficacy, a threshold of at least 12% and at least 200ml is commonly used to determine a clinically meaningful bronchodilator effect.  However, it would not seem reasonable to use this threshold for the end of the dosing interval.  Thus, the analysis of this co-primary endpoint established that the bronchodilator effect of tiotropium remains statistically significant at the end of the dosing interval.  The magnitude of that effect is small compared to what would be expected if this measure were taken at peak effect, but is probably clinically meaningful at the end of the dosing interval.

 

Tiotropium was also statistically superior to placebo on each of the serial spirometry measurements on all test days in both studies.  Consistent with this, tiotropium was statistically superior to placebo in regard to the mean trough, average, and peak FEV1 response on all test days in both studies.  On the first test day, the mean peak FEV1 response was 0.31 liters (Study 205.130) and 0.27 liters (Study 205.137).  The difference between the tiotropium mean peak FEV1 response and the placebo mean peak FEV1 response was 0.19 and 0.16 liters on test day 1 in these two studies.  The serial spirometry FVC data was consistent with the FEV1 data.  The patient-recorded daily PEFR data also supported the efficacy of tiotropium as a bronchodilator.  In both studies, the mean weekly morning and mean weekly evening PEFR values were statistically superior in the tiotropium group, as compared to the placebo group.  The differences between tiotropium and placebo ranged from 14.9 to 27 liters/minute for the morning PEFR, and from 21 to 33 liters/minute for the evening PEFR.

 

Interestingly, in only one of the two studies was tiotropium statistically superior to placebo in regard to the number of puffs of as-needed albuterol used by the patients (Study 205.130).

 

In Study 205.130, there were statistically fewer COPD exacerbations and COPD exacerbation days in the tiotropium group as compared to placebo, but there was no statistically significant difference between these two groups in regard to the number of subjects with at least one COPD exacerbation.  In Study 205.137 there were no significant differences between tiotropium and placebo in regard to the various expressions of COPD exacerbations.  There were no notable differences between tiotropium and placebo in regard to hospitalizations for COPD in either study.

 

Tiotropium was statistically superior to placebo in regard to the Physician’s Global Evaluation on all test days except Week 24 in Study 205.137.  In regard to the total SGRQ scores, the difference between tiotropium and placebo did not reach the generally accepted threshold of a minimal clinically important difference (4 units) on any test day in either study.

 

SUMMARY OF BRONCHODILATOR EFFICACY RESULTS

Existing drugs for COPD are indicated for the relief of bronchospasm associated with COPD.  As such, the standard for approval has been demonstration, through adequate and well-controlled trials, of a bronchodilator effect.  The most commonly used index of bronchodilator effect has been the FEV1.  In most cases, the primary analyses of FEV1 have focused on peak changes.  In this application, the primary focus has been on the “trough FEV1 response.”  This endpoint has the benefit of incorporating important information regarding end-of-dosing-interval bronchodilator efficacy.  The limitation of this endpoint is that there is less experience and consensus regarding what constitutes a minimal clinically meaningful effect. 

 

In the six Phase 3 studies submitted with this application, tiotropium was statistically superior to placebo (or an active control that may be considered a proxy for placebo) in regard to the trough FEV1 response.  The treatment effect size on this endpoint, while less than what might be desired of a peak effect size, may be clinically significant.  Secondary analyses of serial spirometry during the early post-dosing period appear to demonstrate that tiotropium is statistically superior to placebo in regard to peak and average FEV1.  Analyses of the mean peak FEV1 values in the one-year, placebo-controlled studies suggest that the mean peak effect may be clinically meaningful.  It is interesting to note that the time to reach peak FEV1 seems to vary substantially among individual subjects.  Other secondary efficacy variables, such as home PEFR values and supplemental “as-needed” albuterol use, also appear to support the bronchodilator activity of tiotropium.  No consistent, clinically meaningful effect was demonstrated on other indicators of COPD disease activity, such as COPD exacerbations, COPD hospitalizations, and health-related quality of life assessments. 

 

LABELING ISSUES REGARDING BRONCHODILATOR EFFECT

The product labels for orally inhaled bronchodilators customarily provide information (text and figures) that describes the pharmacodynamic effect of the drug.  Typical information that is conveyed includes peak effect, time to peak effect, and duration of action.  For this drug, these concepts are not easy to convey.  One difficulty is the fact that the bronchodilator effect is not maximal after the initial dose.  While it is important to convey the single-dose performance characteristics in the label, because the drug would be indicated for chronic use (maintenance treatment), rather than as a “rescue” medication, it would be equally important to convey the performance characteristics expected with chronic use.  One difficulty conveying the chronic use characteristics is that, due to its demonstrated efficacy throughout the dosing interval, the pre-dose FEV1 after chronic use is greater than the pre-treatment baseline.  Thus, describing the bronchodilator effect as a change from pre-dose values would underestimate the actual clinical benefit.  Because of this issue, the text of the label should be fairly general in this regard, with figures used to illustrate the pharmacodynamic effects.  The Applicant proposes the following figures to convey this information [proposed.pdf/p5].

 


 


For these figures, the Applicant has pooled data from two studies.

 

The Applicant also proposes text that states that tiotropium provided significant improvements in lung function within 30 minutes following the first dose [proposed.pdf/p4].  The term significant when used in regard to bronchodilators usually indicates an improvement of 12% and at least 200ml in the FEV1.  In the two 1-year, placebo-controlled studies, while the mean peak FEV1 did increase by 240ml, the mean FEV1 did not increase by ³200ml at any of the timepoints during the 3-hour post-dose serial spirometry after the first dose.  This was because the time to peak FEV1 varied among individual patients.  In fact, at 30 minutes only 14.7% (Study 205.114/205.117) and 18.8% (Study 205.115/205.128) of patients in the tiotropium group had reached their peak FEV1. 

 

 

2.                  Data Addressing Efficacy in Regard to the Proposed Dyspnea Indication

The Applicant has proposed a unique indication for tiotropium, namely the relief of dyspnea related to COPD.  The primary support of this proposed indication is taken from the results of two studies for which an index of dyspnea (the TDI focal score) was pre-specified as one of two co-primary endpoints (Studies 205.130 and 205.137).  Supporting data may be drawn from other studies in which various indices of the symptom were captured as secondary endpoints.  In the following section of this document, the TDI instrument will be briefly summarized, and the findings of Studies 205.130 and 205.137 will be discussed, along with this additional supporting data. The studies cited in the discussion are reviewed in depth in the Appendix to this document and summarized briefly above.

 

            The Baseline/Transitional Dyspnea Index

The Baseline/Transitional Dyspnea Index (BDI/TDI) is a multidimensional index of the sensation of dyspnea.  Both the BDI and TDI consist of three components.  The components are “functional impairment,” “magnitude of task” (needed to evoke dyspnea), and “magnitude of effort” (needed to evoke dyspnea).  At baseline, each component is assigned a grade, ranging from 0 to 4.  The components may also be graded “W” for “amount uncertain,” “X” for “unknown,” or “Y” for “impaired for reasons other than shortness of breath.”  On subsequent visits, the TDI is administered, with each component assigned a score ranging from –3 (major deterioration) to +3 (major improvement).  A score of +1 indicates improvement within a BDI grade.  The TDI can also be recorded as “Z,” indicating that there was “further impairment for reasons other than shortness of breath.” Reviewer’s Note: For the purposes of the studies, any data recorded as “W,” “X,” “Y,” or “Z” was set to missing for the purposes of data analysis.  The TDI focal score, which consists of the sum of the three components, can thus range from –9 to +9.  The instrument is administered by an observer who has experience in taking a medical history regarding respiratory disease.  The interviewer asks open-ended questions about the patient’s experience of breathlessness and then selects a grade for each component by matching the patient’s responses with the specific criteria of the index.

 

STUDIES 205.130 AND 205.137

In these six-month studies, which are summarized above and reviewed in-depth in the Appendix to this document, one of the co-primary variables was the Mahler Transitional Dyspnea Index (TDI).  The Applicant chose to specify as the primary analysis, a “responder” analysis based on a threshold of 1 unit in the focal TDI score.  During drug development, the Agency informed the Applicant that the clinical validity of both the TDI instrument and of this threshold must be established in order for this primary analysis to be meaningful.  Further, the Applicant was informed that whatever effect was demonstrated in regard to the percentage of “responders” must itself be clinically meaningful in order to merit an indication for dyspnea associated with COPD.  Finally, the Applicant was informed that any claims in regard to dyspnea must be supported by a substantial weight of evidence.

 

At the end of the six-month studies, the percentages of patients with a TDI ³1 unit was 42% and 45% in the tiotropium groups (Studies 205.130 and 205.137, respectively), compared with 26% and 33% in the placebo groups.  These differences were statistically significant in both studies.  The percentages of responders in the active-comparator group (salmeterol) was 35% and 48% in these two studies.

