Safety and Activity of a Dose-Dense Short-Term Chemoimmunotherapy in HIV-Positive Patients with Burkitt Lymphoma (HIV-BL pts): Final Results of the Carmen Phase II Trial

Introduction : Worldwide experience with chemotherapy in HIV-BL pts is still limited. In the rituximab era, only 3 prospective trials, addressing modified GMALL, CODOX-M/IVAC or DA-EPOCH regimens, have been reported. These regimens were delivered in 126-168 days, and were associated with a 2-yr OS ~70%, but showed relevant safety problems: one third of pts did not complete treatment, often due to toxicity, and had severe mucositis, septic complications, and fungal infections. Following a retrospective experience (Ferreri et al, BJH 2012), we designed a dose-dense, short-term program including 7 active drugs and intrathecal chemo that was addressed in a multicenter phase II trial (CARMEN; NCT01516593). Herein, we report safety and activity results of this trial.

Methods : Pts with untreated BL or B-cell lymphoma unclassifiable, with features intermediate between DLBCL and BL (WHO 2008), HIV sero-positivity, age 18-60 years, and ECOG-PS ≤3 were enrolled. Pts with brain lesions were excluded, whereas pts with meningeal disease were considered. The experimental treatment consisted of a 36-day induction phase ofsequential doses of fractionated cyclophosphamide, vincristine, rituximab, high-dose (HD) methotrexate, etoposide, and doxorubicin, with intrathecal chemo. After induction, pts in complete remission (CR) received HD-cytarabine-based consolidation; pts in partial response (PR) received consolidation plus BEAM/ASCT; pts with stable or progressive disease (PD) received intensification (R-ICEx2 + HD-cyclophosphamide + HD-cytarabine + BEAM/ASCT). Leukapheresis was performed after consolidation. Pts with PET+ residual disease received 36-Gy irradiation. CR rate (CRR) after induction was the primary endpoint. The Simon two-stage minimax design was used to test the null hypothesis that the true CRR is 40% (Spina et al. Blood 2005) as opposed to the alternative hypothesis of ≥70%; estimated sample size (one side; type I error 5%; power 80%) was 19 pts. If ≥11 pts reached CR, this regimen would be declared effective in this setting.

Results : From May 2012 to Dec 2015, 20 pts were enrolled (median age 42, range 26-58; 16 males). All pts had high-risk BL: advanced disease in 20 (100%) pts, extranodal disease in 18 (90%), increased LDH serum level in 15 (75%), B symptoms in 12 (60%), meningeal disease in 5 (25%), bone marrow infiltration in 9 (45%). Eight (40%) pts had hepatitis B or C viral infection; 18 (90%) pts received HAART lines ≥2.

Eighteen pts completed the induction (median duration: 47 days; 35-90); two pts died of toxicity (bacterial infections). Drug doses were reduced in 4 pts; HAART was temporarily discontinued in 3. As expected, hematological toxicity was common: G4 neutropenia and thrombocytopenia occurred in 18 (90%) and 11 (55%) pts, respectively; G4 bacterial infections occurred in 3 (15%) pts; no systemic fungal infections were recorded. A single case of G4 non-hematological toxicity (mucositis) was reported. Tumor lysis syndrome occurred in two cases. Patients with HBV or HCV positivity did not experience G4 hepatotoxicity. Toxicity of consolidation phase was mild; G4 neutropenia (11/18) and thrombocytopenia (12/18) were manageable and no cases of G4 non-hematological toxicity were recorded. Thirteen pts were referred to APBSC collection, which was successful in 12 (median: 7.610⁶ CD34+ cells/kg; 3.4-36,5).

After induction, 11 pts achieved a CR and 7 a PR (ORR= 90%; 95%CI=77-100), no pt experienced PD. All responders received consolidation; 5 of 7 pts in PR after induction achieved CR after consolidation + ASCT and are alive and relapse-free at 21-55 months. At the end of the whole therapy, 16 pts achieved a CR (CRR= 80%; 95%CI=63-97).

At a median follow-up of 34 (20-68) months, 14 pts remain event-free, 2 died of toxicity during induction and 4 experienced PD after consolidation, with a 3-yr PFS of 70% (95%CI=52-88). Three pts with CNS disease are alive and relapse-free at 20-34 months. Fifteen pts are alive, with a 3-yr OS of 77% (95%CI=61-93).

Conclusions : The CARMEN program was safe and effective in HIV-BL pts treated in a multicenter setting. With respect to previously reported regimens, the proposed program was delivered in a shorter period, with a better tolerability profile, with a single case of mucositis and without fungal infections. Given its excellent survival effect, this strategy deserves to be further investigated, even in other lymphoma entities.