Dawson, Keith

Abstract Introduction: Follicular lymphoma (FL) is the most common indolent non-Hodgkin lymphoma in the world and has a median age at diagnosis in the seventh decade. FL in young adults (YA; 40 years old or younger) is extremely rare. Currently, there are no standard approaches guiding treatment of YA patients with FL, and very little is known about disease characteristics and outcomes of YA patients with FL given limited research conducted in this vulnerable population. To gain further insight into FL in YA, we analyzed the National LymphoCare Study (NLCS) to describe disease and patient characteristics, as well as features of treatment in YA patients with FL. We previously reported that 2-year progression-free survival (PFS) is an important survival endpoint in patients with FL undergoing chemo-immunotherapy. Hence, we also sought to characterize 2-year PFS in this age group and compare it to older cohorts. Methods: Evaluablepatients were identified in the NLCS, and those between 18–40 years of age with newly diagnosed FL at any stage were classified as YA patients. Patients with mixed histology or transformed disease were excluded, as were patients with progression of disease prior to beginning first-line treatment. Survival probability was estimated by the Kaplan-Meier method. We estimated the association of age group with PFS using hazard ratios (HR) and 95% confidence intervals (CI) from multivariable Cox models. Results: A total of164 YA patients with FL were analyzed, representing 6.2% of the NLCS population, similar to the observed frequency in the Surveillance, Epidemiology, and End Results (SEER) Program data (4.8% of all FL). Sixty nine percent of YA patients had advanced stage disease. The majority of patients (80%) had low-grade histology, and 50% had good risk disease according to the Follicular Lymphoma International Prognostic Index (FLIPI). Nineteen percent of patients (31/164) underwent watchful waiting, 12% received rituximab monotherapy, and 47% received chemo-immunotherapy (61% of whom received R-CHOP [rituximab, doxorubicin, vincristine, prednisone]). There was no significant difference in FLIPI score or other baseline disease characteristics compared to adult patients aged 41–60 years. Eleven deaths occurred among YA with FL; only 5 of these were lymphoma related. Overall survival (OS) at 2 years was 97.4% (95% CI 93.3%, 99.0%), and at 5 years, 93.7% (88.3%, 96.7%), which was similar to patients aged 41–60 (97.2% [96.0%, 98.0%] at 2 years, and 92.0% [90.1%, 93.5%] at 5 years). After a median follow-up of 7.1 years, OS in YA FL was 92%. Through follow-up, there were 64 PFS events. The estimated 2-year PFS (95% CI) for YA and adults 41–60 was 75.9% (67.1%, 82.6%) and 80.9% (78.1%, 83.4%), respectively. After adjusting for FLIPI score, there was no difference in PFS for YA with FL requiring first-line treatment (excluding watchful waiting) compared to adults aged 41–60 years (HR=0.93; 95% CI 0.69, 1.25), and no difference in OS compared to adults aged 41–60 years (HR=1.19; 95% CI 0.64, 2.23). Conclusions: In the largest cohort of YA patients with FL to date, we found few differences in outcomes compared to patients aged 41–60. FLIPI and other disease characteristics were similar to adults aged 41–60 years. There were no differences between YA FL and adults aged 41–60 in PFS for all treated patients. OS in the YA group of patients with FL was outstanding. YA patients with FL have reassuringly similar outcomes to patients aged 41–60. Fertility preservation and survivorship issues should be taken into consideration when defining management strategies, but otherwise these data support that YA patients with FL should not be approached differently from older adults with the same disease. Disclosures Byrtek: Genentech, Inc.: Employment, Equity Ownership. Dawson:Genentech, Inc.: Employment, Equity Ownership. Zhou:RTI-HS: Employee of RTI-HS, which has research contracts with Genentech Other. Flowers:Seattle Genetics: Consultancy; Spectrum: Consultancy, Research Funding; Sanofi: Research Funding; Abbott: Research Funding; Novartis: Research Funding; OptumRx: Consultancy; Millennium/Takeda: Research Funding; Janssen: Research Funding; Celgene: Research Funding; Allos: Consultancy. Farber:Gilead: Speakers Bureau; Janssen/Pharmacyclics: Speakers Bureau; Seattle Genetics: Speakers Bureau; Leukemia Lymphoma Society NJ Chapter: Membership on an entity's Board of Directors or advisory committees; Genentech: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Celgene: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Alexion: Speakers Bureau, Stock ownership Other. Cerhan:Genentech, Inc.: LymphoCare Scientific Advisory Board Other. Link:Genentech, Inc.: Consultancy, Scientific Advisory Board for Genentech Other.

