Page 160 - 18F-FDG PET as biomarker in aggressive lymphoma; technical and clinical validation

Chapter 6b
protocol. The interim PET scan after 4 cycles was performed for observational purposes only. All PET-CT scans were centrally reviewed by the HOVON Imaging Group according to standard procedures as previously described [16] using Deauville score (DS) for visual assessment[15].Scores of 1 to 3 were interpreted as complete metabolic response, and scores of 4 to 5 were consistent with partial metabolic response or progressive disease. CT scans of neck, chest, abdomen, and pelvis were required at 6, 12, 18, and 24 months after completion of induction treatment. Severity of adverse events was defined according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 3.0).
Sample Size Calculation and Statistical Analysis
This trial was designed to compare CR rates on induction treatment between R-CHOP-14 and RR-CHOP-14 (first randomization; R1) and compare failure- free survival (FFS) from second randomization (R2) between no further treatment and rituximab maintenance. The sample size for R1 was 575 patients, accrued over 5 years, with a power of 86% to detect an improvement in CR rate from 77% to 87%. Additional sample size calculation details are provided in the Appendix. The primary end point for R1 was CR on induction. Logistic regression analysis with adjustment for age group (18-65 v ≥ 66-80 years) and age-adjusted IPI score (0 v 1 v 2 v 3; categorical) was applied for the primary analysis, and odds ratios and 95% CIs were determined, with P values < .05 considered statistically significant. Secondary end points were best response on protocol treatment, adverse events, FFS, progression-free survival (PFS), and OS from R1 and disease-free survival (DFS) from CR. For the survival end points, the hazard ratios (HRs) and 95% CIs were determined using univariable and multivariable Cox regression analyses. Kaplan-Meier curves by treatment arm were generated to illustrate survival.
All analyses were performed according to the intention-to-treat (ITT) principle. However, patients initially randomly assigned but considered ineligible in retrospect based on information that should have been available before random assignment were excluded from the respective analyses (modified ITT). The proportion of patients with specific adverse events was compared between arms
2
post hoc using the χ test or Fisher’s exact test, whichever was appropriate. All
reported P values are 2 sided and were not adjusted for multiple testing. Additional details on statistical methods and survival end point definitions are provided in the Appendix.
158