 

Percentage of Patients with TDI³1 After 6 Months of Treatment (Studies 205.130 and 205.137)

Study

Tiotropium

Placebo

Salmeterol

205.130

42%

26%

35%

205.137

45%

33%

48%

 

There are additional data from these two studies that may shed light on the effect of tiotropium in regard to the symptom of dyspnea.  Because this would be a unique indication for tiotropium, some data on the effect of salmeterol on these endpoints is provided for comparison. 

-        Responder analyses for the TDI focal score (based on a threshold of 1 unit) were performed after 8 and 16 weeks of treatment.  In both studies, tiotropium was statistically superior to placebo in these analyses.  The percentages of responders in the tiotropium and placebo groups were 40% vs. 24% and 44% vs. 31% at Week 8, and 43% vs. 27% and 42% vs. 30% at Week 12 in Studies 205.130 and 205.137, respectively.  Of note, in one study (205.130) tiotropium was numerically superior to salmeterol on these analyses, and in the other study (205.137), salmeterol was numerically superior to tiotropium.

-        Using analyses of mean TDI focal scores rather than “responder” analyses, tiotropium was statistically superior to placebo on each test day in both studies.  The effect size was >1 unit on each day except Week 16 in Study 205.130.  Of note, in one study (205.130) salmeterol was not statistically superior to placebo on these analyses, but in the other study (205.137) salmeterol was statistically superior to placebo, with effect sizes ranging from 1.26 to 1.66.

-        A “COPD Symptom Score,” based on the investigator’s assessment, was assigned at each treatment visit.  Tiotropium was statistically superior to placebo in regard to the “shortness of breath” component of this score at most of the treatment visits. The effect size ranged from 0.17 to 0.36 on this 0-3 scale.  Salmeterol was also statistically superior to placebo for “shortness of breath” at most treatment visits.

-        A “shuttle walk test” (SWT) was administered after the first dose and after 8, 16, and 24 weeks of treatment.  The “Modified Borg Dyspnea Scale” was administered before and after each SWT.  In both studies, there was no difference between groups in regard to the distance walked in the SWT.  Of note, in Study 205.137, on each test day the distance walked was numerically superior in the placebo group, as compared to the tiotropium group.  There was no significant difference between tiotropium and placebo in regard to the Modified Borg scores.  There was also no significant difference between salmeterol and placebo on the SWT distance or the Modified Borg score.

 

STUDIES 205.114/204.117 AND 205.115/205.128

In these studies, which are summarized above and reviewed in-depth in the Appendix to this document, the symptom of dyspnea was addressed in two secondary variables, the TDI and the component of the “COPD Symptoms Score” called “shortness of breath.”  The TDI was administered on five occasions during these 1-year studies.  On all occasions the mean TDI focal score was statistically superior in the tiotropium group.  However, the difference between the tiotropium and placebo group was <1 on all but three occasions.  Symptoms of COPD were assessed and recorded by the investigator using a 0-3 scale at each visit.  The tiotropium group was statistically superior to the placebo group in regard to the score on the “shortness of breath” component at most visits.

 

STUDIES 205.122A/205.126A AND 205.122B/205.126B

These active-controlled studies are summarized above and reviewed in-depth in the Appendix to this document.  They TDI assessments on six occasions during the one-year treatment period.  In Study 205.122A/205.126A, the tiotropium group was statistically superior to ipratropium in regard to the mean TDI focal score on four of the six occasions.  However, the difference between the two groups was <0.75 units on each of these occasions.  In Study 205.122B/205.126B, the tiotropium group was superior to ipratropium in regard to the mean TDI focal score on every occasion, with differences exceeding 1 unit on four of the six occasions.



SUMMARY OF THE DYSPNEA EFFICACY RESULTS

The Applicant has proposed a unique indication for tiotropium, the treatment of dyspnea associated with COPD.  The primary support for this proposal is derived from two, six-month, active and placebo-controlled studies in which the TDI, an index of subjective dyspnea, was pre-specified as one of two co-primary efficacy variables.  In those studies, tiotropium was demonstrated to be statistically significantly superior to placebo in the pre-specified primary analysis.  This analysis was a “responder” analysis using a threshold of 1 in the TDI  as the definition of a “responder.”  The utility of this analysis will be discussed in the section below entitled Efficacy Conclusions.  Secondary analyses including TDI responder analyses on other test days, and analyses of mean TDI focal score data also showed statistical superiority of tiotropium over placebo.  It should be noted that in many of these analyses, the effect of tiotropium was not markedly greater than that of the active control, salmeterol.  Finally, in these studies, no difference between groups was seen in regard to the distance walked during a shuttle walk test, or perceived dyspnea during the shuttle walk test, as assessed by the modified Borg scale.

 

In other long-term, placebo controlled studies, the TDI data was analyzed using mean values.  While tiotropium was often statistically superior to its comparator (placebo or ipratropium), the differences were commonly <1 unit. 

 

 

D.                Efficacy Conclusions

The clinical development program for this drug included a total of six large, controlled studies in patients with COPD.  Of these, two were 1-year, placebo-controlled studies, two were 1-year, active-controlled studies, and two were 6-month, active- and placebo-controlled studies.  The one-year studies primarily focused on establishing substantial evidence of efficacy to support the indication traditionally used for COPD drugs, the relief of bronchospasm associated with COPD.  Thus, in these studies the primary efficacy variable was a measure of bronchodilation, FEV1.  One unique aspect of these studies is that the primary endpoint was the pre-dose FEV1, rather than a post-dose assessment, such as peak FEV1, as is more commonly used.  The benefit of using the pre-dose (or “trough”) value is that by showing statistical superiority to the comparator, the proposed dosing interval is supported.  However, there is less of a consensus regarding the minimum magnitude of effect that should be regarded as being clinically meaningful at this timepoint. 

 

The 6-month studies were submitted in order to support a proposal for a unique indication for a COPD drug, the treatment of dyspnea associated with COPD.  Prior to unblinding the data for these studies, the primary efficacy variable for these studies was altered, to include FEV1 and TDI focal score as co-primary variables.  The following discussion will address the proposed indications, treatment of bronchospasm associated with COPD and treatment of dyspnea associated with COPD, separately.

 

The six clinical studies appear to establish the bronchodilator efficacy of tiotropium.  Primary analyses of the six studies all demonstrate that treatment with tiotropium at the proposed dose results in statistically significant improvements in FEV1 at the end of the dosing interval (“trough” FEV1).  In four of the studies the comparator was placebo, and in the remaining two studies the comparator was a short-acting agent whose effect is expected to be negligible at the time the variable was assessed (morning, pre-dose).  The magnitude of effect demonstrated at this timepoint is small, but may be clinically meaningful.  In secondary analyses of the FEV1 data, tiotropium was statistically significantly superior to placebo in regard to standard post-dose variables such as average FEV1 and peak FEV1.  It is noted that the time to reach peak FEV1 is quite variable among individual patients.  Other secondary efficacy variables, such as morning and evening home peak flow measurements and supplemental “as-needed” albuterol use, appear to support the bronchodilator activity of tiotropium in COPD patients.  No consistent, clinically meaningful effect was seen on other indicators of COPD disease activity, such as COPD exacerbations, COPD hospitalizations, and health-related quality of life assessments.

 

The support of a proposed “dyspnea” indication appears to be less convincing.  It is true that the in the two six-month studies tiotropium was statistically superior to placebo on the co-primary analysis of the TDI focal score.  However, several points regarding the TDI  and the analysis of the TDI should be noted. 

·        The package of materials submitted by the Applicant in order to provide details regarding the development of the TDI was very limited.  The instrument was first described in 1984, and it is not clear from the submission that the methodology used to develop the instrument would be considered appropriate using modern day standards.  Currently, appropriate development of a patient reported outcome instrument typically involves: 1) convening of “focus groups” of the specific patient population in order to identify items of importance, 2) reducing the number of these items in order to eliminate highly correlated items, 3) determining the most appropriate response choices, and 4) assigning the most appropriate weight to each item. 

·        Responses to the TDI involve recollection of the baseline status, which may be difficult after many months.  For instance, the baseline assessment of  “Magnitude of Task” is determined in the Baseline Dyspnea Index (BDI) using four Grades of severity.  A score +1 on the TDI for this category, represents an improvement of less than one grade.  The ability of patients to make a determination of a change within one grade after 6 or 12 months is not clear.  It should be noted however, that in Studies 205.130 and 205.137, the TDI was administered at Weeks 8 and 16, in addition to Week 24.

·        There is little consensus in the medical literature regarding the minimal TDI focal score that is considered to be clinically meaningful.  Therefore the selection of the most appropriate “responder” threshold is somewhat uncertain.  There is no evidence that patients were consulted to determine what they believe is clinically meaningful.  The Applicant has proposed that 1 unit is clinically meaningful.  It should be noted that this also represents the smallest improvement that a patient could possibly report.  This means that there is no degree of improvement that could be reported that would not be considered to be clinically meaningful.  Of note, however, according to analyses performed by the Division’s Biometrics Reviewer, tiotropium would have been statistically superior to placebo in the two 6-month studies even if the “responder” threshold were set at 2, rather than 1. Using a response threshold of 2, the percentage of responders in the tiotropium and placebo groups was 33.7% and 23% in Study 205.130, and 40.8% and 29.8% in Study 205.137.  For comparison, the using this threshold, the percentage of responders in the salmeterol groups was 30.7% and 46.6% in Study 205.130 and 205.137, respectively. 

