NDA 20-386, COZAAR (Losartan) Page 1

DEPARTMENT OF HEALTH AND HUMAN SERVICES

PUBLIC HEALTH SERVICE

FOOD AND DRUG ADMINISTRATION

CENTER FOR DRUG EVALUATION AND RESEARCH

STATISTICAL REVIEW AND EVALUATION

(Addendum)

NDA #: 20-386

SERIAL #: SEl-S028

DRUG NAME: COZAAR (Losartan Potassium) tablets

INDICATION: Renal Protection

SPONSOR: Merck & Co., Inc.

DOCUMENT REVIEWED:

1. Vols. 1 and 2 (CDER RECD Date: November 13, 2001)

2. SAS data base in EDR

STATISTICAL REVIEWER: H.M. James Hung, Ph.D. (HFD-710)

MEDICAL REVIEWER: Juan Carlos Pelayo, M.D. (HFD-110)

PROJECT MANAGER: Sandra Birdsong (HFD-110)

STATISTICAL KEY WORDS: Interim analysis, DSMB, sample size change, log rank test, O'Brien-Fleming boundary, Cox regression, geographical region

Distribution: NDA 20-386, SE1-S028

HFD-110/Dr. Throckmorton

HFD-110/Dr. Stockbridge

HFD-110/Dr. Pelayo

HFD-110/Ms. Birdsong

HFD-700/Dr. Anello

HFD-710/Dr. Chi

HFD-710/Dr. Mahjoob

HFD-710/Dr. Hung

HFD-710/chron

JHung/301-594-5436/DB 1/cozaar.doc/4-1-2002

 

 

NDA 20-386, COZAAR (Losartan) Page 2

SUMMARY

 

 

This addendum is to remove the p-values of the component events (doubling serum creatinine, ESRD or death) as the "first" event of the primary composite endpoint, given in Table 1 of my review of March 1, 2002. The removal is to avoid possible misinterpretation of these p-values which originally were provided purely for the description purpose, not for statistical testing. As the result, Table 1 in the original review is changed to Table 1a give below.

Table 1a. Incidence of adjudicated primary events (Reviewer's Analysis)

 

Losartan

=751

Placebo

=762

Hazard ratio

95.2% CI

p-value$

Primary event

Doubling serum

creatinine, ESRD or

death

327 (43.5%)

Median time

= 1303 days

359 (47.1%)

Median time

=1373 day s

0.84 (0.72, 0.97)

0.022

Decomposition of the ls`

primary event

Doubling of SC

162 (21.6%)

198 (26.0%)

0.75 (0.61, 0.92)

ESRD

64 (8.5%)

65 (8.5%)

0.93 (0.65, 1.31)

 

Death

101 (13.5%)

96 (12.6%)

0.98 (0.74, 1.30)

 
 

1st component endpoint

Doubling of SC

162 (21.6%)

198 (26.2%)

0.75 (0.61, 0.92)

0.006

ESRD

147 (19.5%)

194 (25.5%)

0.71 (0.57, 0.89)

0.002

Death

158 (21.0%)

155 (20.3%)

1.02 (0.81, 1.27)

0.88

ESRD or death

255 (34.0%)

300 (39.4%)

0.80 (0.68, 0.95)

0.009

$ nominal p-value, pre-specified primary analysis (Cox model using geographical region as covariate and baseline

proteinuria as a stratification variable)

As said in the conclusion section of my review of March 1, 2002, there was some evidence that losartan might reduce the incidence of the primary composite endpoint, doubling of serum creatinine, ESRD or death, (reduction of risk = 16% with 95% CI of 3% to 28%, p = 0.022). The strength of evidence did not meet the usual standard of statistical evidence, at least lower by an order of magnitude in p-value. Approximately a half of the primary composite events were doubling of serum creatinine and the other half were ESRD or death as the first event. The treatment difference seemed to be largely shown by doubling of serum creatinine. However, in the patients with doubling serum creatinine, 51 % developed ES1RD in the losartan group and 65% in the placebo group; mean time from doubling serum creatinine to ES1RD was about 30 days longer in the losartan group in these patients. In addition, during the study period, 19.5% of the losartan patients and 25.5% of the placebo patients developed ESRD; thus, losartan appeared to give a 29% reduction in risk of having an ES1tD (p = 0.002). There is little difference in death rate between the losartan group and the placebo group.