Safety and activity of a new intensive short-term chemoimmunotherapy in HIV-positive patients with Burkitt lymphoma


The treatment of human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) patients with Burkitt lymphoma (HIV-BL) is a difficult challenge, requiring multidisciplinary efforts and demanding strategies. While the worldwide use of highly active antiretroviral therapy (HAART) resulted in improved tolerability and efficacy of standard chemotherapy in HIV-positive patients with diffuse large B-cell lymphomas, it has not been associated with better outcome in HIV-BL, with respect to the pre-HAART era, suggesting that more intensive treatments should be used (Lim et al, 2005). Currently, a few, small studies addressing feasibility and activity of dose-dense chemotherapeutic regimens in HIV-BL are available (Table S1). Reported regimens display high efficacy, but are often associated with important dose-limiting side-effects, prolonged hospitalization and treatment-related mortality (TRM) of 15–20%.

A dose-dense, short-term chemotherapy program including seven active drugs and intrathecal drug delivery has showed excellent activity and safety profiles in HIV-negative patients with BL in the pre-rituximab era (Di Nicola et al, 2004). We introduced a few changes to this regimen to maintain efficacy and improve tolerability in HIV-BL. In particular, six doses of rituximab were added and methotrexate dose was reduced, from 150 and 250 mg/kg to 3 g/m2, mostly to avoid mucositis, which constitutes an important route of access for infectious agents and one of the main causes of death in these patients (Galicier et al, 2007). We used this modified chemoimmunotherapy regimen in 15 consecutive HIV-BL (median age 42 years old, range 27–63) between July 2009 and July 2011 (Table S2). Treatment consisted of a 36-day induction phase including sequential doses of fractionated cyclophosphamide, high doses of methotrexate and cytarabine, doxorubicin, vincristine, and etoposide, rituximab and intrathecal prophylaxis/treatment (Table 1). Subsequent treatment was then tailored according to the objective response to induction phase (Fig. 1): patients in complete response (CR) were referred to a high-dose cytarabine-based consolidation phase (Table 1); patients in partial response (PR) were referred to consolidation followed by BEAM (carmustine, etoposide, cytarabine, melphalan) plus autologous stem cell transplant (ASCT); patients with stable or progressive disease were referred to intensification phase, followed by BEAM + ASCT. At the end of chemoimmunotherapy, patients with initial bulky disease or with a residual positron emission tomography-positive single lesion were evaluated for 36-Gy involved-field irradiation.

Figure 1.

Therapeutic programme. CR, complete remission; PR, partial response; SD, stable disease; PD, progressive disease; TD, toxic death; RT, radiotherapy; ASCT, autologous stem cell transplant. Intensification phase consisted of one or two courses of R-IVAC (rituximab, etoposide, ifosfamide, cytarabine) or R-ICE (rituximab, ifosfamide, carboplatin, etoposide) chemoimmunotherapy regimen, administered every three weeks as debulking therapy followed by high-dose cyclophosphamide (4 g/m2 day 1; rituximab 375 mg/m2 on days 3 and 10), high-dose cytarabine (2 g/m2 every 12 h for 4 d, supported by reinfusion of ≥ 2 × 106 CD34 +  cells/kg bw; rituximab 375 mg/m2 on day −1 and +11), and BEAM conditioning regimen supported by ASCT. Autologous peripheral-blood stem cells (APBSC) collection was performed with conventional modalities following high-dose cyclophosphamide (days 11–13), and, if needed, a second in-vivo purged APBSC collection was performed following high-dose cytarabine. BEAM conditioning consisted of carmustine 300 mg/m2 on day 1; etoposide 100 mg/m2 every 12 h on days 2–5; cytarabine 200 mg/m2 every 12 h on days 2–5; melphalan 140 mg/m2 on day 6. BEAM was followed by the reinfusion of a minimum of 5 × 106 CD34+ cells/kg in day 8.

Table 1. Induction and consolidation phases
DayDrug/dose/administration schedule
  1. i.v., intravenous route; i.t., intrathecal route.

  2. Dose intensity was maintained using rHuG-CSF; 5 μg/kg body weight) whenever neutrophil count was ≤ 1·5 × 109/l. Antimicrobial prophylaxis (acyclovir 400 mg twice/d, fluconazole 200 mg once a day and trimethoprim 160 mg/sulfamethoxazole 800 mg three times/week) was used. Antibiotic prophylaxis with levofloxacin 500 mg/d was indicated in patients with grade-4 neutropenia.

  3. a

    Central nervous system prophylaxis included conventional triple-drug i.t. chemotherapy delivered every 14 d (days 5, 19 and 33). In the case of meningeal involvement, i.t. therapy consisted of 6 weekly doses (days 5, 12, 19, 26, 33, and 40). Liposomal cytarabine 50 mg in alternative to conventional triple-drug scheme was allowed.

  4. b

    Intravenous alkalinization was used to promote excretion of methotrexate according to institutional guidelines. Calcium leucovorin was administered at a dose of 15 mg/m2 i.v. starting 24 h after completing methotrexate infusion, and continued every 6 h for 12 doses or, in excess, until methotrexate blood levels were less than 0·2 μmol/l. Methotrexate serum levels were monitored at 48, 72 and 96 h from methotrexate infusion and leucovorin dose was adjusted according to methotrexate serum levels.

  5. c

    After 24 h from the end of consolidation phase, recombinant human granulocyte colony-stimulating factor (rHuG-CSF) was administered twice daily to mobilize autologous peripheral-blood stem cells (APBSC).