Abstract Background Disease characteristics, treatment patterns, and outcomes of follicular lymphoma (FL) patients (pts) aged >80 years (yrs) are infrequently reported. Further, the original Follicular Lymphoma International Prognostic Index (FLIPI) has not been formally studied in pts aged >80 yrs. Also, whether FLIPI is associated with overall survival (OS) in pts aged >80 yrs is unknown. Patients/Methods We used the National LymphoCare Study—a Genentech-sponsored, prospective, multicenter registry of FL pts across the United States without study-specific treatment—to comprehensively analyze FL pts aged >80 yrs. Using Pearson chi-squared tests, associations of age groups with disease characteristics and overall response rate (ORR) were examined. Median progression-free survival (PFS) and OS by treatment regimen were estimated using the Kaplan-Meier method. Cox proportional hazards regression that were adjusted for baseline disease factors were used to assess treatment differences in PFS, lymphoma-related mortality (LRM, which includes both disease and treatment-related deaths), and OS, as well as the significance of age by treatment interactions. FLIPI and a prognostic model that was constructed based on the findings were used to predict outcomes. Results Of 2649 evaluable pts, 209 (8%) were aged >80 yrs. Compared with pts aged ≤60, pts aged >80 yrs were more likely to be Caucasian, to have worse performance status, to have lower hemoglobin (Hgb) value (<12 g/dL), to have less advanced stage, and were less likely to be treated at academic institutions. When treated, pts aged >80 yrs had lower ORR compared with younger pts and were less likely to receive rituximab + chemotherapy (P <.001) or anthracyclines (P <.001). Use of maintenance rituximab was not significantly different between pt groups when rituximab induction was given (61% for pts aged ≤60, 38% for pts aged >80, P=.11), but it was more frequent in pts aged >80 yrs when rituximab + chemotherapy induction was administered (44% for pts aged ≤60, 59% for pts aged >80, P=.04). After adjusting for maintenance use, sex, and baseline factors, age significantly differentiated PFS impact of observation, rituximab monotherapy, and rituximab + chemotherapy (P=.01). No treatment regimen provided superior PFS or OS in pts aged >80 yrs. With a median follow-up of 6.5 yrs for all pts (4.3 yrs for pts aged >80 yrs), 5-year OS was 59% and 92%for patients aged >80 and ≤60 yrs, respectively. OS in pts aged >80 yrs varied based on FLIPI score (log-rank test, P=.01; Figure 1). Interestingly, the percentage of deaths that were disease-related was 40% in pts aged >80 yrs, which was comparable to the percentage in pts aged ≤60 yrs of 48%. Cox modeling showed that lower Hgb, B symptoms, and male sex predicted worse OS (P<.01) but not PFS in pts aged >80 yrs. Recognizing that males have inferior OS compared with females in the general population, we constructed a prognostic model composed only of B symptoms and Hgb level <12 (low risk=no factors; high risk=1 or more factors; P<.0001; Figure 2). FLIPI and the proposed prognostic model significantly differentiated both OS and LRM (Table 1). Conclusions In the largest, prospectively published study of FL pts aged >80 yrs, OS, PFS, ORR, and treatment selection varied compared with younger pts. Surprisingly, the proportion of deaths attributed to lymphoma in these older FL pts was 40%. FLIPI was associated with outcome in pts aged >80 yrs and so was a proposed prognostic model comprising lower Hgb and B symptoms. Independent validation of this model is required, and additional prospective studies that are designed for the oldest old are warranted. Disclosures: Nabhan: Genentech Inc.: Advisory Board Other, Honoraria. Byrtek:Genentech, a member of the Roche group: Employment, Equity Ownership; Roche: Equity Ownership, stock options, stock options Other. Dawson:Genentech: Employment; Roche: Equity Ownership. Link:Genentech: Consultancy; Millenium: Consultancy; Pharmacyclics: Consultancy; Spectrum: Consultancy; Genentech: Research Funding; Millenium: Research Funding; Pharacyclics: Research Funding. Zelenetz:GSK: Consultancy; Celgene: Consultancy; Cephalon: Consultancy; Gilead: Consultancy; Seattle Genetics: Consultancy; Sanofi-Aventis: Consultancy; Genentech: Research Funding; GSK: Research Funding; Roche: Research Funding; Cancer Genetics: Scientific Advisor Other. Cerhan:Genentech: Scientific Advisory Board of the National LymphoCare Study Other. Flowers:Abbott, Celgene, Millennium/Takeda, Sanofi, Spectrum, Janssen: Research Funding; Celgene, Genentech Bio-oncology: Consultancy.

