Durability Results of Non-Myeloablative (NMA) Allogeneic Stem Cell Transplantation (alloSCT) for Relapsed Follicular Lymphoma: 17- Year Experience

Purpose: We have previously reported the outcomes of NMA alloSCT in 47 patients with relapsed/chemosensitive follicular lymphoma (FL) who received a matched sibling donor (MSD) after FCR conditioning (Khouri, Blood 2008;111:5530). In subsequent trials, eligibility was expanded to include transplants from matched unrelated donors (MUDs) using a ⁹⁰yttrium ibritumomab tiuxetan (⁹⁰YIT)-based regimen (Khouri, Blood 2012 ;119:6373) or, more recently, BFR (bendamustine, fludarabine, rituximab; Khouri, Blood 20014;124:2306) conditioning. Herein we report on long-term outcomes in 98 FL patients treated on these 3 consecutive trials between 1999-2017. Methods: The BFR regimen (N=20) consisted of bendamustine 130 mg/m² IV daily on days -5 to - 3 prior to transplantation, thus substituting for the cyclophosphamide in the FCR regimen (N=47) . The dose and schedule of fludarabine (30 mg/m² IV daily x3) and rituximab (375 mg/m² IV on day -13 and 1000 mg/m² on days -6, +1, and +8) were similar in both regimens. ⁹⁰YIT-regimens (N=31) : A diagnostic dose of ¹¹¹In-ibritumomab was administered on day-14, followed by a fixed dose of 0.4 mCi/kg ⁹⁰YIT on day -7; FC or BF chemo was then administered at the same dose and schedule (days -5 to -3) as described above. Tacrolimus and methotrexate was used for GVHD prophylaxis. In addition, thymoglobulin of 1 mg/kg was given on days -2, -1 in patients receiving a MUD transplant. Results: Median age was 53 years (range, 29-71) and 24 patients (24%) were >60 years. Twenty-eight patients (29%) had a HCT-CI of ≥ 3. The median prior number of treatments was 3 (range, 2-9); 51 patients (52%) had rituximab-chemo induction at diagnosis. Seventy-one patients (72%) relapsed within 2 years of their induction treatment and median duration of last remission prior to alloSCT was < 1 year in 61% of patients. At transplant, 82 patients (84%) had chemosensitive disease (46% CR, 38% PR); 16% had refractory disease; 22% were PET+; and 20% had elevated serum LDH. Seventy patients (71%) had a transplant from a MSD and 28 (29%) from MUDs; 15 (15%) transplants had female-to-male donors, and 42 (43%) were ABO-mismatched. In addition, CMV was reactive in 80% of patients/ and or donors. Most patients (94%) had alloSCT from peripheral blood. Median number of CD34-positive cells infused was 4.9 x 10⁶/kg. A significant difference in ANC recovery between the 3 conditioning regimens was observed. Neutrophil counts recovered to > 0.5 x 10⁹/L a median of 0 days (range 0-16) for the BFR groups vs. 10 days (range, 0-17) and 11 days (range, 5-17) for the FCR and ⁹⁰YIT-regimen groups, respectively (p<0.0001). This difference was consistent for each transplant type. With a median follow-up time for all patients of 98 months (range, 3-208 months), the overall survival (OS) and progression-free-survival (PFS) rates at 98 months were 82% (95%CI, 73-89) and 74% (95% CI, 64-82), respectively (Figure 1) . The cumulative incidence (CI) of grade II-IV and III-IV acute GVHD was 22% and 9%, respectively. The CI of chronic GVHD was 38%. Treatment-related mortality at 1 year was 9%. Patient demographic, disease characteristics, and transplant characteristics and their correlation with outcomes were evaluated. Disease status of > 1^st relapse, >2 prior chemotherapies, duration of last remission prior to alloSCT < 1 year, ≥ 3 comorbidities, elevated LDH, acute II-IV GVHD, chronic extensive GVHD were associated with inferior OS. By multivariable analysis (MVA), duration of last remission prior to alloSCT (<1year) [HR 6.48 (1.28, 32.69); p=0.024]) and acute II-IV GVHD [HR 8.61 (2.99, 24.83); p<0.001] were associated with inferior OS. No significant prognostic factors on MVA for PFS, or risk for acute or chronic GVHD were noted for this cohort of patients. Conclusions: Nonmyeloablative allogeneic transplant can induce complete responses lasting over a decade in most patients with relapsed FL. Our initial findings published in 2008 were thus confirmed in a larger number of patients including those who received MUD transplants. BFR conditioning has been associated with significantly lesser myelosuppression and a faster neutrophil recovery than other regimens used, validating our initial observation in earlier reports.

View largeDownload slide

View largeDownload slide

Close modal