Abstract
The management of patients with chronic lymphocytic leukemia (CLL) has radically improved over the last few years with the addition of anti-CD20 monoclonal antibodies (MoAbs) to chemotherapy. Chlorambucil has been considered for decades as a suitable therapeutic option for frail patients. Taking into account the advantage offered by the addition of MoAbs to chemotherapy, different studies up to now have explored the feasibility of chlorambucil-based chemoimmunotherapies in treatment-naïve CLL. COMPLEMENT1 is a prospective, randomized, open-label trial evaluating the efficacy and safety of ofatumumab added to chlorambucil, compared with chlorambucil in monotherapy, in the setting of untreated patients with CLL considered unsuitable for a fludarabine-based approach. Progression-free survival was significantly longer in the chemoimmunotherapy arm when compared with the single-agent chlorambucil (22.4 months versus 13.1 months). Response rate and quality were also improved in the combination arm. Furthermore, the addition of ofatumumab did not lead to an unmanageable toxicity. While the employment of anti-CD20 antibodies represents an advantage in the treatment of the CLL symptomatic population, at present different patient selection and treatment schedules do not allow a reliable comparison between chlorambucil-based regimens. The addition of ofatumumab to chlorambucil represents a further therapeutic gain in CLL. Longer follow up and direct comparison with other MoAbs are warranted to establish the preferred first-line treatment in elderly and unfit patients.
Introduction
Chronic lymphocytic leukemia (CLL) is the most common type of leukemia in Western countries with a median age at diagnosis of 72 years [National Cancer Institute, 2010].
Initial therapy for patients with symptomatic disease was historically based on the use of purine analogs and/or alkylating agents [Dighiero, 1997].
The addition of monoclonal antibodies (MoAbs) to chemotherapeutic regimens radically improved treatment outcome in terms of response achievement and survival, conferring substantial clinical benefits.
Notably, while the combination of fludarabine, cyclophosphamide, and rituximab (FCR) represents the gold standard for the fit and young CLL population, the majority of patients, due to age or comorbidity, cannot tolerate a fludarabine-based regimen [Hallek et al. 2010].
Chlorambucil, with its oral bioavailability, low cost and toxic profile, has been considered for decades a reasonable treatment option for frail patients; nevertheless, low response rate and short progression-free survival (PFS), represent the main limitation of this drug [Dighiero et al. 1998; Rai et al. 2000].
Taking into account the favorable chlorambucil toxic profile, its combination with MoAbs could represent a suitable option for less fit patients.
Rituximab is a chimeric type I anti-CD20 MoAb with a cytotoxic action due to complement-dependent cytotoxicity (CDC), antibody-dependent cell cytotoxicity (ADCC), and direct induction of apoptosis [Weiner, 2010].
Treatment with a combination of chlorambucil and rituximab in patients with previously untreated CLL has been explored in clinical trials with the achievement of responses ranging from 66% to 84% [Hillmen et al. 2014; Foà et al. 2014; Goede et al. 2014].
Type I antibodies, inducing the translocation and stabilization of CD20 into lipid rafts, cause strong CDC but little direct cell death. In contrast, type II antibodies do not stabilize CD20 in lipid rafts and consequently show a reduced binding to C1q, resulting in a flat level of CDC [Cragg et al. 2003].
Obinutuzumab is a glycoengineered type II anti-CD20 MoAb that recognizes a CD20 epitope overlapping with that of rituximab, but compared with type I anti-CD20 antibodies it exhibits a different elbow hinge angle and binds CD20 in a different orientation. Regarding its activity, it is primarily due to a direct cellular cytotoxicity rather than a complement-mediated activity [Owen and Stewart, 2012].
Ofatumumab is a fully human type I anti-CD20 MoAb. Its ability to bind to both the small and large loop of the membrane antigen CD20 allows a prolonged dissociation rate. Compared with rituximab, ofatumumab is able to produce greater CDC activity with a similar ADCC activity, especially in low CD20-expressing CLL cells [Teeling et al. 2004; Bologna et al. 2013; Okroj et al. 2013]. Furthermore, complement cascade seems to be easily activated by ofatumumab due to its binding avidity to C1q. Ofatumumab has shown to be effective as a single agent when administered in a population of high-risk patients with refractory CLL [Pawliczkowycz et al. 2009].
