Poster Board II-674
Knowledge of how the disease characteristics and initial treatment (tx) patterns of FL affect the risk of transformation is limited to a few retrospective and single institution studies. Specifically, retrospective studies have shown an increased incidence of transformation in FL patients (pts) undergoing initial observation (J Clin Oncol 2007;25:2426), a decreased incidence in FL pts receiving aggressive, anthracycline based therapy (J Clin Oncol 2008;26:5165), and no difference in risk in FL pts receiving rituximab as first-line versus salvage therapy (Blood 2008;112:837 abstr). The NLCS, a multicenter, longitudinal observational study of 2734 pts with newly diagnosed FL, offers a unique opportunity to better characterize the risk of transformation. Pts who had evidence of transformation at initial diagnosis (n = 27) were excluded from this analysis. At a median follow-up of 37 months, 139 pts (5.1%; 95% CI 4.3% % 6.0%) developed a transformation; 47 pathologically confirmed and 92 clinically suspected. The overall survival curves for the confirmed and suspected transformation cases were not different, indicating similar outcomes for these patient groups. In univariate analysis, > 1 extranodal site (p < 0.001), LDH > normal (p < 0.001), FLIPI score ≥ 3 (p < 0.001), and advanced stage (p = 0.033) at diagnosis were associated with an increased risk of transformation. Age, histologic grade, ECOG performance status, B symptoms, number of nodal sites, hemoglobin, and IPI scores were not associated with transformation. Only > 1 extranodal site remained significant on multivariate analyses of all cases (OR = 1.30; 95% CI 1.06 – 1.61; p = 0.014). In univariate analysis, the risk of transformation was the same in 481 initially observed pts compared to 2253 treated pts (HR 0.99; 95% CI 0.64 – 1.54; p = 0.98). Of treated pts, the risk of transformation for the 944 pts who received first-line anthracyclines was not statistically different (although a trend exists) from the 1309 pts treated with a non-anthracycline regimen (HR 0.73; 95% CI 0.50 – 1.06; p = 0.096). Compared with 450 pts initially treated with non-rituximab containing regimens, the 1803 rituximab treated pts had a lower risk of transformation (6.9% vs. 4.7%), with a HR of 0.59 (95% CI 0.39 – 0.90; p = 0.012). The median follow-up was not significantly different in any of the arms for the subset analyses. Controlling for extranodal sites and FLIPI scores for the initially observed vs. treated, anthracycline vs. non-anthracycline and the rituximab vs. non-rituximab treated groups produced results nearly identical to the respective univariate comparisons. All results were similar in sensitivity analyses comparing confirmed vs. suspected cases, with the exception of a stronger inverse association of rituximab use in confirmed (HR 0.40; 95% CI 0.21 – 0.74; p = 0.003) vs. suspected cases (HR 0.78; 95% CI 0.45 – 1.36; p = 0.38). These findings demonstrate an incidence of FL transformation which is comparable to that reported historically. While most pts received early tx, these prospective data suggest that initial observation does not increase short term risk of transformation. These results also suggest that for initially treated pts, use of an anthracycline does not impact transformation, while use of rituximab may lower the risk of transformation. These results must be interpreted in the context of the limitations of a non-randomized observational study, and replication in other studies as well as longer follow-up will be needed to determine the overall influence of different therapeutic strategies on the risk of transformation.
Link:Genentech: Membership on an entity's Board of Directors or advisory committees, Research Funding. Taylor:genentech: Employment.
Asterisk with author names denotes non-ASH members.