Abstract

FL is incurable with standard therapy, with no consensus on best initial management strategy. Two critical choices - deferral of initial therapy and use or not of anthracyclines - are often “individualized” per clinical circumstances. To date, no data exist regarding whether ‘disease-specific’, ‘patient-specific’, or ‘disease/patient-independent’ characteristics influence these two critical choices for initial therapy. The NLCS is the first prospective observational study in the US designed to assess FL baseline characteristics, prognosis, treatment choices, and clinical outcomes. We utilized this database to correlate the above characteristics with initial therapeutic choices regarding deferral of therapy and use of anthracyclines in pts with FL. NLCS enrolled pts with FL diagnosed within 6 months with no prior lymphoma history. Data collected included demographics, histology, staging, prognostic evaluation, therapy, response, relapses and death. As NLCS is an observational study, no prescribed treatment regimen or intervention is required. Therapy was considered deferred if none was administered within 3 months of diagnosis. Univariate and multivariate analyses were performed. 2254 pts with FL grades 1–3 were enrolled from 237 sites throughout the US from March 2004–January 2007. Median age was 61, 52% were female, and 90% were Caucasian. Grade distribution was 47% 1, 31% 2, and 22% 3. Stages were: I: 18%; II: 15%; III: 30%, and IV: 36%. Initial therapy was deferred in 18%. This frequency varied widely among subgroups defined by disease-, patient-, and provider-specific factors. Among disease-specific factors, strongest association was with grade (Gr1 24%, Gr2 15%, Gr3 9% {p<0.0001}), and with lesser, but still significant, correlation with FLIPI (low 22%, int 20%, hi 10% {p<0.0001}) and stage. Patient-specific factors associated with therapy deferral included age (<60 14%, 60–74 19%, >75 24% {p=0.0004}) and gender (F 20%, M 16% p=0.04}), but not ethnicity. Disease/patient-independent characteristics associated with therapy deferral included geographic region (South/West 14%, Midwest 18%, Northeast 29% {p <0.0001}) and lowest deferral rates with lowest oncologist density, but not academic vs. community treatment site. Initial management included anthracyclines in 34% of all FL pts. Among disease-specific factors, stronger association was with grade (Gr1 21%, Gr2 35%, Gr3 61% {p < 0.0001}), than FLIPI (low 31%, int 34%, hi 39% {p=0.01}). Patient-specific factors associated with anthracycline use included age (<60 41%, 60–74 33%, >75 14% {p<0.0001}) and gender (F 31%, M 37% p=0.0025}), but not ethnicity. Disease/patient-independent characteristics associated with anthracycline use included geographic region (South/West 39%, Midwest 32%, Northeast 21% {p<0.0001}), but not academic vs. community treatment site. As expected, disease-specific characteristics predict intensity of initial management of FL patients with histologic grade a stronger predictor than FLIPI scores. However, lower treatment intensity was also found for factors less clearly predictive of clinical outcome including Northeast location of treating physician and increased density of oncologists. It will be important to evaluate how groups with lower treatment intensity fare when outcome data mature.

Author notes

Disclosure:Employment: Michael D. Taylor and Ming Lin are employed by Genentech. Consultancy: Brian Link and Julie Vose have served as consultants for Genentech, Inc. Ownership Interests:; Michael D. Taylor and Ming Lin have ownership interests in Genentech Inc. Research Funding: University of Iowa and University of Nebraska have received research funding from Genentech Inc, Biogen Idec, and multiple other pharamceutical companies. Membership Information: Brian Link and Julie Vose serve on the Genentech’s Lymphocare advisory committee.