The COMPLEMENT1 trial
COMPLEMENT1 is a prospective, randomized, open-label trial evaluating the efficacy and safety of ofatumumab added to chlorambucil, compared with chlorambucil in monotherapy [Hillmen et al. 2015].
As previously reported [Catovsky et al. 2007; Hillmen et al. 2014], authors adopted for this study a chlorambucil dosage of 10 mg/m2 for 7 days, while ofatumumab was administered monthly at 1000 mg, with a split dosage during the first cycle (300 mg day 1; 1000 mg day 8).
A total number of 447 patients were randomly assigned to receive either ofatumumab plus chlorambucil or chlorambucil alone (221 and 226 cases, respectively).
The primary endpoint of the study was PFS; secondary included overall survival (OS), time to progression, overall response rate (ORR), complete remission rate, time to next treatment (TTNT), and safety evaluation.
Even if not restricted to those with elevated Cumulative Illness Rating Scale (CIRS) score, all patients included in this trial were considered unsuitable for a purine analogs-based treatment. At the investigators’ judgment, age was the main criteria for protocol inclusion. Median age was 69 years, with half of patients being older than 70 years; median CIRS score was 9. Clinical and disease characteristics were well balanced between the two groups.
Patients could receive the assigned treatment up to 12 cycles or until best response; in the great majority 6 or more cycles were administered, two thirds of patients interrupted treatment after 6 cycles for clinical judgment of satisfactory response.
Overall and complete responses resulted significantly in favor of the ofatumumab–chlorambucil arm compared with chlorambucil only (82% versus 69% and 13% versus 1%, respectively), in this case also irrespective of clinical and biologic parameters. Of the 212 cases tested for minimal residual disease (MRD) in the combination arm, 17 achieved MRD negativity, with only 1 in the chlorambucil group.
PFS results significantly improved with the addition of ofatumumab compared with chemotherapy alone (22.4 months versus 13.1 months,p < 0.0001), thus meeting the primary endpoint of the study. The PFS advantage of immunochemotherapy was evident in all prognostic subgroups, stratified for clinical and biologic parameters, and did not differ in the older population. In particular, patients with positive zeta-chain-associated protein kinase 70, unmutated immunoglobulin heavy (IgHV) status, β2-microglobulin > 3500 mg/ml, 11q deletion, showed an advantage with ofatumumab. Furthermore, a trend of improvement in patients with 17p was noted. In line with the PFS, TTNT was significantly longer for the combination arm (median 39.8 months versus 24.7 months,p < 0.0001).
After a median follow up of 28.9 months, OS survival was not reached in both groups.
Treatment was well tolerated overall. Despite a higher number of grade 3 extra-hematologic adverse events reported in the combination arm, this did not lead to an increased discontinuation rate in patients treated with ofatumumab, even in the older population. Grade 3–4 infusion reactions were the most relevant among the extra-hematologic adverse events, being reported in 10% of patients regardless of age. All-grade and grade 3–4 infection rate did not differ between the ofatumumab–chlorambucil arm (42% and 9%, respectively) and the chlorambucil arm (46% and 12%, respectively); also incidence of deaths during treatment or within 60 days from the last drug administration,was 3% in both groups.
Regarding hematologic toxicity, all-grade anemia and thrombocytopenia were more common with chlorambucil (13% versus 9% and 26 versus 14%, respectively); neutropenia was more often reported with ofatumumab–chlorambucil (27% versus 18% for any grade), thus not leading, as previously mentioned, to a difference in infection rate between the two arms. All those data did not differ even when stratifying the two populations according to age.
Overview on COMPLEMENT1 trial data and chlorambucil-based chemoimmunotherapies
Despite the fact that most patients with CLL are old and present comorbidity, very few trials seem to address this ‘real life’ population.