·        The six-month studies were multinational.  The issue of cross-culture interpretation and translation is not addressed.  The Applicant has not provided data to establish the validity of the TDI when translated into languages other than English and when used in other cultures.  

·        The effect size demonstrated is questionable to merit a specific indication for dyspnea.  In one study 42% of tiotropium patients were classified as “responders,” while 26% of placebo patients were “responders.”  In the second study, the difference between the groups was even smaller (45% vs. 33%).  According to “number needed to treat” (NNT) analyses performed by the Division’s Biometrics Reviewer, approximately 7.5 patients would need to be treated with tiotropium in order that one patient would note a dyspnea benefit above that expected with the use of placebo.  (This figure is derived from the pooled data from the two studies.  The NNT was 6.45 in Study 205.130, and 8.6 from Study 205.137).  

·        The robustness of the dyspnea effect is called into question by the fact that in analyses of mean TDI focal scores in the six pivotal studies the difference between tiotropium and placebo was often less than 1 unit. 

·        The comparator drug used in the two six-month studies (salmeterol) does not have an indication for dyspnea, yet its performance in the “responder” analyses was not different from that of tiotropium.

·        The studies were not designed with TDI as a primary efficacy endpoint.  The conduct of the studies reflect this the following ways: 1) there is no indication that the observers, who completed the TDI questionnaire, were blinded to other study data, either at the time of the visit, or over the duration of the study.  Knowledge of the patient’s clinical data and status as well as possible adverse events (e.g. dry mouth) could have introduced bias into the grading of the TDI.  2) The observer first reviewed the SGRQ results prior to interviewing the patient for the TDI.

 


 

VIII.     Integrated Review of Safety

 

A.        Brief Statement of Conclusions

The clinical development program included adequate numbers of subjects exposed.  The types of safety assessments used in these studies was adequate, and was generally consistent with development programs for other inhalation drug products for a COPD indication.  The adverse event data indicated that anticholinergic effects were more frequent in the treated group.  Dry mouth was quite common, and was more frequent in women and in older patients.  Other anticholinergic effects included constipation and urinary effects.  Upper respiratory tract infections were also more common in the tiotropium-treated patients. 

 

The safety database contains subtle suggestions that tiotropium may be associated with increased adverse cardiac effects, particularly in the category of “heart rate and rhythm disorders.”  The cardiac safety database contains relatively few 24-hour Holter monitors.  Given the potential, based on mechanism of action, pharmacokinetics, and intended patient population, for adverse cardiac effects with this drug, this issue will be raised for discussion at the PADAC meeting.

 

B.        Description of Patient Exposure

 

1.                  Clinical Studies

The Phase 3 development program included six “pivotal” clinical studies.  Four of these were randomized, double-blind, active- or placebo-controlled studies with treatment durations of approximately one year.  These studies were conducted in the U.S., Netherlands, and Belgium.  The two remaining studies were randomized, double-blind, active and placebo-controlled studies with treatment durations of six months.  Three additional studies are described by the Applicant as Phase 3 “characterization” studies.  These were: 1) a six-week placebo-controlled study comparing morning to evening dosing (“AM/PM dosing study”; 205.123); 2) a mucociliary clearance study (205.116); and 3) a sleep study (205.124).

 

Additional clinical studies include eleven human pharmacology studies, three Phase 2 single-dose studies (205.119, 205.120, and 205.139), one Phase 2 multiple-dose, dose ranging study (205.108), one multiple-dose dose-ranging study using tiotropium inhalation solution and inhalation powder, and four studies in patients with asthma.

 

This Clinical Briefing Document will focus primarily on safety data derived from the six “pivotal” clinical studies.  Following the approach taken by the Applicant in the Applicant’s Integrated Summary of Safety, the pooled safety data from the two 1-year placebo controlled studies, the pooled safety data from the two 1-year ipratropium-controlled studies, and the pooled safety data from the two 6-month salmeterol and placebo-controlled studies will be discussed separately.  Additional relevant safety information from the remainder of the clinical studies will be discussed as well.

 

2.                  Exposure

A total of 4,124 subjects participated in the clinical program [iss.pdf/p102].  This included 224 healthy volunteers, 3,411 COPD patients, 471 asthma patients, and 18 patients with renal impairment.  A total of 2,117 subjects were exposed to tiotropium by inhalation of the powder capsule formulation.  This included 57 healthy volunteers, 1,723 COPD patients, and 337 asthma patients.  A total of 1,701 subjects were exposed to the 18mcg dose of tiotropium.  Of these 1,701 subjects, 48% were exposed to the drug for more than 200 days, and 34% were exposed to the drug for more than 330 days.

 

The table below summarizes the numbers of patients exposed to tiotropium in the six “pivotal” Phase 3 studies.

 

Patient Exposure to Tiotropium in the Six “Pivotal” Phase 3 Studies                            [iss.pdf/p113-4]

 

Total

³101 days

³200 days

³ 330 days

One-year, placebo-controlled studies

550

501

(91%)

482

(88%)

302

(55%)

One-year, ipratropium-controlled studies

356

325

(91%)

316

(89%)

260

(73%)

Six-month, salmeterol- and placebo-controlled studies

402

353

(88%)

354

(88%)

not applicable


 

 

The mean age for all patients was 65 years in the one-year placebo-controlled studies, and 64 years in the one-year ipratropium-controlled studies and the six-month salmeterol and placebo-controlled studies.  Nearly all patients were caucasian, and 65% to 85% were male [iss.pdf/p127, 133]. In the one-year, placebo-controlled studies, the mean FEV1 was approximately 1 liter, representing 38-39% of the predicted value.  In the one-year, ipratropium-controlled studies the mean baseline FEV1 was 1.18 to 1.25 liters, representing 41-44% of the predicted value [iss.pdf/p129]. In the six-month, salmeterol- and placebo-controlled studies the mean baseline FEV1 was 1.07 to 1.12 liters, representing 39-41% of the predicted value [iss.pdf/p134].


 

C.        Methods and Specific Findings of Safety Review

 

1.                  Safety Evaluations Performed

In the tiotropium clinical studies safety was monitored using the following assessments: clinical adverse events, vital signs, physical examinations, clinical laboratory results, and electrocardiograms (ECGs).  Adverse events were classified using the Boehringer Ingelheim – World Health Organization – Adverse Reaction Terminology List (BI-WHO-ART) [iss.pdf/p82].  The respiratory system events were further divided into “upper” and “lower” respiratory system disorders.  One of the (non-pivotal) clinical studies included 24-hour Holter monitoring (Study 205.123, one of the Phase 3 “characterization” studies).

           

2.                  Significant/Potentially Significant Events (Deaths, Serious Adverse Events, and Discontinuations Due to Adverse Events)

 

The table below summarizes the incidences of deaths, serious adverse events (SAEs), and adverse events leading to discontinuation in the three sets of “pivotal” Phase 3 studies.

 

Significant/ Potentially Significant Adverse Event Profile                                                                 [iss.pdf/p33, 44]

 

1-year, placebo-controlled studies

1-year, ipratropium-controlled studies

6-month, salmeterol and placebo-controlled studies

 

Tiotropium

Placebo

Tiotropium

Ipratropium

Tiotropium

Salmeterol

 Placebo

 

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

Total Treated

550

100

371

100

356

100

179

100

402

100

405

100

400

100

Deaths

7

1.3

7

1.9

9

2.5

3

1.7

1

0.2

6

1.5

5

1.3

SAEs

99

18

78

21

57

16

46

25.7

37

9.2

50

12.3

55

13.8

AEs leading to discontinuation

 

53

 

9.6

 

50

 

13.5

 

35

 

9.8

 

22

 

12.3

 

29

 

7.2

 

60

 

14.8

 

64

 

16.0

 

A total of 26 deaths occurred among the 1456 patients enrolled in the one-year studies [iss.pdf/p145].  None of the deaths were considered by the investigators to be related to study medication.  In general, the causes of death were consistent with what might be expected in this patient population.  Two causes of death were reported in the tiotropium groups but not in the comparator groups.  These were myocardial infarction (4 deaths) and arrhythmia (1 death).  The incidence of death was similar in all groups.  In the one-year, placebo-controlled studies, there were 7 (1.3%) deaths in the tiotropium group and 7 (1.9%) in the placebo group.  In the one-year, ipratropium-controlled studies there were 9 (2.5%) deaths in the tiotropium and 3 (1.7%) deaths in the ipratropium group.  Narrative summaries of all deaths were reviewed by the Medical Reviewer. 

 

The seven deaths among the tiotropium patients in the one-year, placebo-controlled studies were due to [iss.pdf/p149-54]:

·        acute myocardial infarction:

-        67 year-old man, after 227 days of treatment.

