- Top of page
- Patients and methods
- Conflict of Interest
Specific trials on adult Burkitt lymphoma (BL) and ‘unclassifiable’ lymphomas with features intermediate between BL and diffuse large B-cell lymphoma (BL/DLBCL) are advocated which include substantial numbers of older patients, to improve treatment feasibility, while countering risks of systemic and central nervous system (CNS) recurrences. We prospectively evaluated a modified CODOX-M/IVAC (CODOX-M: cyclophosphamide, vincristine, doxorubicin, high-dose methotrexate; IVAC: ifosfamide, etoposide and high-dose cytarabine) regimen by the addition of rituximab (R) and liposome-encapsulated cytarabine (D) to increase antitumour activity and halve the number of intrathecal treatments. Thirty adults (40% >60 years) with BL (n = 15) and BL/DLBCL (n = 15) were accrued. Primary endpoints were progression-free survival (PFS), CNS recurrence, and liposomal cytarabine-associated toxicity. Eighty percent of patients received the whole treatment programme, the remaining cases received at least three full courses. Application of the RD-CODOX-M/IVAC regimen resulted in remarkable 4-year PFS (78%) and complete remission (CR) rates (93%). However, PFS was significantly lower in patients older than 60 years as compared to younger ones (49%vs 93%, P = 0·03; median, 36 months), despite high actual dose-intensity, CR rate and tolerability. Reduced-intensity intratechal prophylaxis through liposomal cytarabine was effective because the CNS failure rate was low (3·4%) and without severe neurological toxicities. The RD-CODOX-M/IVAC strategy is feasible and highly effective, but improving outcomes in elderly patients remains a priority.
Highly aggressive mature B-cell neoplasms in adults encompass classical Burkitt lymphoma (BL) and a number of hybrid malignancies representing a continuum between well-defined clinicopathological extremes, such as BL and diffuse large B-cell lymphoma (DLBCL). Most of these tumours, formerly defined as atypical BL (aBL) or Burkitt-like lymphoma (BLL), are now listed in the updated 2008 World Health Organization (WHO) classification as ‘lymphoma unclassifiable with features intermediate between BL and DLBCL’ (BL/DLBCL) (Jaffe et al, 2001; Leoncini et al, 2005; Swerdlow et al, 2008) and are under active review for underlying genetic and molecular heterogeneity (Salaverria & Siebert, 2011). The current management of such malignancies relies on sparse trial-based evidence, mainly due to the very low incidence of these lymphomas and to the heterogeneous categorizations through subsequent histological classification systems (Perkins & Friedberg, 2008). The outcome of these diseases has been reported to be generally poor after treatment with front-line regimens specifically designed for DLBCL (Dave et al, 2006; Nomura et al, 2008; Smeland et al, 2004), whereas chances of cure up to 70% hinge upon the use of more intense and rapidly recycling multi-agent chemotherapy, coupled with aggressive intrathecal (IT) and systemic prophylaxis for central nervous system (CNS) disease (Blum et al, 2004; Hoelzer, 2009; Kasamon & Swinnen, 2004; Nomura et al, 2008). This strategy, borrowed from paediatric literature, is aimed at countering rapid tumour regrowth/spreading between chemotherapy courses, avoiding early emergence of chemoresistant clones, and preventing the risk, which is expected to be high, for CNS (predominantly leptomeninges) involvement (12–17%) and recurrence (6–11% and 30–50%, with and without IT chemoprophylaxis, respectively) (Bernstein et al, 1986; Hill & Owen, 2006; Magrath et al, 1984; Sariban et al, 1983). Recent studies suggest that this approach can be implemented by targeting the strong expression of the B-cell lineage restricted marker, CD20, on tumour cells of BL and BL/DLBCL, and have reported promising results by the addition of the monoclonal anti-CD20 antibody rituximab to an abbreviated-intensive regimen in human immunodeficiency virus (HIV)-positive adults and patients older than 60 years of age (Oriol et al, 2008; Thomas et al, 2006). If confirmed, this would represent a major advance given that elderly patients, under-represented in most trials despite accounting for about 30% of newly diagnosed cases, have a dismal prognosis and represent an ongoing therapeutic challenge (Kasamon & Swinnen, 2004; Kelly et al, 2009; Perkins & Friedberg, 2008).
