Danish population-based study in DLBCL: post-treatment EFS for 24 months is a potential surrogate end point for clinical trials

In January 2017, in the Journal of Clinical Oncology, Lasse Hjort Jakobsen from Aalborg University Hospital, Denmark, and colleagues published the results of a retrospective study of 1,621 newly-diagnosed DLBCL patients entered into the Danish Lymphoma Registry between 2003 and 2011.

The aim was to compare the survival of patients in remission following R-CHOP (or R-CHOP-like) treatment compared with a matched general population and to verify whether Post-treatment Event Free Survival for 24 months (pEFS24) could be used as a surrogate end-point in clinical trials.

 Key Highlights:

• pEFS defined as time from treatment end until censoring, relapse, or death
• General population matched to patient grouping based upon age, gender, and calendar year
• Median 5-year follow-up: general population pOS = 87% vs 78% DLBCL (95% CI, 76–80, P<0.001)
• Standardized mortality ratio baseline = 1.75 (95% CI, P<0.001), pEFS24 = 1.27 (95% CI, P<0.001), pEFS48 = 1.32 (95% CI, P<0.001)
• Residual lifetime estimation in CR/CRu DLBCL pts compared with matched general population:
  • Ten years post-treatment = 1.07 mo/y reduction (95% CI, 0.87–1.27 mo/y)
  • pEFS24 pts = 0.31 mo/y (95% CI, 0.11–0.5 mo/y)
  • pEFS36 pts = 0.29 mo/y (95% CI, 0.09–0.49 mo/y)
  • pEFS48 pts = 0.29 mo/y (95% CI, 0.08–0.51 mo/y)
• Pts older than 50yr had higher mortality
• Five-year overall DLBCL relapse incidence rate = 18% (95% CI, 16%–20%), or in pEFS24 pts = 8% (95% CI, 7%–10%)
• Five-year risk of DLBCL relapse pEFS24 pts ≤ 50yr = 4% (95% CI, 2%–8%), 50–60yr 7% (95% CI, 5%–11%), ≥ 60yr = 10% (95% CI, 8%–12%)

In conclusion, the authors stated that the data in this study does not support other data showing that following pEFS24 DLBCL pts had normalized survival. Loss of residual lifetime was found to be significant for pts achieving pEFS24, albeit minimal. Furthermore, the authors state that although pEFS24 may be useful as a surrogate clinical trial end point, it does risk losing the effects of late adverse events and toxicities. 

Abstract:

Purpose. The general outlook for patients with diffuse large B-cell lymphoma (DLBCL) in first remission is important information for patients and for planning post-treatment follow-up. The purpose of this study was to evaluate the survival of patients with DLBCL in remission compared with a matched general population. Methods. A total of 1,621 patients from the Danish Lymphoma Registry who were newly diagnosed with DLBCL between 2003 and 2011 were included in this study. All patients were ≥ 16 years of age at diagnosis and had achieved complete remission or complete remission unconfirmed after first-line rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) or R-CHOP–like therapy. Results. The 5-year post-treatment DLBCL survival was inferior to survival in the matched general population (78%; 95% CI, 76 to 80; v 87%; standardized mortality ratio, 1.75; P < .001). Excess mortality was present but reduced for patients achieving post-treatment event-free survival for 24 months (pEFS24; standardized mortality ratio, 1.27; P < .001). In age-stratified analyses, the survival of patients < 50 years of age was normalized to the general population after achieving pEFS24 (P = .99). During the first 8 years after pEFS24, the average loss of lifetime was 0.31 mo/y (95% CI, 0.11 to 0.50 mo/y). Excess mortality diminished when analyzing death from lymphoma as competing event to death from other causes, suggesting that early and late relapse is responsible for increased mortality in patients with DLBCL. Conclusion. Although this population-based study does not support complete normalization of survival for patients with DLBCL achieving pEFS24, the estimated loss of residual lifetime was low for patients in continuous remission 2 years after ending treatment. Therefore, pEFS24 is an appealing and relevant milestone for patient counseling and could be a surrogate end point in clinical trials.