Abstract Abstract 3708 Background: There are few prospective studies on disease characteristics, patterns of care, response, and outcomes in elderly FL patients (pts) in the US. The NLCS is a Genentech-sponsored prospective multicenter registry study that collects this information without study-specific treatment. We utilized the NLCS database to better understand the impact of age on FL outcome. Patients and Methods: All evaluable pts with FL histology in the NLCS were included except pts with FL plus other lymphoma histology or pts who progressed before first treatment or before being assigned to watchful waiting (WW). Using Pearson Chi-Square tests, associations of age groups (≤60, 61–70, >70) with disease characteristics and overall response (ORR) were examined. Median PFS and OS by treatment regimen were estimated using Kaplan-Meier methods for each age group. Cox proportional hazards regression adjusted for baseline factors (grade, number of nodal sites, LDH, Hgb, stage, performance status (PS), bone marrow (BM) involvement, race, and treatment center type) were used to assess treatment differences in PFS and OS and the significance of age by treatment interactions. Results: Of 2,647 pts, 47% (n=1,254) were ≤60 yrs, 25% (n=666) were 61–70 yrs, and 27% (n=727) were >70 yrs (min age of 22; max of 97). Compared with pts ≤60 yrs, pts 61–70 and >70 were more likely to be white (93% >70, 92% 61–70, and 88% ≤60, P=.02 and .02 respectively), have stage I/II disease (37% >70, 36% 61–70, and 29% ≤60, P=.0008 and .0003), have <5 nodal sites (73% >70, 69% 61–70, and 61% ≤60, P=.001 and <.0001), and have poor-risk FLIPI (53% >70, 51% 61–70, and 15% ≤60, P<.0001 and <.0001). Compared with pts ≤60, elderly pts (>70) were more likely to have FL grade 3 (24% vs 18%, P=.01). While there were no differences in geographic distribution by age, elderly pts were more likely to receive therapy at community practices (86%) versus academic institutions than pts ≤60 (77%, P<.0001) or 61–70 (81%, P=.004). Treatments varied significantly by age (P<.0001). More elderly pts were observed compared to pts ≤60 (23% vs19%). When treated, elderly pts (22%) were more likely to receive rituximab (R) monotherapy compared with patients aged 61–70 (12%) or ≤60 (10%). When chemotherapy alone or plus R was given, elderly pts were less likely to receive anthracyclines (45% vs 65% (61–70) and 68% (≤60)). Among all variables, only grade 3 histology predicted anthracycline use in all age groups. Lack of BM involvement predicted anthracycline use for younger pts (≤60 and 61–70). Of those ≤60, white pts were more likely to receive anthracyclines, and of those 61–70, those with ≥5 nodal sites were more likely to receive anthracyclines. ORRs were similar across age groups receiving similar regimens with R plus chemo providing the highest ORR. Adjusting for baseline factors, treatment (WW, R, R-Chemo, or other) benefit varied for each age group in terms of PFS (P=.02), with treatment outcomes being most differentiated among younger pts (Table). PFS appeared shorter in elderly pts regardless of the treatment received. No significant interaction between age and efficacy of anthracycline in terms of PFS or OS was observed (P-values >.65), but the overall effect of anthracycline for all pts was beneficial for PFS (HR=0.80, P=.02) and OS (HR=0.67, P=.003). Median OS was 8 years for elderly and not reached for others. After adjusting for baseline factors, no significant differences in treatment impact by age on OS were seen. Elevated LDH, reduced Hgb, stage III/IV, PS ≥1, and BM involvement were all significantly associated with shortened OS. These factors were also significantly associated with treatment choice, as worse-prognosis elderly pts were more likely to receive either R or R+chemo than WW or other treatment. Conclusions: FL pts >70 yrs more commonly received R alone and less commonly received anthracyclines when treated with chemotherapy. The impact of anthracyclines on PFS did not vary by age, but differences in PFS for other treatment regimens showed a stronger