·        coronary artery disease

-        49 year-old man developed severe chest pain after 91 days of treatment.  Cardiac catheterization revealed single vessel disease (60% lesion).  Cardiac medications were begun during a seven day hospitalization, but twelve days later he developed recurrent chest pain and expired.

·        cardiac arrhythmia:

-        85 year-old man found dead after 33 days of treatment.  No autopsy.

·        sudden death:

-        59 year-old man found dead in bed after 45 days of treatment.

·        cardiac arrest:

-        61 year-old man with history of hypertension and coronary artery disease (status post coronary artery bypass grafting) experienced cardiac arrest after 15 days of treatment. He was initially resuscitated and placed on a ventilator, but died two days later.

·        congestive heart failure/ cardiomyopathy:

-        65 year-old woman with a baseline diagnosis of cardiomyopathy, who was hospitalized for congestive heart failure after 339 days of treatment.  She was hospitalized for 15 days for diagnostic testing and treatment.  She was readmitted 5 days later with congestive heart failure, and died.

·        suicide:

-        51 year-old man with history of post-traumatic stress disorder died of suicide (opiate, cocaine, and diphenhydramine intoxication) after 112 days of treatment.

 

Thus, five of these seven deaths among the tiotropium patients in the one-year, placebo-controlled trials were attributable to cardiac ischemia or arrhythmia.  For comparison, only one of the seven deaths in the placebo group was attributed to cardiac ischemia or arrhythmia.  (This was a 65 year-old man with a history of hypertension who died after 240 days of treatment.  Details of the circumstances of his death are not provided, but an autopsy revealed atherosclerotic coronary disease without signs of acute myocardial infarction.)  The remaining deaths in the placebo group were due to worsening COPD (1 patient), cor pulmonale (1 patient; recorded as “myocardial insufficiency”), and carcinoma (4 patients). 

 

In the one-year, ipratropium-controlled studies there were two deaths in the tiotropium group (out of a total of 7) due to myocardial infarction, and no deaths in the ipratropium group (out of a total of 3) due to myocardial infarction [iss.pdf/p154-9].  The remaining causes of death were carcinoma, pulmonary emboli, respiratory insufficiency, and meningitis in the tiotropium group, and pneumonia, aortic aneurysm rupture, and carcinoma (with treatment-related leukopenia and sepsis) in the ipratropium group. 

 

In the six-month studies there was only one death in the tiotropium group.  This was due to ruptured abdominal aortic aneurysm [iss.pdf/p183].  For comparison, there were five deaths in the placebo groups of the six-month studies. These deaths were due to cardiac arrest (two events, one of which occurred in association with COPD exacerbation), respiratory insufficiency, bronchial carcinoma, and “death” (patient was found dead, cause not specified) [iss.pdf/p184-8].

 

Fewer patients in the tiotropium groups reported serious adverse events, as compared with both the placebo and the active comparator groups.  As indicated in the table above, the percent of patients reporting SAEs in the tiotropium group was 18% in the one-year, placebo-controlled studies and 16% in the one-year, ipratropium-controlled studies, compared with 21% of placebo patients and 26% of ipratropium patients [iss.pdf/p159].  In the six-month studies, 9.2% of tiotropium patients, 14% of placebo patients, and 12% of salmeterol patients reported SAEs.

 

The most common SAEs were COPD exacerbation and pneumonia.  None of the SAEs were considered to be related to tiotropium. COPD exacerbation SAEs were less common in the tiotropium groups (5.8% vs. 8.1% in the placebo-controlled one-year studies, 6.5% vs. 12% in the ipratropium controlled studies, and 3.5% vs. 5.8% in the placebo group and 5.9% in the salmeterol group in the 6-month studies) [iss.pdf/p39, 47].  The table below indicates the SAEs that occurred in >1 patient in the tiotropium group and occurred more frequently in the tiotropium group in the one-year, placebo-controlled studies.

 

Serious Adverse Events Occurring More Frequently in the Tiotropium Group And Occurring in >1 Patient in the Tiotropium Group (One-year, Placebo-Controlled Studies)  (number [%] of patients)                                                                                                                [iss.pdf/p161-6]

Event

Tiotropium Group

Placebo Group

Chest Pain

8 (1.5)

4 (1.1)

Dehydration

5 (0.9)

0 (0)

Neoplasm Malignant

4 (0.7)

0 (0)

Syncope

3 (0.5)

0 (0)

Myocardial Infarction

3 (0.5)

1 (0.3)

Angina Pectoris

2 (0.4)

1 (0.3)

Fibrillation, Atrial

2 (0.4)

1 (0.3)

Prostatic Disorder

2 (0.4)

0 (0)

Diabetes Mellitus, Aggravated

2 (0.4)

0 (0)

Hyperglycemia

2 (0.4)

0 (0)

Accident, Vehicular

2 (0.4)

0 (0)

Manic Reaction

2 (0.4)

0 (0)

Infection

2 (0.4)

0 (0)

 

As shown in the table above the SAEs that occurred in the tiotropium group but did not occur in the placebo group in the one-year, placebo-controlled studies were: dehydration (5 events), syncope (3 events), prostatic disorder (2 events), vehicular accident (2 events), diabetes mellitus aggravated (2 events), hyperglycemia (2 events), manic reaction (2 events), and infection (2 events).  In addition, one event of each of the following occurred in the tiotropium group, but did not occur in the placebo group: allergic reaction, arrhythmia, cardiac arrest, angina pectoris aggravated, coronary thrombus, sick sinus syndrome, tachycardia, tachycardia supraventricular, aneurysm, aortic stenosis, cardiomyopathy, hemoptysis, hypoxia, sinusitis, constipation, ileus, colitis, dysphagia, gastrointestinal disorder NOS, gastroesophageal reflux, esophagitis, goiter, hyperkalemia, colon carcinoma, larynx neoplasm malignant, malignant melanoma, neoplasm malignant, uterine carcinoma, neuritis, anxiety, delirium, depression, suicide attempt, cerebellar infarction, thrombus arterial leg, lymphadenopathy, herpes zoster, hydronephrosis, and renal calculus [iss.pdf/p161-6].

 

In the six-month studies, SAEs occurring in the tiotropium group but not in the placebo group were: upper respiratory tract infection (2 events), gastroenteritis (2 events), and one episode each of the following: tachycardia supraventricular, skeletal pain, aneurysm, breast neoplasm malignant (female), epididymitis, prostatic disorder, testis disorder, abdomen enlarged, accident household, cor pulmonale, arthritis rheumatoid aggravated, duodenal ulcer, skin ulceration, urticaria, epistaxis, and cataract [iss.pdf/p191-3].

 

The incidence of discontinuation due to adverse events was lower in the tiotropium groups as compared with both the placebo and the active comparator groups.  In the one-year, placebo-controlled studies, 53 (9.6%) of tiotropium patients and 50 (14%) of placebo patients discontinued due to an adverse event [iss.pdf/p167].  In those studies, events leading to discontinuation that were seen in more than two patients in a treatment group are listed in the table below.  Dry mouth, the only event that occurred more frequently in the tiotropium group, is shaded.

 

Adverse Events Leading to Discontinuation, occurring in more than 2 patients in the one-year, placebo-controlled studies  (number [%] of patients)                                                            [iss.pdf/p167]

Event

Tiotropium Group

Placebo Group

COPD Exacerbation

20 (3.6%)

19 (5.1%)

Dyspnea

0 (0%)

5 (1.3%)

Pneumonia

4 (0.7%)

5 (1.3%)

Cardiac Failure

2 (0.4%)

3 (0.8%)

Dry Mouth

3 (0.5%)

1 (0.3%)

 

In the 6-month studies, 7.2% of tiotropium patients and 16% of placebo patients discontinued due to an adverse event [iss.pdf/p194].  In these studies, COPD exacerbation and dyspnea were the only AEs that led to discontinuation of more than 2 patients in a treatment group.  COPD exacerbation was the cause of discontinuation in 3.5% of tiotropium patients and 7.5% of placebo patients, and dyspnea was the cause of discontinuation in 1.2% of tiotropium patients and 3.3% of placebo patients.  Dry mouth led to discontinuation in 1 tiotropium patient (0.2%) and in 0 patients in the placebo and salmeterol groups.

 

3.                  Other Safety Findings: Adverse Events, Lab Findings, Vital Signs, and ECGs

 

The table below summarizes the overall incidence of adverse events and the incidence of those adverse events that were considered by the investigator to be possibly drug-related.  The overall incidence of adverse events was similar among the groups.  Of note, the incidence of drug-related adverse events was greater in the tiotropium group, as compared to placebo and as compared to each of the active comparators examined (ipratropium and salmeterol).  This is due to the increased incidence of drug-related dry mouth with tiotropium.