In several clinical studies, when comparing chlorambucil with more intensive agents, despite the achievement of higher overall and complete response rate and a prolongation of PFS, no improvement of OS could be demonstrated. Furthermore, updated results from the CLL5 trial showed a prolonged survival in patients treated with chlorambucil compared with those who received fludarabine (68.1 months versus 49.8 months, respectively) [Eichhorst et al. 2009]. Similar data were reported with the long-term follow up from the CLL4 trial comparing chlorambucil with fludarabine and FC [Else et al. 2015]. After 10 years observation of the survivors, OS did not differ among patients treated with chlorambucil compared with FC or the fludarabine arm; this could be explained taking into account the fact that a significantly better complete response rate was obtained following first-line chlorambucil or fludarabine progression, when compared with those who received FC as first line.
As previously mentioned, physically fit and young patients with CLL are likely to benefit from the standard treatment with the FCR platform. Conversely, FCR very often led to treatment-related toxicities causing delay or therapy discontinuation in elderly patients, who represent the majority in CLL, thus precluding the clinical benefits [Hallek et al. 2010].
In this setting, the addition of MoAbs to chlorambucil has been explored as a possible compromise between effectiveness and tolerability.
Two studies so far have focused on the rituximab–chlorambucil combination [Hillmen et al. 2014; Foà et al. 2014]. In both trials, chemoimmunotherapy compared favourably with historical data of single agent chlorambucil as ORR resulted >80% in both studies with a rate of 10% CR in the study from Hillmen et al. [Hillmen et al. 2014] and 18.9% with 2% MRD negativity in the one from Foà et al. [Foà et al. 2014]. PFS also significantly improved with the addition of rituximab, 23.5 months [Hillmen et al. 2014] and 34.7 months [Foà et al. 2014] compared with less than 20 months in previous studies with chlorambucil [Catovsky et al. 2007; Hillmen et al. 2007; Knauf et al. 2009].
In preclinical studies and in the experience of non-Hodgkin lymphomas, obinutuzumab showed superiority when compared with rituximab [Mössner et al. 2010; Sehn et al. 2015; Patz et al. 2011]; this led to its use in combination with chlorambucil in frail patients in the phase III CLL11 trial, which compared obinutuzumb–chlorambucil versus rituximab–chlorambucil versus single-agent chlorambucil [Goede et al. 2014].
Rituximab as well as obinutuzumab showed superiority in terms of response achievement and response quality, in respect to chlorambucil alone.
Direct comparison between rituximab and obinutuzumab, associated with the same chlorambucil schedule, demonstrated the superiority of the latter in terms of ORR (66.1% versus 77.7%, respectively, p < 0.001); complete responses (7% versus 20.7%, respectively), and MRD negativity (2.6% versus 19.5%, respectively). Furthermore, these results translated to a PFS benefit of obinutuzumab as well as rituximab added to chlorambucil compared with chlorambucil alone (median PFS 26.7 months and 15.2 months versus 11.1 months, respectively). Finally, after about 3 years follow up, obinutuzumab, but not rituximab, combined with chlorambucil, led to a significant survival advantage in respect to single-agent chlorambucil. A survival prolongation, albeit not significant, was obtained with obinutuzumb–chlorambucil compared with rituximab–chlorambucil. Despite adverse events, in particular neutropenia, occurring more frequently in the MoAbs arm, rate of grade ⩾ 3 infections did not differ among combination and chlorambucil treatment groups.
Taking into account these results [Foà et al. 2014; Goede et al. 2014; Hillmen et al. 2014, 2015], the addition of any anti-CD20 MoAbs to chlorambucil treatment appears clearly as a successful option for elderly or unfit patients with CLL. Furthermore, especially when considering an older population, it seems advisable to administer the most effective regimen upfront as patients treated with chlorambucil alone in the CLL11 trial, despite a similar ORR, obtained a significantly shorter PFS, even following the crossover, than those who received a MoAb-based regimen from the start. This represents once again, a confirmation of chemoimmunotherapy superiority in respect to chemotherapy alone, even in the setting of low-expressing CD20 cells in CLL. In such a population in fact, combination therapy obtains better results than chlorambucil alone in terms of response rate and PFS, without adding unmanageable toxicity.