association among younger pts Disclosures: Nabhan: Genentech: Research Funding, Speakers Bureau. Byrtek:Genentech, Inc., a member of the Riche Group: Employment, Equity Ownership. Dawson:Genentech, Inc., a member of the Riche Group: Employment, Equity Ownership. Link:Genentech, Inc., a member of the Riche Group: Consultancy; Celgene: Consultancy; Spectrum: Consultancy. Friedberg:Genentech: Consultancy. Cerhan:Genentech: National LymphoCare Scientific Advisory Board Other. Flowers:Celgene: Consultancy; Prescription Solutions: Consultancy; Seattle Genetics: Consultancy; Millennium: Consultancy, Research Funding; Genentech: Consultancy; GIlead: Research Funding; Spectrum: Research Funding; Janssen: Research Funding; Lymphoma Research Foundation: Membership on an entity's Board of Directors or advisory committees.

Abstract Abstract 3702 Background: Both drug choice and delivered dose of intended treatment potentially influence outcome in patients treated for follicular lymphoma (FL). Historically, observational studies have solely evaluated drug choice. The National LymphoCare Study (NLCS) is a prospective, observational study that enrolled >2700 patients with FL in the United States between 2004 and 2007. We report on measures of delivered dose and its impact on outcomes for the most common first-line regimens. Methods: All evaluable patients treated with R-CHOP, R-CVP, or R-fludarabine (R-Flu)–containing regimens in the NLCS were included, except those with FL plus other lymphoma histology, non-FL histology, or those who progressed before first treatment or before being assigned to watchful waiting. Associations between baseline factors and choice of treatment (R-CHOP, R-CVP, or R- Flu–containing regimen), number of cycles received (≤4, 5–6, 7+), and completion of therapy were assessed using the Pearson chi-square test. Factors associated with risk of early discontinuation were assessed using logistic regression, and a generalized logits model was used to assess the relationship between baseline factors and number of cycles received. Kaplan-Meier methods and the log-rank test were used to estimate progression-free survival (PFS), overall survival (OS), and time to next treatment. Results: Among 1165 evaluable patients, a total of 648 received R-CHOP, 296 received R-CVP, and 221 received R-Flu combinations. Patient characteristics were similar between groups with the following exceptions: patients receiving R-CHOP were more likely to have grade 3 FL (38% vs 11% for R-CVP and 13% for R-Flu), and patients receiving R-CVP were older (25% ≥75 vs 8% for R-CHOP and 13% for R-Flu) and had higher Follicular Lymphoma International Prognostic Index (FLIPI) risk scores (50% poor FLIPI vs 41% for R-CHOP and 36% for R-Flu). The distribution of delivered cycles demonstrated a high degree of compliance with published regimens, with most patients receiving ≥5 cycles of treatment (R-CHOP 86%, R-CVP 81%, R- Flu 67%, P<.0001). The median cycle length was 21 days for both R-CHOP and R-CVP, and 29 days for the R-Flu regimens. Cycle intensity, defined as the expected cycle length (21 days for R-CHOP and R-CVP, and 28 days for R-Flu) divided by the actual cycle length, was between 0.9 and 1.1 for 68% of R-CHOP patients, 75% of R-CVP patients, and 45% of R-Flu patients. The percentage of patients with longer-than-expected time between cycles (cycle intensity <0.9) was similar for the 3 treatment groups; however, R-Flu patients were more likely to have shorter-than-expected time between cycles (24% vs 4% for R-CHOP and 3% for R-CVP). We sought to understand the reasons for patients receiving ≤4 cycles. Most patients who received ≤4 cycles were considered to have completed the intended duration of therapy (58% for R-CHOP, 62% R-CVP, and 75% R-Flu); rates of early discontinuation from the intended treatment were significantly less for patients receiving ≥5 cycles. Early discontinuation was associated with age ≥75 years and treatment with R-Flu (particularly in patients with grade 1–2 FL). Toxicity was the most commonly cited reason for early discontinuation (47% for R-CHOP, 46% for R-CVP, and 71% for R-Flu). The higher rate of R-Flu discontinuation due to toxicity is consistent with findings from other studies. Disease progression was rarely cited as the reason for early discontinuation. Time to retreatment was significantly shorter for patients receiving ≤4 cycles for patients receiving R-CHOP and R-CVP compared with patients receiving 5–6 cycles and 7+ cycles (P=.002 and P<.001 respectively; medians not yet reached). Delivery of ≤4 treatment cycles was associated with lower response rates (76% CR/PR vs 90% for 5–6 cycles and 92% for 7+ cycles) and worse PFS and OS, among patients who received R-CVP (all P<.01) but not among other regimens. Conclusions: Most patients with FL receiving chemoimmunotherapy in the NLCS completed ≥5 cycles of treatment. Most who received fewer cycles were considered to have completed therapy, and therapy commonly was given at or above the expected cycle intensity. Strategies to improve dose delivery appear unlikely to impact patient outcomes, except possibly in patients receiving R-CVP. Disclosures: Martin: Genentech: Speakers Bureau. Link:Genentech: Consultancy, Membership on an entity's Board of Directors or advisory committees. Byrtek:Genentech: Employment. Dawson:Genentech, A Member of the Roche Group: Employment. Friedberg:Genentech: Membership on an entity's Board of Directors or advisory committees. Cerhan:Genentech: Membership on an entity's Board of Directors or advisory committees. Flowers:Genentech: Membership on an entity's Board of Directors or advisory committees.

We aimed to comprehensively study sex differences in disease and patients' characteristics, treatment and outcomes in patients with follicular lymphoma (FL) in the United States (USA) utilizing the National LymphoCare Study registry (2004–2014). Among evaluable males (n = 1277) and females (n = 1375) with FL, females less commonly received anthracyclines and were more likely to receive rituximab monotherapy. Overall response rates were comparable between sex groups. With a median follow‐up of 8.1 years, male sex emerged as an adverse factor for PFS (HR, 0.84, 95% CI, 0.72–0.97). Lymphoma‐related mortality (HR, 0.46; 0.23–0.93) and overall survival (HR, 0.63; 0.41–0.97) favored females aged ≤60 years. There are subtle differences in outcomes between male and female FL patients diagnosed and treated in the contemporary era. These data represent the largest prospective analysis of FL patients in the USA based on sex and can aid design of clinical trials for this disease. Am. J. Hematol. 91:770–775, 2016. © 2016 Wiley Periodicals, Inc.

SummaryTo examine the effectiveness of an initial management strategy of watchful waiting for follicular lymphoma (FL) in clinical practice, we compared outcomes for patients diagnosed 2004–2007 in the United States initially managed with watchful waiting with outcomes following initial rituximab monotherapy and chemoimmunotherapy. In total, 1754 stage II–IV patients in the National LymphoCare Study underwent watchful waiting (n = 386), rituximab monotherapy (n = 296) or rituximab plus chemotherapy (n = 1072) as initial management strategy. Female patients and those who received treatment in the Northeast or in an academic setting more commonly underwent watchful waiting versus initial chemoimmunotherapy; whereas patients with grade 3 histology, anaemia, elevated lactate dehydrogenase, extranodal involvement, B symptoms or performance status ≥1 more commonly received chemoimmunotherapy. Although time to new treatment and progression‐free survival following first‐ and second‐line therapy were improved with chemoimmunotherapy, and time to chemotherapy was improved with rituximab monotherapy, there were no differences in overall survival between watchful waiting and chemoimmunotherapy or rituximab monotherapy. With 8‐year overall survival estimates of 74%, initial management with watchful waiting in the context of sequential therapy remains a viable option for FL patients in the modern era. This trial was registered at www.clinicaltrials.gov (NCT00097565).