FRCR ONCOLOGY PART 1 - ALL FREE PRACTICE MCQS: FRCR Oncology Part 1: MEDICAL STATISTICS - 12 (20 QUESTIONS, ANSWERS, EXPLAINATION BELOW)

In a phase II/III seamless adaptive trial evaluating a novel PARP inhibitor for BRCA-mutated pancreatic cancer, the interim analysis uses a Bayesian predictive probability threshold of 0.85 to proceed to phase III. Given a posterior probability of success of 0.88 based on 60 patients (40 responders), what is the most appropriate decision?
A. Stop the trial for futility
B. Proceed to phase III
C. Expand the phase II cohort
D. Adjust the randomization ratio
E. Terminate the trial for efficacy
A multi-arm, multi-stage (MAMS) trial for metastatic colorectal cancer tests three experimental therapies against a control arm, using a drop-the-loser design with a family-wise error rate (FWER) controlled at 0.05 via Dunnett correction. If the interim analysis p-value for arm B is 0.015, what is the most likely outcome for arm B?
A. Arm B is dropped for futility
B. Arm B continues to the next stage
C. Arm B is declared superior
D. Arm B requires sample size re-estimation
E. Arm B is terminated for ethical reasons
In an RCT comparing proton FLASH radiotherapy to conventional IMRT for glioblastoma, an interim analysis employs O’Brien-Fleming boundaries with α-spending of 0.005 at 50% information fraction. If the z-score is 2.8, what is the appropriate action?
A. Continue the trial
B. Stop for efficacy
C. Stop for futility
D. Adjust the sample size
E. Re-randomize patients
A basket trial evaluates a KRAS inhibitor across five tumor types (lung, colon, pancreas, bile duct, endometrial). The trial uses a Simon’s two-stage design per cohort with a type I error of 0.05 and power of 0.9. If the lung cohort has 8/20 responses in stage 1 (minimum 7 required), what is the next step?
A. Stop the lung cohort for futility
B. Proceed to stage 2 for the lung cohort
C. Expand the lung cohort sample size
D. Terminate the entire trial
E. Adjust the response threshold
In a platform trial evaluating multiple immunotherapies for NSCLC, a response-adaptive randomization algorithm increases allocation to the most promising arm. If arm A has a response rate of 45% (n=50) and arm B has 30% (n=50), what is the most likely adjustment?
A. Equal allocation to all arms
B. Increased allocation to arm A
C. Increased allocation to arm B
D. Termination of arm A
E. Termination of arm B
A phase III trial of a CDK4/6 inhibitor in HR-positive breast cancer uses a group sequential design with Lan-DeMets boundaries. At 75% information fraction, the z-score is 3.1, and the boundary is 2.9. What is the appropriate action?
A. Continue the trial
B. Stop for efficacy
C. Stop for futility
D. Adjust the sample size
E. Re-estimate the effect size
In an umbrella trial for pancreatic cancer, patients are stratified by KRAS mutation status. The trial uses a Bayesian hierarchical model to borrow information across strata. If the posterior probability of efficacy for KRAS G12D is 0.92, what is the most likely decision?
A. Stop the stratum for futility
B. Declare efficacy for the stratum
C. Expand the stratum sample size
D. Adjust the borrowing parameter
E. Terminate the entire trial
A trial of carbon ion therapy for osteosarcoma employs a futility boundary based on a beta-binomial model. If the interim response rate is 10/30 (threshold 12/30), what is the most appropriate action?
A. Continue the trial
B. Stop for futility
C. Stop for efficacy
D. Adjust the sample size
E. Re-randomize patients
In a factorial trial evaluating chemotherapy (A vs. no A) and radiotherapy (B vs. no B) for head and neck cancer, the interaction term has a p-value of 0.04. What does this indicate?
A. No interaction between treatments
B. Significant interaction between treatments
C. Chemotherapy is superior
D. Radiotherapy is superior
E. Trial is underpowered
A crossover trial of two sequencing strategies for metastatic melanoma uses a mixed-effects model to account for period effects. If the carryover effect p-value is 0.01, what is the most appropriate action?
A. Ignore the carryover effect
B. Analyze only the first period
C. Adjust for period effects
D. Terminate the trial
E. Increase the washout period
In a non-inferiority trial of a biosimilar chemotherapy agent for lymphoma, the hazard ratio for progression-free survival is 1.05 (95% CI: 0.92–1.20) with a non-inferiority margin of 1.25. What is the correct conclusion?
A. Biosimilar is inferior
B. Biosimilar is non-inferior
C. Biosimilar is superior
D. Result is inconclusive
E. Trial is underpowered
A cluster-randomized trial evaluates a screening program for colorectal cancer across 20 clinics. If the intraclass correlation coefficient (ICC) is 0.03, what is the primary impact on sample size?
A. No impact
B. Decreased sample size
C. Increased sample size
D. Invalidates randomization
E. Reduces power
In a trial of proton therapy for pediatric medulloblastoma, a Bayesian adaptive design uses a posterior probability threshold of 0.90 for efficacy. If the posterior probability is 0.87 after 50 patients, what is the most appropriate action?
A. Declare efficacy
B. Stop for futility
C. Continue recruitment
D. Adjust the threshold
E. Terminate the trial
A seamless phase I/II trial of a novel immunotherapy for sarcoma uses a continual reassessment method (CRM) for dose-finding. If the posterior probability of dose-limiting toxicity at dose level 3 is 0.25 (target 0.30), what is the next step?
A. Escalate to dose level 4
B. De-escalate to dose level 2
C. Continue at dose level 3
D. Stop the trial
E. Expand the cohort
In a trial of adjuvant therapy for breast cancer, the primary endpoint is disease-free survival (DFS). If the trial uses a Haybittle-Peto boundary (p<0.001) for interim analysis, what is the primary advantage?
A. Increased power
B. Reduced type I error
C. Simplified analysis
D. Flexible stopping
E. Higher sample size
A platform trial for glioblastoma tests multiple targeted therapies. If the response rate for arm C is 20% (n=40) and the historical control rate is 10%, what is the most appropriate statistical test?
A. Chi-square test
B. Fisher’s exact test
C. Log-rank test
D. T-test
E. Wilcoxon rank-sum test
In a trial of carbon ion MRT for chordoma, the interim analysis uses a Pocock boundary with α=0.05. If the z-score is 2.1 and the boundary is 2.3, what is the appropriate action?
A. Stop for efficacy
B. Stop for futility
C. Continue the trial
D. Adjust the sample size
E. Re-randomize patients
A basket trial for MSI-high cancers uses a Bayesian hierarchical model. If the posterior probability of response for the colorectal cohort is 0.95, what is the most likely decision?
A. Stop the cohort for futility
B. Declare efficacy for the cohort
C. Expand the cohort
D. Adjust the model parameters
E. Terminate the trial
In a factorial trial of immunotherapy (A vs. no A) and chemotherapy (B vs. no B) for NSCLC, the main effect of A has a p-value of 0.02. What does this indicate?
A. Immunotherapy is ineffective
B. Immunotherapy is effective
C. Interaction is significant
D. Chemotherapy is superior
E. Trial is underpowered
A trial of alectinib for ALK-positive NSCLC uses a Simon’s optimal two-stage design. If stage 1 requires ≥5/15 responses and 6 responses are observed, what is the next step?
A. Stop for futility
B. Proceed to stage 2
C. Expand stage 1
D. Terminate the trial
E. Adjust the threshold