What is still not defined is the best MoAb to choose and its correct dosage as well as the preferred chlorambucil schedule.
Even if different chlorambucil administrations and, probably, patient selection criteria used in trials, do not allow a direct comparison of the results obtained with MoAbs added to chemotherapy, some consideration could be extrapolated from the published data.
In UK studies with rituximab or ofatumumab plus chlorambucil, chemotherapy was administered with the intensive dosage of 10 mg/m2 for 7 days [Hillmen et al. 2014, 2015]. A similar dose was employed in an Italian study with rituximab–chlorambucil [Foà et al. 2014], while a significantly lower chlorambucil dosage (0.5 mg/kg on day 1 and day 15) has been reported by a German group [Goede et al. 2014], and translated, for the rituximab–chlorambucil arm, to a poorer performance when compared with previous studies with the same combination.
In addition to the different chlorambucil dose intensity, it is important to note that chemotherapy duration, in terms of number of cycles, varied among first-line chlorambucil-based chemoimmunotherapy trials ranging from 6 cycles in the German schedule, 8 in the Italian one, and up to 12 in both UK studies employing rituximab and ofatumumab.
Nevertheless, when comparing ofatumumab and rituximab combined in the same chlorambucil schedule [Hillmen et al. 2014, 2015], the response rate, as well as PFS, did not seem to differ despite the use of a more active anti-CD20 MoAb. Moreover, even with the administration of a lower chlorambucil dose, the obinutuzumab combination appears to be slightly superior. It is noteworthy that obinutuzumab was prescribed during the first cycle with a higher dose intensity compared with ofatumumab or rituximab and this could explain its better performance [Goede et al. 2014]. This is particularly important in the comparison with ofatumumab. In fact, while rituximab was also prescribed with the increased dose of 500 mg/m2 from the second cycle in all the three schedules employing it [Foà et al. 2014; Goede et al. 2014; Hillmen et al. 2014], in the COMPLEMENT1 trial, following the 300 mg test dose, there was essentially one 1000 mg full dose ofatumumab. Obinutuzumab dose-response activity has been explored in a phase II study addressed to symptomatic, untreated CLL [Byrd et al. 2015].Two different MoAb dosages (1000 mg and 2000 mg, with a powered dose during the first cycles in both schedules) were administered as monotherapy. Obinutuzumab was demonstrated to be effective when administered with the 1000 mg schedule; nevertheless, ORR and complete response rate (69% and 20%, respectively) resulted in favor of the higher dose, thus suggesting a dose–response relation. Similarly, the first clinical study evaluating ofatumumab efficacy and tolerability was performed by Coiffier and colleagues who reported the experience with 33 patients in the Hx-CD20-402 dose-escalation study [Coiffier et al. 2008]. Three cohorts of patients experienced different ofatumumab dosages (maximum dose 500 mg in cohort A, 1000 mg in cohort B, and 2000 mg in cohort C). Patients received four weekly infusions with a starting dose of 100 mg, 300 mg, and 500 mg, respectively. This trial was addressed to a heavily pretreated population with a median number of three prior therapies.
Ofatumumab used as a single agent was effective in 13/14 responding patients belonging to cohort C. Objective responses were rapid in a large number of patients on the 2000 mg schedule, with lymph node size halving from week 4 and then gradually progressing. An improvement in baseline cytopenia was also observed. On the basis of these results, the 2000 mg dosage was chosen in subsequent studies with monotherapy ofatumumab. Some trials are currently exploring the efficacy and safety of dose-dense ofatumumab in association with chemotherapy, in particular the use of a double dose compared with that previously experienced in combination with chemotherapy in treatment-naïve fit patients with CLL [Mauro et al. 2015; Wierda et al. 2011].
Conversely, higher rituximab dosage combined with chemotherapy did not lead to an advantage [O’Brien et al. 2005].