 

Adverse Event Profile                                                                                                                            [iss.pdf/p33, 44]

 

1-year, placebo-controlled studies

1-year, ipratropium-controlled studies

6-month, salmeterol and placebo-controlled studies

 

Tiotropium

Placebo

Tiotropium

Ipratropium

Tiotropium

Salmeterol

 Placebo

 

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

N

(%)

Total Treated

550

100

371

100

356

100

179

100

402

100

405

100

400

100

All Adverse Events

 

495

 

90

 

338

 

91.1

 

318

 

89.3

 

162

 

90.5

 

298

 

74.1

 

305

 

75.3

 

307

 

76.8

Drug-related Adverse Events

 

 

104

 

 

18.9

 

 

34

 

 

9.2

 

 

73

 

 

20.5

 

 

22

 

 

12.3

 

 

44

 

 

10.9

 

 

33

 

 

8.1

 

 

31

 

 

7.8

 

The tables below summarizes adverse events that were reported by ³3% of patients in the tiotropium group and occurred more frequently in the tiotropium group as compared to the placebo group.  For purposes of reference, the tables contain data on the incidence of these AEs in the active comparator groups.  The active comparator data is drawn from separate studies in the case of the one-year studies, and from the same studies in the case of the 6-month studies.  The first table contains the one-year studies, and the second table contains the 6-month studies. 

 

Adverse Events Reported by ³3% of Patients in the Tiotropium Group and Occurring More Frequently In the Tiotropium Group as Compared with the Placebo Group  [One-Year Studies]                                  [iss.pdf/p143-4]

 

1-year, placebo-controlled studies

1-year, ipratropium-controlled studies

 

Tiotropium

Placebo

Tiotropium

Ipratropium

 

N

(%)

N

(%)

N

(%)

N

(%)

Total Treated

550

100

371

100

356

100

179

100

Body as a Whole

Accidents

Chest Pain

Edema, dependent

Influenza-Like Symptoms

 

73

38

25

45

 

13.3

6.9

4.5

8.2

 

42

17

13

30

 

11.3

4.6

3.5

8.1

 

16

19

10

39

 

4.5

5.3

2.8

11.0

 

14

4

9

25

 

7.8

2.2

5.0

14.0

Gastrointestinal System

Abdominal Pain

Constipation

Dyspepsia

Mouth Dry

Vomiting

 

26

19

32

88

19

 

4.7

3.5

5.8

16.0

3.5

 

11

6

17

10

9

 

3.0

1.6

4.6

2.7

2.4

 

20

2

5

43

3

 

5.6

0.6

1.4

12.1

0.8

 

11

2

1

11

3

 

6.1

1.1

0.6

6.1

1.7

Musculoskeletal System

Arthritis

Myalgia

 

26

21

 

4.7

3.8

 

17

11

 

4.6

3.0

 

15

13

 

4.2

3.7

 

7

6

 

3.9

3.4

Resistance Mechanism Disorders

Infection

Moniliasis

 

23

20

 

4.2

3.6

 

12

9

 

3.2

2.4

 

5

10

 

1.4

2.8

 

5

3

 

2.8

1.7

Respiratory System

Coughing

Epistaxis

Pharyngitis

Rhinitis

Sinusitis

Upper Respiratory Tract Infection

 

26

20

49

30

62

226

 

4.7

3.6

8.9

5.5

11.3

41.1

 

17

7

27

20

35

138

 

4.6

1.9

7.3

5.4

9.4

37.2

 

30

4

23

9

12

153

 

8.4

1.1

6.5

2.5

3.4

43.0

 

17

2

5

4

4

62

 

9.5

1.1

2.8

2.2

2.2

34.6

Skin and Appendages

Rash

 

23

 

4.2

 

8

 

2.2

 

7

 

2.0

 

4

 

2.2

Urinary System

Urinary Tract Infection

 

40

 

7.3

 

19

 

5.1

 

14

 

3.9

 

4

 

2.2

 

In the one-year placebo-controlled studies, the most notable adverse events were related to the gastrointestinal system (abdominal pain, constipation, dyspepsia, dry mouth, and vomiting).  The occurrence of AEs in the category of “Gastrointestinal System Disorders” was 38.5% in the tiotropium group and 29.1% in the placebo group [issa.pdf/13].  Of these, by far the most common was dry mouth, with an incidence of 16% in the tiotropium group.  Of note, the one-year ipratropium-controlled studies demonstrated that the frequency of dry mouth is greater with tiotropium than with the related drug, ipratropium.  Upper respiratory tract infections were also remarkably more common in the tiotropium groups as compared to both the placebo group and the ipratropium group.  Other upper respiratory tract AEs, such as epistaxis, pharyngitis, and sinusitis may reflect drying effects of this anticholinergic compound on the airway mucosa.  The mechanism that might be responsible for the observed increased incidence of urinary tract infections in the tiotropium group is not known, but may relate to urinary stasis due to anticholinergic effects on the genitourinary system.

 

The table above includes adverse events reported by ³3% of subjects in a treatment group.  The listings of all AEs reported in the one-year, placebo-controlled studies, by treatment group, were reviewed [issa.pdf/p7-34].  The following observations are derived from these listings: 

·        The AE “allergic reaction” occurred in 14 (2.5%) patients in the tiotropium group and 3 (0.8%) patients in the placebo group.

·        The AE “tooth caries” occurred in 4 (0.7%) patients in the tiotropium group and 0 patients in the placebo group. 

·        The occurrence of AEs in the category “Metabolic and Nutritional Disorders” was 6.4% in the tiotropium group and 2.7% in the placebo group. This difference is primarily the result of the following disparities in the occurrence of AEs in this category: 

-        1) “diabetes mellitus”, “diabetes mellitus aggravated,” or “hyperglycemia” was reported in 14 (2.5%) tiotropium patients and was reported in only 1 (0.3%)  placebo patients;

-        2) “dehydration” was reported in five tiotropium patients (0.9%) and was not reported in any placebo patients; Reviewer’s Comment: It is not clear if the occurrence of dehydration was related to the reported hyperglycemia/ diabetes.

-        3) “hypercholesterolemia” was reported in 6 (1.1%) tiotropium patients and 1 (0.3%) placebo patients.

·        “Urinary retention” occurred in 4 (0.7%) tiotropium patients and 0 placebo patients.

·        “Micturation disorder” or “micturation frequency” occurred in 6 (1.1%) tiotropium patients and 0 placebo patients.

Reviewer’s Note: With the exception of hypercholesterolemia, which was slightly more common in the tiotropium group than the placebo group (1.0% vs. 0.3%), the adverse event data from the six-month studies did not confirm these observations [issa.pdf/p201-22].

 

 

Adverse Events Reported by ³3% of Patients in the Tiotropium Group and Occurring More Frequently In the Tiotropium Group as Compared with the Placebo Group  [6-month Studies]                                    [iss.pdf/p182]

 

Tiotropium

Placebo

Salmeterol

 

N

(%)

N

(%)

N

(%)

Total Treated

402

100

400

100

405

100

Body as a Whole

Accidents

Back Pain

Chest Pain

Influenza-Like Symptoms

 

17

16

16

27

 

4.2

4.0

4.0

6.7

 

10

12

15

16

 

2.5

3.0

3.8

4.0

 

21

16

14

21

 

5.2

4.0

3.5

5.2

Gastrointestinal System

Mouth Dry

 

33

 

8.2

 

9

 

2.3

 

7

 

1.7

Respiratory System

Pharyngitis

Sinusitis

Upper Respiratory Tract Infection

 

18

13

78

 

4.5

3.2

19.4

 

12

10

64

 

3.0

2.5

16.0

 

14

1

69

 

3.5

0.2

17.0

 

The AE data from the 6-month studies indicate fewer differences between tiotropium and placebo.  Several of the AEs that were more common in the tiotropium group in the 1-year studies were also noted to be more common in the tiotropium group in the 6-month studies.  Most notable among these were dry mouth, upper respiratory tract infection, influenza-like symptoms, and pharyngitis.

 

In these studies, the investigators were asked to indicate which adverse events were considered to be possibly related to study drug.  The most common adverse event that was considered to be possibly drug related was dry mouth [iss.pdf/p37-8, 46-7].  Drug-related dry mouth was reported in 14% of the tiotropium patients in the placebo-controlled studies (compared with 2.2% of placebo patients), in 11% of tiotropium patients in the ipratropium-controlled studies (compared with 5.6% of ipratropium patients), and in 6.5% of tiotropium patients in the 6-month studies (compared with 1% in the placebo and the salmeterol patients).  Drug-related dysphonia was also more common in the tiotropium group as compared with placebo in the one-year studies (1.5% vs. 0.3%), but not in the 6-month studies.  Finally, drug-related pharyngitis was slightly more common in the tiotropium group (1.1% vs. 0.8% in the placebo-controlled studies, and 1.1% vs. 0% in the ipratropium-controlled studies) in the one-year studies.

 

Vital signs were measured at the screening visit and on test days at the same intervals as the pulmonary function testing for the first three hours post dose (vital signs were measured just prior to pulmonary function measurements) [iss.pdf/p82].  The mean values for heart rate and blood pressure were similar in the tiotropium and the placebo groups [issa.pdf/p263-5].  The Applicant defined “marked changes” in vital signs as follows [iss.pdf/p229]:

-        Systolic BP increase: an increase of ³25mmHg above baseline

-        Systolic BP decrease: below 100mmHg if not at that level at baseline, and a decrease of greater than 10mmHg below baseline

-        Diastolic BP increase: above 90mmHg and an increase of greater than 10mmHg from baseline

-        Diastolic BP decrease: below 60mmHg if not at that level at baseline and a decrease of >10mmHg below baseline

-        Pulse increase: greater than 100bpm if not at that level at baseline and an increase of >10% above baseline.

