Statistical Reviewer’s Guide (SRG)

Study ID: ONC-201
Submission Type: NDA
Indication: Metastatic Non-Small Cell Lung Cancer (NSCLC)
Version / Date: v1.0 / 15-Mar-2026

1. Study Overview

Item	Description
Study Title	A Phase II, Randomized, Double-blind Study of Drug X versus Placebo in Patients with Metastatic NSCLC
Phase	Phase II
Study Design	Randomized, double-blind, placebo-controlled, parallel-group
Analysis Objective	To evaluate the efficacy and safety of Drug X
Analysis Type	Confirmatory (primary), Exploratory (secondary)

2. Analysis Data Overview (ADaM)

ADaM Dataset	Description	Granularity (One record per…)	Key Variables
ADSL	Subject-level analysis dataset	Subject	TRT01A, ITTFL, SAFFL, AGE, SEX
ADAE	Adverse events analysis dataset	AE per subject	AETERM, AESEV, AESER, TRT01A
ADLB	Laboratory analysis dataset	Test per visit per subject	PARAMCD, AVAL, BASE, CHG, AVISIT
ADTTE	Time-to-event dataset	Subject	CNSR, AVAL, PARAMCD

3. Analysis Populations

Population Name	Flag Variable	Definition
Intent-to-Treat (ITT)	ITTFL = 'Y'	All randomized subjects
Safety	SAFFL = 'Y'	Subjects who received ≥1 dose of study treatment
Per-Protocol (PP)	PPFL = 'Y'	ITT subjects without major protocol deviations

4. Endpoint Definitions and Derivations

4.1 Primary Endpoint

Item	Description
Endpoint Name	Progression-Free Survival (PFS)
Endpoint Description	Time from randomization to disease progression or death
ADaM Dataset(s)	ADTTE
ADaM Variable(s)	AVAL (time), CNSR
Baseline Definition	Date of randomization
Derived Variable	AVAL = (event date − randomization date + 1)
Time Point	Through data cutoff date

4.2 Key Secondary Endpoint – Objective Response Rate (ORR)

Item	Description
Endpoint Name	Objective Response Rate
ADaM Dataset(s)	ADRESP
ADaM Variable(s)	AVALC (response), ANL01FL
Derivation Summary	Best overall response derived from tumor assessments

5. Handling of Missing Data

Analysis	Missing Data Approach	Notes
PFS	Censoring	Subjects without event censored at last adequate assessment
ORR	Non-responder imputation	Subjects without post-baseline assessment treated as non-responders

6. Statistical Methods Summary

Endpoint	Method	Covariates	Multiplicity Handling
PFS	Stratified Cox proportional hazards model	Treatment, baseline disease stage	Hierarchical testing
ORR	Cochran-Mantel-Haenszel test	Stratification factors	Controlled within hierarchy

Detailed statistical specifications are provided in the SAP.

7. Traceability: ADaM → TLF

8. Sensitivity and Supportive Analyses

Analysis	Difference from Primary	Rationale
Sensitivity PFS	Alternative censoring rules	Assess robustness to censoring assumptions
Supportive ORR	Excluding non-evaluable subjects	Sensitivity to response definition

9. Known Data Issues and Statistical Limitations

Issue ID	Description	Potential Impact	Mitigation
SRG-01	Missing tumor assessments at later visits	Potential bias in ORR	Conservative non-responder imputation
SRG-02	Early study discontinuations	Reduced power for PFS	Sensitivity analyses performed

10. Appendices

Appendix A. Key Variable Mapping

Appendix B. Abbreviations