One highly effective program, CODOX-M/IVAC [CODOX-M: cyclophosphamide, vincristine, doxorubicin, high-dose methotrexate (MTX); IVAC: ifosfamide, etoposide and high-dose cytarabine], was pioneered in children and young adult patients with BL and BLL at the National Cancer Institute (Magrath et al, 1996). Modified versions were subsequently developed for older patients to ensure a safer toxicity profile while maintaining the original alternating sequence of CODOX-M and IVAC courses with fractionated schedule of alkylators, high-dose CNS-penetrating agents, and IT cytarabine/MTX injections for CNS prophylaxis (Lacasce et al, 2004; Mead et al, 2008, 2002). This latter CNS-directed programme has been regarded as a critical component of the strategy against these tumours because, when given only as high-dose intravenous (iv) load, cytarabine and MTX were unable to abrogate completely the risk of early leptomeningeal recurrences (Jabbour et al, 2005; Magrath et al, 1984). However, since cytarabine and MTX are cell-cycle S-phase-specific agents with short terminal half-lives (3·4 and 4·5 h, respectively), repeated IT courses are needed, 8–11 for CODOX-M/IVAC (Magrath et al, 1996) and 10–16 for other treatment platforms (Divine et al, 2005; Thomas et al, 1999), to warrant sustained cytotoxic concentrations in the cerebrospinal fluid (CSF) (Fleischhack et al, 2005; Shapiro et al, 1975). A suitable alternative therapy to such a high number of IT injections, which are cumbersome for patients and increase risks of procedural complications (DeAngelis, 1998; Jabbour et al, 2005), can be exemplified by liposome-encapsulated cytarabine. This agent has shown sustained cytotoxic concentrations in the CSF, up to 10–14 d following a single IT injection (Kim et al, 1993; Rueda Dominguez et al, 2005), and more activity than native cytarabine and MTX (Cole et al, 2003; Glantz et al, 1999).
This prospective study evaluated the feasibility and efficacy of CODOX-M/IVAC plus rituximab (R) and liposomal cytarabine (D) (RD-CODOX-M/IVAC), a modified version of the Magrath regimen (Magrath et al, 1996), in adult patients with BL and BL/DLBCL. It was designed to increase tolerability in older individuals, contained rituximab (R) to foster antitumour activity and included liposomal cytarabine (D) to halve the load of IT therapy.
- Top of page
- Patients and methods
- Conflict of Interest
In this prospective study, RD-CODOX-M/IVAC showed higher CR and 4-year PFS rates in BL and BL/DLBCL patients as compared with standard CODOX-M/IVAC (Magrath protocol, Magrath et al, 1996). Also, a less demanding injection schedule with liposomal cytarabine safely replaced the traditional ‘double agent’ cytarabine/MTX IT prophylaxis. Comparison with the historical group was limited by small patient numbers and by the lack of an appropriate summary score estimation to simultaneously control for multiple confounders. It is of note, however, that the two groups showed a substantial overlap due to consecutive accrual, no patients were excluded from evaluation, identical eligibility criteria, comparable distribution of the principal ‘observed’ variables, uniform modalities for diagnosis, response assessment and follow up as well as analogous supportive care.
Despite the study group having over one-third of patients older than age 60 years and median age two-fold higher than Magrath’s original study population (Magrath et al, 1996), 80% of patients completed the full four cycles of RD-CODOX-M/IVAC, without unacceptable toxicity or treatment-related deaths. These results compared favourably with the two baseline prospective, international multicentre Phase II trials, UK Lymphoma Group trial LY06 and Medical Research Council/National Cancer Research Institute trial LY10, which were conducted in a much younger patient population and wherein adapted-CODOX-M/IVAC programmes were completed by 43% and 73% of patients, respectively (Mead et al, 2008, 2002).
RD-CODOX-M/IVAC yielded a 4-year OS rate of 82%, which is superior to the survival rates (67–73% at 2 years) reported for modified-Magrath protocols, not including rituximab (Lacasce et al, 2004; Mead et al, 2008, 2002), as well as to OS rates observed for other well-known short, multiagent, cyclical intensive regimens at 2 years (70%)(Divine et al, 2005) and at 3–4 years (49–54%) (Hoelzer et al, 1996; Rizzieri et al, 2004; Thomas et al, 1999).