-        Pulse decrease: below 60bpm if not at that level at baseline and a decrease of >10bpm below baseline

 

The incidence of “marked changes” from baseline (as defined by the Applicant) were generally similar in the tiotropium and placebo groups [iss.pdf/p236-7].  On Test Day 1 in the one-year, placebo-controlled studies, more patients in the tiotropium group developed a marked decrease in systolic blood pressure (defined as: below 100mm Hg if not at that level at baseline, and a decrease of greater than 10mm Hg below baseline [iss.pdf/p229]), as compared with placebo (3.1% vs. 0.5%).  Because patients with potentially significant changes in pulse rate due to the anticholinergic effects of the drug might not be captured by the definition of a “marked change” for increased pulse, the Applicant was asked to submit shift tables for pulse rate increases of various magnitudes.  This data was submitted on July 31, 2002.  In all of the placebo-controlled studies, no remarkable difference was seen between tiotropium and placebo in regard to the percentages of patients who exhibited increases in heart rates of >5, >10, >15, or >20 beats per minute at any test day [Submission dated 7/31/02, pages 4-9].

 

In the four 1-year Phase 3 studies, laboratory testing was performed at baseline and at three-month intervals throughout the treatment period [iss.pdf/p83].  In the two 6-month Phase 3 studies, laboratory testing was performed at baseline and at the end of the study.  Laboratory tests included hematology, clinical chemistry, and urinalysis.  The mean values for all parameters both at baseline and conclusion of patient participation were similar between treatment groups [iss.pdf/p244-50].  The incidence of “marked” changes in laboratory values (as defined by the Applicant) from baseline to final evaluation was similar among groups in the one-year and the six-month studies [iss.pdf/p251-6].  In the one-year, ipratropium-controlled studies there was a relatively high percentage of subjects in each group who demonstrated “marked” increase in LDH (12.7%  in the tiotropium group and 9.9% in the ipratropium group).  Marked elevations in other liver enzymes were not seen, nor were marked changes in hemoglobin or hematocrit to suggest hemolysis as a source of the LDH.

 

he THe

 

 

In the one-year, placebo-controlled studies there was no difference between groups in the percentage of patients with clinically significant changes in physical examination (defined by the Applicant) from baseline to final examination.  In the one-year, placebo-controlled studies there were 46 (8.4%) such patients in the tiotropium group and 36 (9.7%) such patients in the placebo group [iss.pdf/p257].

 

Electrocardiogram data are discussed in the section below, entitled “Adverse Events Related to the Pharmacologic Actions of the Drug.”

 

Paradoxical bronchospasm, defined as a decline in FEV1 by at least 15% from baseline within 30 minutes of administration of study drug, was less frequent in the tiotropium group than in the placebo group in the one-year, placebo-controlled studies (4.5% vs. 12%) [iss.pdf/p240].  In five of the 25 tiotropium patients who exhibited paradoxical bronchospasm, the event occurred on two test days.   In the one-year, ipratropium-controlled studies 15 (4.2%) patients in the tiotropium group and 1 (0.6%) patient in the ipratropium group experienced paradoxical bronchospasm.  In the six-month studies 10 (1.0%) patients in the tiotropium group, 22 (2.1%) patients in the salmeterol group, and 33 (3.2%) patients in the placebo group experienced paradoxical bronchospasm.  There were no discontinuations of tiotropium due to paradoxical bronchospasm.

 

4.                  Pregnancy

No pregnancies were reported during the conduct of any of the clinical studies for tiotropium [Submission date 7/24/02, page 4].

 

5.                  Interactions

 

Drug-Demographic Interactions

In order to asses the effect of age on the safety of tiotropium, adverse events were analyzed according to age groups (£60 years, 61-70 years, and ³71 years) [iss.pdf/p41].  In the one-year placebo-controlled studies, two specific adverse events were noted to occur with increasing frequency in the older age groups in the tiotropium group only, suggesting a drug-age interaction.  These were dry mouth, and constipation. A third AE, urinary tract infection, occurred with greater frequency in older patients in both treatment arms, although the effect was more marked in the tiotropium group. 

 

The adverse event “dry mouth” increased in frequency with age in the tiotropium group.  In the one-year, placebo-controlled studies the percent of patients with dry mouth was 11% in the younger age group, 16% in the middle age group, and 21% in the older age group [iss.pdf/p176].  In contrast, the incidence of this adverse event in the placebo group was 3.0%, 1.9%, and 3.5% in the three age groups.  This observation was also made in the one-year, ipratropium-controlled studies, in which the percentages of patients with dry mouth also increased with age (7.7%, 15%, and 14%).  In contrast, the percentages declined with age in the ipratropium group (8.2%, 6.1%, and 4.2%) in these studies [iss.pdf/p41].  Drug-age interaction was not suggested in the 6-month studies [iss.pdf/p49].

 

In the one-year, placebo-controlled studies constipation was also more frequent with increasing age in the tiotropium group (2%, 2.8%, and 6%), but not in the placebo group (3.0%, 0.6%, and 1.7%).  In these studies, urinary tract infection occurred with increased frequency in the older age groups in the both the tiotropium group (3.3%, 5.2%, and 12%), and the placebo group (2.0%, 3.9%, and 6.1%), although the frequency was greater in the tiotropium group.  These observations were not made in the one-year, ipratropium-controlled studies or the six-month studies [iss.pdf/p49].

 

Reviewer’s Comment: The observation that dry mouth, constipation, and urinary tract infection occur more frequently with increasing age in the tiotropium group, along with the observation of increased systemic drug exposure with increasing age (see discussion of pharmacokinetics in Section IV of this Clinical Briefing Document) suggest that these adverse events represent systemic effects of the drug.  

 

The majority of patients in the pivotal clinical studies were men.  The proportions of patients with adverse events was generally similar between genders within each treatment group, with the exception of dry mouth.  In the one-year, placebo-controlled studies the frequency of dry mouth in the tiotropium group was 23% among women, and 13% among men.  For comparison, the frequencies in the placebo group were 2.9% in women and 2.6% in men [iss.pdf/p177].  This pattern was also seen in the six-month studies, with dry mouth being reported by 14.3% of women and 6.4% of men [iss.pdf/p49].  In the one-year, ipratropium-controlled studies women also reported more dry mouth than men [iss.pdf/p42].

 

Genitourinary effects also showed evidence of a gender effect in the one-year studies.  The adverse events “urinary retention” and “micturation disorder” were reported solely in men, and there was an increase in the frequency of urinary tract infection among men.  Urinary retention occurred only in men in the tiotropium group (1.1%) in the one-year, placebo-controlled studies.  Micturation disorders occurred only in men receiving either tiotropium (1.1% in the placebo-controlled studies and 0.3% in the ipratropium-controlled studies), or ipratropium (0.6%).

 

In the six-month studies, pharyngitis and sinusitis were more common in women (7.7% and 7.7%) than in men (3.5% and 1.9%) [iss.pdf/p49].

 

Because very few patients in these studies were non-white, analyses for drug-race safety interactions were not informative.

 

Drug-disease interactions

There was no evidence of a drug-disease severity interaction, based on categories of diseases severity (FEV1 <35%, FEV1 35-49%, and FEV1 ³50% predicted) [iss.pdf/p43, 50].  There was no evidence of a drug-smoking status interaction, based on smoking status at entry into the trial [iss.pdf/p43, 50]. 

 

Drug-drug interactions

The clinical development program did not include specific drug-drug interaction studies.  Subgroup analyses of adverse event data from the “pivotal” Phase 3 studies were performed for baseline users vs. non-users of theophylline, oral corticosteroids, and inhaled corticosteroids.  While the incidence of COPD exacerbations was greater in steroid users compared with steroid non-users, the Applicant states that there was no evidence of interaction of tiotropium with oral steroids or inhaled steroids on reported adverse events.  In the one-year placebo controlled studies, the incidence of dry mouth in the tiotropium group was greater in theophylline users than in non-users (20% vs. 15%) [iss.pdf/p175].  No such difference was seen in the placebo group.  In the 1-year ipratropium-controlled studies, reports of dry mouth were equally distributed in those receiving tiotropium who were theophylline users and non-users.  Finally, in the six-month studies, the pattern was reversed, with a lower incidence of dry mouth among tiotropium patients who were theophylline users vs. non-users (2.7% vs. 10%) [iss.pdf/p48].

 

6.                  Safety Findings from Other Clinical Studies

 


COPD Studies

COPD studies discussed in this section include the AM/PM dosing trial (205.123), the mucociliary clearance trial (205.116), the sleep trial (205.124), the dose-ranging trials (205.119, 205.120, and 205.108), a pharmacokinetic trial in the elderly (205.133), and  a trial conducted with the Respimat device (205.127) [iss.pdf/p50].

