Inspections, Compliance, Enforcement, and Criminal Investigations
Smithkline Beecham Pharmaceuticals Co. 01-Jul-02
Food and Drug Administration
466 Fernandez Juncos Avenue
Puerta De Tierra
San Juan, Puerto Rico 00901-3223
July 1, 2002
Mr. Jean-Pierre Gamier
Chairman Executive Officer
SB Pharmco Puerto Rico &
Smithkline Beecham Pharmaceuticals Co.
1 Frankline Plaza, 16th & Vine Street
P.O. Box 7929
Philadelphia, PA 19101
Dear Mr. Gamier
On 2/7, 8, 12-14, 20-22, 25, 26, 3/11, 5, 7, & 11 -13 & 4/l0/02, the Food and Drug Administration (FDA) conducted an inspection of your drug manufacturing facility SB Pharmco Puerto Rico, Inc. and Smithkline Beecham Pharmaceuticals Co. located al Rd. 172, Km 9.2,Bo. Certenejas Cidra, Puerto Rico 00739. The inspectional findings revealed that your products Bactroban Ointment, Paxil Oral Suspension and Thorazine tablets are adulterated within the meaning of Section 501(a)(2)(B) of the Federal Food, drug and Cosmetic Act; and in violation of Title 21 Code of Federal Regulations, Parts 210 & 211. The CGMP violations documented during the most recent inspection include:
Quality Control Unit
I. Failure of your quality control unit to exert its responsibility and authority as required by 21 CFR 211.22 to reject all drug products that fail to meet the established specifications; to assure that all records related to a product manufactured are adequately reviewed and that no errors that may impact the identity, strength, purity and/or safety of the drug products have occurred, or if an error has occurred that an investigation including corrective actions is performed in a timely manner. You also fail to have procedures in place to prevent microbial contamination of products as requited by 21 CFR 211.l 13. For example:
a. Bactroban 2% Ointment lot #50-lB25 was released to the market on June 1, 2001 contaminated with Pseudomonas fluorescens. However, it was not until 10/26/01].
during an Annual Review of the Microbiology profile of Bactroban Ointments 22g, that your firm noticed that this lot had been released with Pseudomonas contamination. No action was taken against the lot until a conference meeting was held with the FDA on 2/15/02. When you were informed of the serious cocerns the agency had with the affected lot, a recall was initiated. This incident demonstrates inadequate QA/QC control over product quality, your review and decision-making processes.
Although in your response you suggest the implementation of a Product Incident Review Committee, the response does not describe the manner in which this committee will prevent recurrence of a similar situation in the future.
b. Failure to assure that batches of Baclobran ointment, contaminated will questionable gram negative microorganisms (e.g. CDC Group IV C-2/ Ralstonia paucula, Bulkolderia picketti), were not released to the market. It was not until FDA raised a serious concern regarding Bactroban ointment batch 29-1B25,62-1B25,84-1B25,94-1B25 and 106-lB25, and Bactoderm Ointment, lot #32-lB25, that you decided to recall the questionable lots. This example also shows a lack of adequate procedures in place to prevent contaminated products from being released for distribution.
Production and Process Controls
2. Failure to assure batch uniformity and integrity of drug products and to assure that the production and process are adequate to consider your process as validated and in conformance with the specifications as required by 21 CFR 211.100 and 21 CFR 211.110. For example:
a. Failure to have a validated process for Paxil Oral Suspension (OS). Inspections that ended on 6/l3/01 and 4/10/02, respectively, have documented your failure to manufacture Paxil Oral Suspension in accordance to the established specifications and to demonstrate that you have a reproducible and reliable manufacturing process,. The inspection of 6/01 reported OOS assay uniformity results for batches of Paxil OS lot Xl-8Pl5 ([redacted], lot X2-8P15 [redacted] and lot X4-0p15 (validation batch rejected).
The current inspection revealed 00s assay uniformity results for the following lots of Paxil OS: lot Xl-lP15 (failed during stability and rejected), X7-lP15 (rejected due to OOS). After the process was modified lot X9-LPI5 also obtained OOS results. Following a process modification, development batch #X10-1P15 showed individual assay uniformity OOS results for samples collected after the initial mixing and for samples collected from the [redacted] prior to filling. A new validation was performed and lot Xll-lP15 also obtained low OOS assay uniformity results. The process was again modified and batch X6-2Pl5 (commercial scale) showed an OOS assay uniformity result in the sample collected from the top of the mixing tank; batch X7-2Pl5 (commercial scale) also showed an OOS result of [redacted] ml from a sample collected from the top and lot X8-2P15 (commercial scale) showed 00s assay uniformity results for samples representing the top and middle or the mixing tank.
b. Failure to have a robustly developed manufacturing process in that:
- approximately 50% of the total batches manufactured (pilot/development /commercial scale batches) have shown an assay uniformity 00s result at one of the critical stages of the process (after initial [redacted] minutes, after mixing for [redacted] additional minutes or in filled bottles).