This is the first prospective study exploring the activity of CODOX-M/IVAC plus rituximab, but some retrospective data on this combination are available from two small and inhomogeneous series of patients, though not so large nor homogeneous, from Japan (15 patients, nine receiving rituximab: 3 BL and 6 BL/DLBCL) (Maruyama et al, 2010) and the UK (23 patients: 14 BL, 5 BL/DLBCL, 4 DLBCL) (Mohamedbhai et al, 2011). A further retrospective analysis was recently reported by US investigators (Barnes et al, 2011). The addition of rituximab to implement abbreviated intensive regimens for BL and BL/DLBCL has been explored in few other prospective trials, mostly published only in abstract form. Results from these studies are depicted in Table III: CR rates of 80–90% seem regularly attainable as well as survivals approaching 80% in the absence of late relapses beyond 2 years. RD-CODOX-M/IVAC is aligned to these performances, for example those achieved in a comparable series of 31 adult patients, one third of whom were older than 60 years, treated at M.D. Anderson Cancer Center with Rituximab plus Hyper-CVAD (cyclophosphamide, vincristine, doxorubicin and dexamethasone) alternating with methotrexate and cytarabine (Thomas et al, 2006). The proportions of CR rate and 3-years OS and DFS were comparable between RD-CODOX-M/IVAC (CR 93%, OS 82%, DFS 84%) and R-Hyper-CVAD (CR 86%, OS 89%, DFS 88%) (Thomas et al, 2006). Nevertheless, differently from R-Hyper-CVAD (Thomas et al, 2006), but similarly to the Cancer and Leukemia Group B (CALGB) 10002 study (Rizzieri et al, 2010) and to preliminary data from the Massachusetts General Hospital Cancer Center and Dana Farber Institute (Barnes et al, 2009), RD-CODOX-M/IVAC regimen did not specifically improve outcome in patients ≥60 years given that PFS and median survival overlapped with controls. These evidences are consistent with previous reports suggesting that highly aggressive mature B-cell malignancies may have a worse prognosis in the elderly despite adequate actual dose-intensity of treatment (Divine et al, 2005; Mead et al, 2008), with the majority of failures (73%) occurring due to inadequate disease control (i.e. progression within 6 months) rather than toxicity (Mead et al, 2008).
Table III. Prospective studies of Rituximab plus intensive multiagent chemotherapy in adult patients with Burkitt or Burkitt-like lymphoma/leukaemia.
|Study group (reference)||Protocol||No. Pts||Median age, years (range)||CR, %||TD, %||DFS, %||EFS, %||OS, %|
|MDACC (Thomas et al, 2006)||R-Hyper-CVAD||31||46 (27–77)||86||0||88 at 3 years||80 at 3 years||89 at 3 years|
|Age < 60 years||22|| || || ||88 at 3 years||76 at 3 years||90 at 3 years|
|Age ≥ 60 years||9|| || || ||100 at 3 years||89 at 3 years||89 at 3 years|
|PETHEMA (Oriol et al, 2008)||GMALL B-ALL/NHL 2002 (adapted)||36||36 (15–55)||86||13||90 at 2 years||–||77 at 2 years|
|HIV negative||17|| ||88|| ||93 at 2 years|| ||82 at 2 years|
|HIV positive||19|| ||84|| ||87 at 2 years|| ||73 at 2 years|
|GMALL (Hoelzer, 2008)*||GMALL B-ALL/NHL 2002||146||36||90||3||–||–|| |
|Age < 55 years||120|| || || || || ||91 at 3 years|
|Age ≥ 55 years||26|| || || || || ||84 at 3 years|
|CALGB (Rizzieri et al, 2010)*||CALGB 10002||105||−(19–79)||82||–||87||77 at 2 years||79 at 2 years|
|Age < 60 years||77|| || || || ||87 at 2 years|| |
|Age ≥ 60 years||28|| || || || ||67 at 2 years|| |
|NCI (Dunleavy et al, 2011)*||DA-EPOCH-R||29||35 (16–88)||100||0||–||97 at 4 years||100 at 4 years|
|HIV negative||19|| || || || || || |
|HIV positive||10|| || || || || || |
|Present study||RD-CODOX-M/IVAC||30||52 (25–77)||93||0||84 at 4 years||78 at 4 years||82 at 4 years|
|Age < 60 years||18|| || || || ||93 at 4 years|| |
|Age ≥ 60 years||12|| || || || ||49 at 4 years|| |
Albeit results from rituximab-containing regimens appear better than the former protocols without rituximab (Hoelzer et al, 1996; Rizzieri et al, 2004; Thomas et al, 1999), we believe that these nonrandomized and generally small-sized trials are difficult to interpret as they seem also to examine different patient populations. Some series include very large proportion of leukaemia patients (R-Hyper-CVAD) while others are devoid of patients with marrow involvement (DA-EPOCH-R), some included concomitantly HIV-infected patients (GMALL B-ALL/NHL 2002, DA-EPOCH-R) while others excluded them upfront (CALG 10002) or at analysis (R-Hyper-CVAD), some comprised only young patients (GMALL B-ALL/NHL 2002) others presented large proportions of elderly (CALGB 10002, R-Hyper-CVAD) (Table III). In such a context, evidentiary support for the addition of rituximab to current regimens does not seem plausible, as also suggested by a recent population-based study (Wasterlid et al, 2011). Determinants of improved results other than rituximab should be considered, such as enhanced supportive care and adaptation of historical treatment schedules to elderly or HIV-infected patients. Accordingly, the improved survival outcomes achieved, especially in the elderly, by R-Hyper-CVAD in comparison with Hyper-CVAD alone may have benefitted from protective environment and laminar air-flow room predisposed ‘per protocol’ for patients older than 60 years in case of grade 4 neutropenia after the first R-Hyper-CVAD course. This may have favoured R-Hyper-CVAD by zeroing the 33% infection-related induction mortality rate observed among elderly patients treated with Hyper-CVAD alone (Thomas et al, 1999, 2006). A further potential bias in the survival analysis favouring R-Hyper-CVAD may be claimed due to the high prevalence (85 vs. 45%) of patients with leukaemia (i.e., at worse prognosis) in the group treated without rituximab (Thomas et al, 2006). Likewise, in the case of RD-CODOX-M/IVAC, two factors other than rituximab may have contributed to the overall effectiveness of the strategy.