−2, −1Methylprednisolone 0·5–1 mg/kg/d i.v.
0Methylprednisolone 0·5–1 mg/kg/d i.v.
Cyclophosphamide 500 mg/m2 over 1 h infusion
Vincristine 2 mg total dose i.v. bolus
1Methylprednisolone 0·5–1 mg/kg/d i.v.
Cyclophosphamide 500 mg/m2 over 1-h infusion
2Rituximab 375 mg/m2
5Methotrexate 12 mg + cytarabine 50 mg + steroids, i.t.a
7Methotrexate 3 g/m2 i.v. over 6 h with leucovorin rescue therapyb
14Rituximab 375 mg/m2
15Etoposide 250 mg/m2 every 12 h
19Methotrexate 12 mg + cytarabine 50 mg + steroids, i.t.a
21Methotrexate 3 g/m2 i.v. over 6 h with leucovorin rescue therapyb
29Rituximab 375 mg/m2
Doxorubicin 50 mg/m2 i.v. bolus
33Methotrexate 12 mg + cytarabine 50 mg + steroids, i.t.a
36Rituximab 375 mg/m2
Vincristine 2 mg total dose i.v. bolus
50–51Cytarabine 2 g/m2 in a 3-h infusion, twice a day (every 12 h)c
52Rituximab 375 mg/m2
60Rituximab 375 mg/m2

Treatment was safe and well tolerated. All patients but one completed the induction phase (median duration: 49 d; range 34–108); methotrexate and etoposide occasionally required delivery delay due to G3 transaminase increase or G4 neutropenia. Cytostatics dose reductions were recorded only in two patients. There was a single toxic death, which was due to pneumonia in a patient with extensive bone marrow infiltration and baseline CD4+ count of 0·017 × 109/l. As expected, haematological toxicity was common, but manageable (Table S3); with conventional antimicrobial prophylaxis and rHuG-CSF support, infective complications were mild and no systemic fungal infections occurred. The most relevant toxicity was recorded after consolidation: the original cytarabine-cisplatin consolidation regimen (Di Nicola et al, 2004) used in the first four patients was too toxic, with prolonged G4 neutropenia and severe infections in all cases. The exclusion of cisplatin, the change of cytarabine administration schedule and the addition of rituximab (Table 1) were associated with a strikingly improved tolerability in subsequent patients, with only two cases of well-controlled febrile neutropenia. Of note, there was a single case of G4 non-haematological toxicity (transient diarrhoea). Patients with hepatitis B virus or hepatitis C virus positivity completed the planned treatment (ASCT in three), and experienced only transient G3 increase of transaminase serum level, without significant chemotherapy delay. HAART was discontinued during chemoimmunotherapy in four patients: three patients exhibited an increase of plasmatic HIV-RNA levels at day 45 followed by undetectable levels at day 90 in two cases, while the third patient needed for a further HAART line; the fourth patient was the toxic death. CD4+ cell counts remained unchanged in the three assessed patients, but were < 0·05 × 109/l in two of them.

Response after induction phase was complete in six patients and partial in seven [overall response rate risk (ORR) = 87%; 95% confidence interval CI: 70–100%], one patient had meningeal dissemination and one died of sepsis (Fig. 1). Thirteen patients were referred to consolidation phase, 11 of them were referred to APBSC collection, which was successful in nine (median: 14·106 CD34+ cells/kg; range 7·20–20·02). The six patients in CR after induction received consolidation phase and remained in CR at 17–33 months [median progression-free survival (PFS): 27+ months). Five of the seven patients in PR after induction phase achieved CR after consolidation and BEAM + ASCT, the other two patients experienced progressive disease after consolidation and were referred to intensification (Fig. 1). At the end of the whole treatment, 12 patients achieved a CR (complete response ratio = 80%; 95% CI: 60–100%). At a median follow-up of 25 months (range 9–33), 11 patients remain disease-free, with a 2-year PFS for the whole series of 73 ± 28%. Eleven patients are alive, three died of lymphoma and one patient died of sepsis, with a 2-year overall survival (OS) of 73 ± 28%. Interestingly, the nine patients with baseline CD4+ cell count ≥ 0·200 × 109/l are alive and disease-free, with a median OS of 25+ months, whereas five of the six patients with < 0·200 × 109/l CD4+ cells had residual disease after induction and experienced unfavourable events, with a median OS of 9 months.

The proposed intensive, short-term chemoimmunotherapy regimen is fast, safe, cost-effective and active in HIV-BL, especially in those patients responsive to HAART and with adequate CD4+ cell counts. It showed tolerability and efficacy similar to those reported with the original regimen in HIV-negative patients with BL (Di Nicola et al, 2004). Moreover, its activity and efficacy are similar to those attained with more demanding and resource consuming regimens in HIV-BL, with an apparently better tolerability profile (Table S1). In contrast to previously reported regimens (Cortes et al, 2002; Galicier et al, 2007; Oriol et al, 2008; Montoto et al, 2010; Sparano et al, 2010; Barnes et al, 2011; Rodrigo et al, 2012), the present chemoimmunotherapy programme was delivered in a shorter period (median 90 d; range 52–143), without cases of mucositis, opportunistic infections and interruption due to toxicity, with manageable haematological toxicity, only mild infectious and a single toxic death. In line with recent studies (Dunleavy et al, 2010; Sparano et al, 2010), the addition of rituximab was not associated with increased immunosuppression. In fact, CD4+ cell counts remained unchanged in the nine patients with a follow-up longer than one year; with counts > 0·200 × 109/l in eight of them. These encouraging results will be confirmed in an ongoing multicentre prospective phase II trial ( Identifier: NCT01516593).


AJMF, MS and AR designed research, performed analyses and wrote the paper. MBV, GD, CC, LF, MF, SM, and SG performed research, treated patients and collected data. MDN, GR, UT and FCC contributed logistic resources, analytical tools and reviewed the paper.

Competing Interests

The authors have no competing interests.