 

There were three deaths in these studies.  The causes of death were myocardial infarction (11 weeks after the last dose of tiotropium in the pharmacokinetic trial in the elderly), respiratory failure (in a placebo patient in the sleep study), and non-Hodgkin’s lymphoma (108 days following the two-week study period of a dose-ranging trial [205.120]).  Few SAEs were reported in these relatively short studies.  Few adverse events led to discontinuation, and such events were generally less common in the tiotropium groups.  In the six-week AM/PM dosing study, one patient receiving PM tiotropium developed cystitis, hematuria, and orchitis requiring hospitalization.  Study drug was discontinued [iss.pdf/p206].  One patient in the Respimat study (205.127) who was receiving tiotropium 2.5mcg developed worsening of hematuria that was considered unexpected and related to the study drug [iss.pdf/p206-7].  Of note, male rats developed proteinaceous material in the urinary bladder in the majority of preclinical studies [4/18/02 submission, iss.pdf/p272].  This was associated with a mild inflammatory response and diffuse hyperplasia of the bladder transitional epithelium, and prostatitis.

 

Asthma Studies

Among the four asthma studies (205.121, 205.201, 205.202, and 205.203), no deaths were reported and SAEs and AEs leading to discontinuation were few [iss.pdf/p208-9].

 

Healthy Volunteer Studies

The most common AEs in the single-dose studies were headache and taste perversion [iss.pdf/p209].  In the multiple-dose studies the most common AEs were dry mouth and taste perversion. 

 

7.                  Adverse Events Related to the Pharmacologic Actions of the Drug

The Application included specific attention to adverse effects that might result from the anticholinergic effect of tiotropium.  These include gastrointestinal effects (dry mouth, constipation, and dysphagia), cardiovascular effects (tachycardia), genitourinary effects (urinary retention, urinary tract infection), and ophthamologic effects (glaucoma).

 

Dry Mouth

Dry mouth was consistently more common in tiotropium groups as compared with placebo and as compared with the active comparators, ipratropium and salmeterol.  Dry mouth was more common in older patients and in women.  The median onset of dry mouth, which was generally of mild or moderate intensity, was 15 to 35 days [iss.pdf/p211-3].  Severe dry mouth and discontinuation due to dry mouth were uncommon (three patients in each category in the one-year studies).   In the one-year, placebo-controlled studies, tiotropium was also associated with increased frequency of certain adverse events that may be related to the drying effects of the drug.  These include epistaxis (3.6% vs. 1.9%), pharyngitis (8.9% vs. 7.3%), sinusitis (11.3% vs. 9.4%), and moniliasis (3.6% vs. 2.4%) [iss.pdf/p212].  Among these, the frequency of pharyngitis, sinusitis, and moniliasis were greater in the tiotropium group (6.5%, 3.4%, and 2.8%, respectively) compared with the ipratropium group (2.8%, 2.2%, and 1.7%, respectively) in the one-year, ipratropium-controlled studies.

 

Constipation

In the one-year, placebo-controlled studies constipation was reported more frequently in the tiotropium group (3.5%) than in the placebo group (1.6%) [iss.pdf/p214].  One patient in the tiotropium group required hospitalization due to fecal impaction.

 

Dysphagia

Dysphagia was reported by three patients in the one-year studies.  All three were in the tiotropium group [iss.pdf/p215].  Two of the patients underwent endoscopy as a result of the symptom.

 

Urinary Retention and Micturation Disorders

Urinary retention occurred in four patients (0.7%) receiving tiotropium in the one-year placebo-controlled studies [iss.pdf/216]. The four cases occurred between treatment days 18 and 174, in men between the ages of 69 and 77.  All four required the placement of a Foley catheter and three were started on medication for BPH.  Urinary retention also occurred in one patient receiving tiotropium in the six-month studies, but did not occur in any patients in the one-year, ipratropium-controlled studies [iss.pdf/p216].  In addition, there were four reports of micturation disorders in the tiotropium group (in men aged 64 to 81 years) and none in the placebo group of the one-year, placebo-controlled studies, and one case of micturation disorder in each of the two treatment arms of the one-year, ipratropium-controlled studies.

 

Urinary Tract Infection

In the one-year, placebo-controlled studies, the incidence of urinary tract infection was greater in the tiotropium group (6.5% vs. 4.0%) [iss.pdf/p217].  In the one-year, ipratropium-controlled studies the incidence of UTI was not different between the tiotropium and the ipratropium groups. However, the incidence of cystitis was greater in the tiotropium group in those studies (2.5% vs. 0.0%).  In the six-month studies the incidence of UTI was 1.2% in the tiotropium and 0.5% in the placebo group.

 

Cardiovascular Effects

The incidence of death due to cardiac events was not different in the tiotropium and placebo groups in the one-year studies (0.5% vs. 0.3%).  However, there were subtle indications that tiotropium may be associated with increased frequency of adverse cardiac effects, specifically in the category of “heart rate and rhythm disorders.”  (Note: Cardiac AEs are divided into three categories: “general,” “heart rate and rhythm disorders,” and “myo-, endo-, pericardial and valve disorders.”)  In the one-year, placebo-controlled studies the incidence of “heart rate and rhythm disorders” was greater in the tiotropium group (4.4%, 24 patients) than in the placebo group  (2.2%, 8 patients) (see table below) [iss.pdf/p231].  It should be noted that in the one-year ipratropium-controlled studies the incidence of “heart rate and rhythm disorders” was greater in the ipratropium group (5.0%) than in the tiotropium group (3.9%).  The incidence of serious “heart rate and rhythm disorders” in the one-year, placebo-controlled studies was 1.3% in the tiotropium group and 0.5% in the placebo group [iss.pdf/p232]. This included two SAEs of supraventricular tachycardia, both of which occurred in patients on tiotropium.  In the one-year studies, there were four discontinuations due to heart rate and rhythm disorders, all in the tiotropium group [iss.pdf/p234]. In the one-year, placebo-controlled studies, there were two deaths due to heart rate and rhythm disorders, both in the tiotropium group [iss.pdf/p233].  Although there was no difference between groups for “Myo-, Endo-, Pericardial and Valve Disorders” AEs in the one-year, placebo-controlled studies, there was a slightly greater incidence of SAEs in this category (2.0% vs. 1.3%) [iss.pdf/p232].

 

Cardiac Adverse Events, by WHO System Organ Class (1-year studies)                                              [iss.pdf/p231]

 

1-year, placebo-controlled studies

1-year, ipratropium-controlled studies

 

Tiotropium

Placebo

Tiotropium

Ipratropium

 

N

(%)

N

(%)

N

(%)

N

(%)

Total Treated

550

100

371

100

356

100

179

100

Cardiovascular Disorders, General

5

0.9

5

1.3

3

0.8

3

1.7

Cardiac Failure

Cardiac Failure, Right

Cardiomegaly

Cor Pulmonale

Heart Disorder

Heart Valve Disorder

5

0

0

0

0

0

0.9

0

0

0

0

0

4

0

1

1

0

0

1.1

0

0.3

0.3

0

0

1

2

0

0

0

1

0.3

0.6

0

0

0

0.3

1

0

1

0

1

0

0.6

0

0.6

0

0.6

0

Heart Rate and Rhythm Disorders

24

4.4

8

2.2

14

3.9

9

5.0

Arrhythmia

AV Block

Bradycardia

Bundle Branch Block

Cardiac Arrest

Extrasystoles

Fibrillation Atrial

Palpitation

Sick Sinus Syndrome

Tachycardia

Tachycardia supraventricular

4

0

1

0

1

2

5

4

1

4

2

0.7

0

0.2

0

0.2

0.4

0.9

0.7

0.2

0.7

0.4

1

0

1

0

0

0

3

2

0

1

1

0.3

0

0.3

0

0

0

0.8

0.5

0

0.3

0.3

0

1

0

1

0

0

5

3

0

4

0

0

0.3

0

0.3

0

0

1.4

0.8

0

1.1

0

1

1

0

0

0

0

4

3

0

0

0

0.6

0.6

0

0

0

0

2.2

1.7

0

0

0

Myo-, Endo-, and Pericardial and Valve Dis

15

2.7

10

2.7

10

2.8

6

3.4

Angina  Pectoris

Angina Pectoris Aggravated

Cardiomyopathy

Coronary Artery Disorder

Heart Murmur

Myocardial Infarction

Thrombosis Coronary

4

2

1

4

0

3

1

0.7

0.4

0.2

0.7

0

0.5

0.2

2

0

0

4

2

2

0

0.5

0

0

1.1

0.5

0.5

0

6

1

0

0

0

3

0

1.7

0.3

0

0

0

0.8

0

4

1

0

0

0

1

0

2.2

0.6

0

0

0

0.6

0

 

 

In the one-year, placebo-controlled studies ECGs were done at baseline and every 90 days for the duration of the study.  Unfortunately, the protocol did not specify the timing of the ECGs in relation to study drug and that information was not captured on the case report forms [Submission date 7/16/02, page 5].  Therefore, it cannot be assumed that the ECGs were obtained at or near the expected Cmax.  In the these studies, there was no difference between groups in the incidence of ECG changes (1% vs. 1.8%).  The Applicant states that there was no imbalance in regard to the type of ECG abnormalities noted.  One patient in the one-year, ipratropium-controlled studies developed tachycardia 30 minutes after the first dose of tiotropium and discontinued the study. 