Firstly, patients with BL/DLBCL received ‘per protocol’ two additional courses of R-CODOX. This may have contributed to achieving outcome similar to BL and a PFS rate significantly higher than the historical cohort. As their disease is deemed biologically intermediate between BL and DLBCL, patients with BL/DLBCL may have benefitted from more protracted treatment and higher load of anthracycline and rituximab.
Secondly, the full planned RD-CODOX-M/IVAC programme was administered to all patients regardless of risk, whether low or high. Modified-Magrath protocols have so far assigned low-risk patients to CODOX-M courses only, and reserved the full CODOX-M/IVAC alternating sequence to high-risk patients (Lacasce et al, 2004; Mead et al, 2008, 2002), as originally designated by Magrath in a population of 21 children (median age,12 years) and 20 adults (median age, 25 years)(Magrath et al, 1996). However, while studies in children/adolescents supported that low-risk patients deserve a shorter and less intensive treatment, there is no evidence that this concept is equally valid as applied to the adult population. We designed this study to deliver the whole RD-CODOX-M/IVAC sequence regardless of risk because (i) in adults with BL and BLL the low-risk profile as defined by Magrath et al (1996) is uncommon, and no alternative clinical and/or biological prognostic system has been validated yet (Perkins & Friedberg, 2008), (ii) CNS recurrence frequently occurs in adults with early stage BL (resected stage I and abdominal stage II) (Divine et al, 2005), and (iii) chemoresistant clones can emerge early also in patients with limited stage disease and salvage strategies have previously proven ineffective (Blum et al, 2004; Hoelzer, 2009; Thomas et al, 1999).
RD-CODOX-M/IVAC yielded a low CNS failure rate (3·4%), comparable with the original Magrath regimen (11·4%) and the modified-Magrath protocols (6–11%) (Magrath et al, 1996; Mead et al, 2008, 2002), with the number of IT prophylactic administrations reduced from 8 to 4. The dual results of a less intensive but equally effective IT program may be related to the favourable pharmacokinetics of liposomal cytarabine over standard cytarabine and MTX. Liposomal cytarabine maintains cytotoxic concentrations in CSF for up to 14 d following a single IT injection (Bleyer, 1999; Glantz et al, 1999; Kim et al, 1993), compared with 24–48 h for standard cytarabine and MTX (Shapiro et al, 1975; Zimm et al, 1984). This may allow improved exposure of lymphoma cells, which have a very slow growth rate in CSF, to the cell-cycle S-phase-specific agent cytarabine gradually released from liposome particles (Kim et al, 1993). Furthermore, liposomal cytarabine distributes more homogeneously than standard MTX and cytarabine within the subarachnoid space, particularly in the cerebral ventricles, where tumour cells often accumulate (Chamberlain & Kormanik, 1996; Kim et al, 1993; Muldoon et al, 2007; Shapiro et al, 1975). Co-administering IT liposomal cytarabine with systemic high-dose cytarabine has been associated with significant neurotoxicity (encephalopathy, cauda equina syndrome, seizure and pseudotumour cerebri) in the Hyper-CVAD regimen (Jabbour et al, 2007). However, no severe neurotoxicity occurred in the present study, possibly because of (i) a reduced pre-exposure of CNS tissues to cytarabine deriving from high-dose systemic treatment (2 g/m2 vs. 3·0 g/m2 of Hyper-CVAD, (ii) a longer time interval between delivery of systemic high-dose cytarabine and IT application of liposomal cytarabine, and (iii) intralumbar and systemic administration of dexamethasone. Liposomal cytarabine-associated adverse neurological events were mild-to-moderate and interestingly affected the same group of 13 patients while all others were completely free from any toxicity. Similar results have been reported in a recent study of HIV-seropositive patients (Spina et al, 2010). It is, therefore, conceivable that neurotoxicity to liposomal cytarabine may occur in an especially predisposed subset of patients.
In conclusion, the RD-CODOX-M/IVAC strategy may represent an effective and tolerable treatment for both BL and BL/DLBCL and, within the limits of a retrospective comparison, improve Magrath protocols in adults. Further progress is needed for elderly patients, whose outcome still remains unsatisfactory despite safe and adequate treatment/dose delivery.