CHOP-R in refractory patients

Chris was kind enough to publish a link to the ASH abstracts for this year and I have just started looking at them. The first one from the German group tickles all the boxes that I have been concerned about:

2860 Cyclophosphamide, Adriamycin, Vincristine and Prednisone Plus Rituximab (CHOP-R) in Fludarabine (F) Refractory Chronic Lymphocytic Leukemia (CLL) or CLL with Autoimmune Cytopenia (AIC) or Richter’s Transformation (RT): Final Analysis of a Phase II Study of the German CLL Study Group Petra Jenke1, Barbara Eichhorst, Raymonde Busch, Nadine Anheier, Ulrich Duehrsen, Jan Duerig, Martin H. Dreyling, Manuela Bergmann, Maria Elisabeth Goebeler, Hans Juergen Hurtz, Martina Beate Stauch, Stephan Stilgenbauer, Hartmut Doehner, Prof. Dr. med., Anna-Maria Fink, Kirsten Fischer, Clemens-Martin Wendtner, and Michael Hallek,

Introduction: In the last decade, important progress has been achieved in the treatment of CLL through the use of purine analog-based chemoimmunotherapies. Several conditions remain a challenge, often with a poor outcome. Amongst these therapeutic problems are Richter’s transformation (RT), refractoriness to F-based therapies (Fref), and the occurrence of AIC, which are sometimes induced by F. Fref and RT pts have a very poor prognosis with an estimated overall survival (OS) of only 10 and 8 months (mos), respectively. Therefore, therapeutic alternatives are urgently warranted. CHOP-R has improved the outcome of pts with aggressive non-Hodgkin’s lymphoma. To test the efficacy and tolerability of the CHOP-R regimen in CLL patients with RT, Fref, or AIC, the GCLLSG initiated a prospective phase II trial.

With the usual German efficiency, rather than rely on anecdotes this is a formal phase 2 (reasonably sized) clinical trial. Of course being a phase 2 it is primarily looking at response rate, but since we have ample historical controls over what usually happens in this group of patients, we probably won't need a phase 3 to determine outcome unless this trial should prove to be a stupendous outlier.

Material and Methods: 62 patients were included in the study. Due to protocol violations, 2 patients were excluded. Within the group of Fref pts, the medical review detected 11 patients who had received pre-treatment with F (Fpret), but were not refractory according to the updated 2008 guidelines. Thus, 26 pts were classified as Fref/pret, 19 pts as AIC and 15 pts as RT. All patients received CHOP every 3 weeks (cyclophosphamide 750mg/m², adriamycin 50mg/m² and vincristine 1,4mg/m² d1; prednisone 100mg/m² d1-5). Rituximab was added starting with the 2nd cycle (375mg/m² on each d0, and 21 days after the last CHOP-R). RT pts received up to 8, Fref/pret and AIC up to 6 courses of CHOP-R. In case of PD after 3 cycles, pts went off-study. The primary endpoints were remission rate, quality and duration of response.

Note: this was the lymphoma dose of R, not the CLL dose.

Results: 79%, 73%, and 40% of AIC, Fref/pret, and RT pts were male, respectively. The median age was 65 years (y) for Fref/pret-pts, 66y in the AIC and 69y in the RT group. Binet stages for Fref/pret pts were: A: 8%; B: 27% C: 65 %. All but 3 AIC pts were at Binet stage C. Initial RT stages according to Ann Arbor were: II: 13%, III: 13%, IV 73%. The median number of previous therapies were 3 for Fref/pret, 2 for AIC and 2 for RT.

Note: these were younger patients, but most had advanced disease and were a poor risk lot with many previous types of treatment.

A total of 314 cycles were administered, with a median number of 3 cycles for AIC and Fref pts and a median number of 4 cycles for the RT group. Due to toxicity 73% of cycles in the Fref/pret group, 66% in the AIC and 87% in the RT group were dose-reduced. 69% of Fref/pret-pts and 58% of AIC-pts received full 6 cycles of therapy and only 40% of RT-pts completed 8 cycles of therapy. Treatment was stopped in 6 pts because of PD. Due to treatment related toxicity treatment was stopped in 16 pts (27%). Treatment related mortality was 3% (2 pts). Treatment toxicity was reported according to NCI common toxicity criteria (CTC) version 2.0. Adverse events grade 3 or 4 for anemia, neutropenia and thrombocytopenia were documented in 75%, 55% and 65% of patients, respectively. Infections were the most common non-hematologic toxicity and occurred in 67%; severe infections CTC grade 3 or 4 occurred in 28%.

Treatment was not administered with the dose intensity that was planned for. This was because of toxicity - as would be expected with so many stage C patients in the mix.


All 26 Fref/pret-pts were available for response evaluation. CHOP-R achieved 54% PR, 35% SD and 12% PD. The median progression-free survival (PFS) and median treatment-free survival (TFS) were 11 and 14 mos. OS was 27 mos with a significant difference concerning F-ref (n=15) and F-pret (n=11) pts (17 vs. 35m; p=0.05). We evaluated the response of all 15 RT-pts with 60% PR, 7% CR, 13% SD and 20% PD. The PFS was 15 mos, TFS was 17 mos and OS 27 mos. 17 AIC pts were available for response evaluation with 82% PR, 6% SD and 12% PD. The PFS and TFS were only 14 and 16 mos. The OS was 50 mos.

Note: although the PR rate was respectable, it is artificially high because only some of the pret patients would have been refractory to fludarabine and the difference between overall survival for Fref and pret was more than double and statistically significant. If patients lived long enough they could benefit from other treatment including lenolidemide, alemtuzumab and transplant.

The population had a high incidence of unfavourable genetic markers: deletion of chromosome 17p [del(17p)] was detected in 24%, del(11q) in 34% and unmutated IGHV in 70%. 85% had high levels of serum thymidine kinase (sTK > 10 U/l), and 49% had high levels of ß2-microglobulin (ß2M > 3.5 mg/l). Patients with del(17p) had an unfavourable response rate and achieved significant less a PR or CR (36% vs. 76%; p=0.03). Multivariate analyses showed that del(17p) and ECOG performance status had a negative prognostic impact on OS (p = 0.0001). Moreover the presence of a del(17p) predicted a short PFS (6 vs. 16.9 mos; p=0.001).

Note: This is of course old news, but we are now discerning other factors involved in Richter Syndrome and fludarabine refractoriness.

Conclusion:

CHOP-R achieves promising response rates in CLL patients with Fref and RT and very good response rates in patients with AIC. However, the progression-free survival and overall survival remain unsatisfactory. Therefore, CHOP-R might be used as induction therapy prior to allogeneic stem cell transplantation in physically fit patients.

Or the agents that interfere with BCR stimulation, or alemtuzumab, or lenalidemide especially early in the disease. Even Bendamustine-R (or Ofatumumab) has not been evaluated in this scenario. Still CHOP-R for AIC might be promising.