 

In the six-month studies, ECGs were performed at baseline and at the completion of the study.  The incidence of ECG changes was 1.7% in the tiotropium group and 0.8% in the placebo group.  In the four-week, parallel-group, placebo-controlled, dose-ranging study (205.108) ECGs were performed at baseline and at one, three, and five hours after drug administration on Day 1 and Weeks 1, 2, and 4.  No differences in the occurrence of ECG changes was noted between active and placebo groups [iss.pdf/p225].  Tachyarrhythmias were seen in three tiotropium patients (ventricular tachycardia in a patient receiving 4.5mcg, atrial fibrillation in a patient receiving 9mg, and sinus tachycardia in a patient receiving 18mcg) and in one patient in the placebo group (sinus tachycardia). 

 

The ECG database is supported by timed ECGs (1, 3, and 5 hours post-dose) that were performed in the multiple-dose, dose-ranging study (205.108).  In that placebo-controlled study, doses of 5.5mcg, 11.0mcg, 22.0mcg, and 44.0mcg were studied (33-35 patients per treatment group).  In addition to a baseline pre-dose ECG, timed ECGs (and 2-minute rhythm strips) were obtained after 8, 15, and 29 days of treatment [U96-3068.pdf/p111-14].  The ECGs and rhythm strips were centrally read by a cardiologist.  There were no differences seen between placebo and active treatment in regard to ECG changes.  Borderline QT interval was reported as intermittent in one placebo patient and transient in on tiotropium patient (22.0mcg).  Specific QT or QTc interval data was not submitted.

 

Only one study included 24-hour Holter monitoring.  This was the six-week, AM/PM dosing trial (205.123), in which there were three, double-blind treatment groups (tiotropium 18mcg AM dosing, tiotropium 18mcg PM dosing, and placebo, using the Handihaler device) [U00-0121.pdf].  The study was performed in the UK and the Netherlands, during the period May, 1997 to July, 1998 [U00-0121.pdf/p20].  Exclusion criteria were similar to other clinical studies.  Patients with a history of significant disease other than COPD and patients with a recent history of heart failure or any cardiac arrhythmia requiring drug therapy were excluded.  A total of 121 patients entered the trial (43 in the tiotropium PM dosing group, 38 in the tiotropium AM dosing group, and 40 patients in the placebo group).  The mean age was 65.8 years, and 62% of the population was male.  In this study, Holter monitoring was performed on two occasions.  The baseline, 24-hour Holter monitor was placed on Day 0, and continued until Day 1, approximately 2 hours after the first dose of study medication [U00-0121.pdf/p44].  The second 24-Holter monitor was placed at Visit 4, at the end of the 6-week treatment period.  The protocol does not specify when the monitor was placed in relation to study drug administration [U00-0121.pdf/p45].  The analysis of the Holter tapes was performed by a central facility (Hertford Medical, The Netherlands) [U00-0121.pdf/p41]  Data on supraventricular and ventricular ectopy, heart rate, and heart rate variability were collected and assessed [U00-0121.pdf/p101].  Reviewer’s Comment: Interpretation of the comparisons of “baseline” and on-treatment Holter results is complicated by the fact that the first dose of study drug was given during the recording of the “baseline” Holter.  On-treatment Holter monitor results were available for 35 tiotropium PM patients, 37 tiotropium AM patients, and 31 placebo patients [U00-0121.pdf/p103].  No clear effect on the frequency of supraventricular or ventricular ectopy was observed.  No episodes of atrial fibrillation or atrial flutter were observed, either at baseline or on treatment [U00-0121.pdf/p104].  One subject in the tiotropium PM group developed a four-fold increase in ventricular ectopy after medication [U00-0121.pdf/p104].  None of the treatment groups had a remarkable change in mean heart rate, minimum heart rate, or maximum heart rate.  There were no episodes of AV block.  The Applicant states that assessment of heart rate variability is a sensitive marker of anticholinergic effects on the heart.  In general, an increase in variability is said to indicate an improvement in cardiac autonomic function.  The Applicant states that tiotropium was associated with a minor decrease in heart rate variability [iss.pdf/p228].

 

Ocular Events

The clinical studies did not suggest a drug-associated increase in the occurrence of glaucoma.  In the one-year, placebo-controlled studies, glaucoma was reported in two patients receiving tiotropium and one patient receiving placebo [iss.pdf/p238].  In the one-year, ipratropium-controlled studies one case of glaucoma was reported in a patient receiving tiotropium.  In the six-month studies, glaucoma was reported in one patient in the tiotropium group, one patient in the salmeterol group, and two patients in the placebo group.  In a study evaluating the safety following ocular administration of single increasing doses of a solution of tiotropium ranging from 0.02 to 0.40 mcg, tiotropium did not increase pupillometric pressure or affect pupillary diameter in healthy volunteers (Study 205.138).

 

D.        Adequacy of Safety Testing

The safety assessments performed in the pivotal studies were generally satisfactory, with one exception.  The cardiac safety database is limited and does not provide sufficient evidence of cardiac safety for this drug.  There are several reasons to be concerned about possible cardiac effects of tiotropium.  First, anticholinergic drugs, such as tiotropium, might be expected to have effects on cardiac rate and rhythm.  Second, the drug is associated with detectable plasma concentrations, particularly with chronic use.  Third, underlying cardiac disease is common in the proposed patient population.  As with most clinical development programs, subjects with significant cardiac disease (e.g. myocardial infarction within 1 year, heart failure within three years, cardiac arrhythmia requiring drug therapy, and significant disease other than COPD), subjects with hypoxemia requiring daytime oxygen therapy, and subjects with a creatinine >2.0 mg/dL were excluded from participation in the Phase 3 clinical studies.  Such patients, who will receive the drug if it is approved, may be at increased risk of adverse drug-related cardiac effects.  Finally, because of the large volume of distribution and long elimination half-life, subjects who develop adverse drug effects will continue to be exposed for weeks after discontinuing the drug.

 

The cardiac safety database includes insufficient Holter monitor data.  Holter monitors were performed in only one study (205.123).  In that study, “baseline” Holters included a period of time on drug, complicating the interpretation of the comparison of baseline to on-treatment data.  On-treatment Holters were only available for 37 patients treated with the proposed dose in the morning, 35 patients treated with the proposed dose in the evening, and 31 placebo patients.  For comparison, the product label for Serevent Inhalation Aerosol (GlaxoSmithKline) indicates that Holter monitoring was performed on 284 COPD patients during five 24-hour periods

 

Although the ECG monitoring in the one-year, placebo-controlled studies was less than optimal because the on-treatment ECGs were not obtained at or near the expected Cmax (and may have been obtained pre-dose), the ECG database is supported by the timed ECGs from the multiple-dose, dose-ranging study (205.108). 


 


 

 

E.                 Four-Month Safety Update

The Four-Month Safety Update, dated April 18, 2002, was submitted electronically.  The references cited in this section of the Clinical Briefing Document refer to the April 18, 2002, submission.  The submission included an updated Integrated Summary of Safety including new safety data covering the period of December 14, 2000 to December 13, 2001, and Clinical Trial Reports for two studies (205.131 and 205.222).  Study 205.131 is discussed briefly below.  Study 205.222 was a study of the effect of concomitant cimetidine and ranitidine once daily on the single dose pharmacokinetics of tiotropium, performed in 18 subjects in Germany [iss.pdf/p93].  The updated ISS includes preliminary unblinded safety data from four studies [iss.pdf/p24]:

-        an exercise study with a treatment period of six-weeks (205.131) [iss.pdf/p28];

-        a study evaluating changes in inspiratory capacity with a treatment period of fur-weeks (205.218) [iss.pdf/p28];

-        a study to evaluate the effect of a single dose of ipratropium after 19 days of tiotropium treatment in healthy volunteers aged 40-65 years(205.239) [iss.pdf/p30];

-        and a placebo-only HandiHaler ease-of-use and learning retention study (205.220) [iss.pdf/p28].


 

However, the updated ISS safety database includes only 18 subjects not reported in the original ISS [iss.pdf/p109].  These are the 18 healthy volunteers who participated in the IV pharmacokinetic trial (single doses of 14.4 mcg).  The  preliminary safety data from the four unblinded studies listed above, are discussed separately [iss.pdf/p287-95].  Review of that discussion did not reveal any new potential safety concerns.

 

The submission also provides information on 4 previously unreported deaths, which occurred in Study 205.214, an ongoing study evaluating the effect of tiotropium on the severity and incidence of COPD exacerbations [iss.pdf/p277-9].  The causes of death were pulmonary embolism, moncytic leukemia, myocardial infarction, and intestinal obstruction (post-operative).  The treatment assignment has not been unblinded.