ANSWERS

B. Proceed to phase III
The posterior probability of success (0.88) exceeds the Bayesian predictive threshold (0.85), indicating sufficient evidence to proceed to phase III in the seamless adaptive trial.
B. Arm B continues to the next stage
The p-value (0.015) is less than the Dunnett-corrected threshold (α ≈ 0.0167 for three comparisons), allowing arm B to continue in the MAMS trial.
A. Continue the trial
The z-score (2.8) is below the O’Brien-Fleming boundary at 50% information (z > 3.471 for α = 0.005), so the trial continues.
B. Proceed to stage 2 for the lung cohort
The lung cohort meets the Simon’s two-stage criterion (8 ≥ 7 responses in stage 1), allowing progression to stage 2.
B. Increased allocation to arm A
The response-adaptive randomization algorithm favors arm A (45% response rate) over arm B (30%), increasing its allocation probability.
B. Stop for efficacy
The z-score (3.1) exceeds the Lan-DeMets boundary (2.9) at 75% information, indicating sufficient evidence to stop for efficacy.
B. Declare efficacy for the stratum
The posterior probability of efficacy (0.92) is high, supporting a decision to declare efficacy for the KRAS G12D stratum in the Bayesian hierarchical model.
B. Stop for futility
The interim response rate (10/30) is below the futility threshold (12/30), warranting trial termination for futility.
B. Significant interaction between treatments
A p-value of 0.04 for the interaction term indicates a significant interaction, suggesting the effect of one treatment depends on the other.
B. Analyze only the first period
A significant carryover effect (p=0.01) suggests contamination across periods, so analyzing only the first period avoids bias.
B. Biosimilar is non-inferior
The upper bound of the CI (1.20) is below the non-inferiority margin (1.25), confirming non-inferiority of the biosimilar.
C. Increased sample size
An ICC of 0.03 indicates clustering, increasing the effective sample size required due to reduced independence of observations.
C. Continue recruitment
The posterior probability (0.87) is below the efficacy threshold (0.90), so recruitment continues to gather more evidence.
C. Continue at dose level 3
The posterior probability of toxicity (0.25) is close to the target (0.30), suggesting dose level 3 is appropriate for further evaluation.
B. Reduced type I error
The Haybittle-Peto boundary (p<0.001) is conservative, minimizing type I error during interim analyses.
B. Fisher’s exact test
For small sample sizes (n=40) and binary outcomes (response rates), Fisher’s exact test is appropriate to compare arm C (20%) to the historical control (10%).
C. Continue the trial
The z-score (2.1) is below the Pocock boundary (2.3), so the trial continues without stopping for efficacy.
B. Declare efficacy for the cohort
The posterior probability of response (0.95) is high, supporting efficacy declaration for the colorectal cohort.
B. Immunotherapy is effective
The main effect p-value (0.02) indicates immunotherapy has a significant effect, independent of interaction or chemotherapy.
B. Proceed to stage 2
The observed responses (6 ≥ 5) meet the stage 1 criterion, allowing progression to stage 2 in the Simon’s design.

Sunday, May 18, 2025

FRCR Oncology Part 1: MEDICAL STATISTICS - 12 (20 QUESTIONS, ANSWERS, EXPLAINATION BELOW)