Survival in Follicular Lymphoma: The Stanford Experience, 1960–2003.

Background: Recent data from retrospective and prospective studies indicate an improvement in the overall survival (OS) of follicular lymphoma (FL) patients (pt). We sought to evaluate survival data in FL pt managed at our institution and to analyze their presenting features, management and disease course.

Methods: Previously untreated FL (grades 1 and 2) pt referred from 1960–2003 and followed at Stanford were identified from the Lymphoma Database for this retrospective study. Data were reviewed for quality and consistency of coding for histology (pathology reviewed at Stanford) and treatment courses. Four eras were defined by date of diagnosis 1 (1960–75, n=180); 2 (1976–86, n=426); 3 (1987–96, n=471); and 4 (1997–2003, n=257). Patient (age, gender) and disease (stage, histology) characteristics, management (time from diagnosis to referral, time to first treatment, courses/types of treatment), and outcomes (progression-free survival [PFS], OS, transformation) were evaluated, overall and with respect to era.

Results: The OS for the 1334 FL pt in the dataset was 12.6 years (yr). Age (median = 49 yr), gender (54% male) and stage (83% III-IV) did not differ across the eras. Median OS improved from 11.2 yr [95% CI 10.2,12.1] in eras 1 and 2 (1960–86) to 18.4 yr [95% CI 14.5,22.4] in eras 3 and 4 (1986–2003), p<0.001. Greatest gains were observed in pt less than 60 yr old and in advanced stage disease. OS was not different in eras 3 and 4 (p=0.56) but median follow-up is just 5 yr in era 4. Although treatment varied over the eras, PFS was no longer for the later eras (3 and 4) compared to early eras (1 and 2), p=0.73. Initial therapy was more likely to be deferred in the later eras: 16%, 44%, 59%, 60% for eras 1, 2, 3 and 4, respectively. Rituximab use, as expected, varied over the eras: 1 (1% pt), 2 (8% pt), 3 (23% pt), 4 (42% pt). The incidence of transformation was similar over the eras and OS after transformation was significantly longer if patients were chemotherapy-naive at transformation (p<0.001).

Conclusions: OS of patients with FL managed at Stanford University significantly improved in the 1986–2003 eras, particularly among younger and advanced stage patients. This improvement was not associated with longer PFS after the first course of treatment or differences in the incidence of histologic transformation. Application of a time-variant Cox regression model is in progress to better understand the contributions of treatment, particularly rituximab, as well as other factors in these patients.