- multiple process changes made as, a result of these uniformity failures have not adequately resolved batch inconsistency problems.
We also note that your revised process validation for Paxil OS, including lots X11-1P15, Xl2- 1P15, Xl-2P15 and X2-2P15, required that approximately [redacted] Kg of the batch (representing the last [redacted] of the total batch) be discarded because high assay results have been sporadically observed towards the end of the filling process.
Although some investigations have attributed several OOS results to unexpected manufacturing events, records confirm that OOS assay uniformity results have been obtained for samples collected from the either the top, middle, or bottom of the mixing lank or from filled bottles representing the end of the batches.
FDA has determined this product is medically necessary and we have informed you that we are exercising regulatory discretion with respect to lots of Paxil Oral Suspension released to the market not showing OOS results. However, it is still your responsibility to assure that all products are manufactured in accordance to the Good Manufacturing Practice regulations.
C. Failure to adequately validate the manufacturing process for Torazine Tablets 10 mg, 25 and 200 mg tablets. The records reviewed indicate that the thickness and hardness parameters for Torazine 10 mg were out of the established specifications during the validation runs. The records also show that validation lots of Torazine 25 mg and 200 mg, respectively, were not tested for friability nor content uniformity. In addition, your third validation lot for Torazine (#X301-9T74J) failed the assay test during the 3 month stability interval.
Your response does not adequately address these deficiencies and only indicates that you, will revalidate the process prior to releasing additional lots to the market. It makes no reference to lots or portions of lots that were released to the market to assure that the distributed product conforms to its quality standard. Please include in your response the assay results of all the retain sample lots, represented by X301-9T74J, that were tested (if any) as part of the investigation.
3. Failure to conduct investigations in a timely manner and to take corrective actions to prevent recurrence in accordance to 21 CFR 211.192. For example:
a. Investigation No. 0lLB190 includes an incident related to High Total Plate Count results obtained in water sampling point WIG2-014-H02 collected on September 5, 01. The investigation was not completed until 12/l5/01. The preventive and corrective actions were not verified until 2/12/02.
b. Investigation No. OlLI3201 related, to a Total Plate Count OOS result obtained on October 3, 2001 for sampling point DW-1-017-H13 located in the production area.
Sphingomonas paucimobilis was isolated from the sample. The investigation was not concluded until December 18, 2001.
The records reviewed fail to demonstrate that the corrective actions were implemented and verified.
During our review of the records provided the following significant deficiencies related to media fill lot #802-lSTM14 were also noticed:
Media fill vials are not being incubated for the required lime that would assure bacterinl growth for both slow and fast growth microorganisms. The records show that the above media fill vials were incubated at conditions of 22.52 +/- 2.5?C for a period of only 4 days, and nor for 7 days as required. Media fills units should be incubated for 14 days.
The information presented in the Media Fill Certificate is questionable and shows a significant discrepancy. This record indicates that the media fill was performed on l/30/02 and that the vials were incubated on 11/30/02. This same record also appears signed by the Q.A. Leader on 01/28/01. The inspection was conducted during the period of 2/7- 4/10/02. According to the media till records the exercise was performed on 1l/30/01, which raises a question regarding the accuracy of the records provided.
You should take prompt action to correct these deviations and to prevent their future recurrence. Failure to make prompt corrections could result in regulatory action without further notice. Possible actions include seizure and/or injunction.
Please notify this office in writing, within 15 working days of receipt of this letter, of the specific steps you have taken to correct the noted violation, including an explanation of each step being taken to prevent the recurrence of similar violations. Copies of revised labeling for the products should also be submitted, If corrective actions can not be completed within 15 working day, state the reason for delay and the time within which corrections will be completed.
Your reply should be sent to he Food & Drug Administration, Sari Juan District Office, 466 Fernandez JUNCOS Ave., San Juan, PR 00901-3223, Attention: Carmelo Rosa, Compliance Officer.
C. James Shen Ph.D.
Acting District Director