With a median follow up of 29 months, PFS prolongation of the ofatumumab–chlorambucil regimen did not translate in a survival advantage, while patients treated with obinutuzumb–chlorambucil had improved OS when compared with chlorambucil alone. However, it is important to highlight that, while 3 years OS was similar among the two MoAb combination groups (about 80%), survival with chlorambucil as monotherapy, significantly differs in COMPLEMENT1 and CLL11 trials (at 3 years, 83% and less than 70%, respectively). A longer follow up is probably needed to better clarify if a survival advantage in ofatumumab–chlorambucil combination will emerge, particularly in those patients who obtained an MRD negative status following treatment. Furthermore, recent updated results from the CLL11 trial [Goede et al. 2015], after 39 months observation, failed to demonstrated a significant OS benefit of the obinutuzumb–chlorambucil combination compared with the rituximab-based one. Nevertheless, a trend for obinutuzumb–chlorambucil survival advantage continues to be present in all the subgroups analyzed except for those with a creatinine clearance lower than 50 ml/min.
Table 1 summarizes schedules and results of the chlorambucil-based chemoimmunotherapies described above in previously untreated patients with CLL.
Chlorambucil and monoclonal antibody combinations in previously untreated chronic lymphocytic leukemia.
Chl, chlorambucil; nr, not reached; Ob, obinutuzumab; Ofa, ofatumumab; R, rituximab.
Regarding toxicity, up to now, ofatumumab has demonstrated a lower severe infusion reaction rate when compared with obinutuzumab as grade 3–4 reactions occurred in 20% of patients treated with the latter and 10% of those randomized in the chemoimmunotherapy group in COMPLEMENT1. In both trials, baseline lymphocyte count did not influence the incidence or severity of infusion reactions, which were anyway limited to the first or second infusion and did not lead to treatment discontinuation in any patient. Nevertheless, in the CLL11 trial, even the enforcement of additional prophylactic measures showed only a limited benefit on infusion-related adverse events.
A ‘head-to-head’ comparison between ofatumumab and obinutuzumab would be desirable to better define the most effective first-line approach using CD20 MoAbs for elderly and/or unfit patients.
Moreover, economic analysis performed with either ofatumumab or obinutuzumab combinations showed them to be cost-effective when taking into account response rate and survival benefit translating into an improved long-term health outcome confirming the superiority of chemoimmunotherapy [Herring et al. 2015; Yagudina et al. 2015].
Bendamustine-based combinations also appear to be promising and well tolerated in elderly patients [Fischer et al. 2012; Cuneo et al. 2014; Laurenti et al. 2015]. In the CLL10 trial, the bendamustine–rituximab regimen did not differ from FCR in terms of PFS in the older population [Eichhorst et al. 2014]. Furthermore, the direct comparison between chlorambucil and bendamustine, demonstrated a higher efficacy of the latter with a comparable tolerability [Knauf et al. 2012], and the phase II study OMB115991, in which bendamustine was administered with ofatumumab in a previously untreated CLL population, confirmed the high efficacy and safety of this combination [Offner et al. 2014].
Pentostatin could also represent a suitable option for patients older than 65 years. In a population with a median age of 72 years, this purine analog, which has previously demonstrated its efficacy and tolerability [Kay et al. 2007; Shanafelt et al. 2013], brought an elevated overall and complete response rate, including 19% of MR-negative complete responses, with a manageable toxicity when combined with ofatumumab and cyclophosphamide [Tedeschi et al. 2015]
Finally, the role of tyrosine kinases inhibitors, such as ibrutinib, idelalisib, and more recently, venetoclax, must be defined in first-line treatment of the elderly, in particular in 17p-deleted patients who gained marginal or no advantage from the administration of ofatumumab or obinutuzumab.
In 64 previously untreated older patients with CLL, the administration of eight weekly rituximab infusions in association with idelalisib, a PI3Kd inhibitor, led to a 97% ORR with 19% complete responses. All del(17p)/TP53-mutated patients obtained a response as well as almost all the unmutated IgHV. Furthermore PFS was 83% at 36 months [O’Brien et al. 2015].
While idelalisib and rituximab resulted in a successful combination, some concerns exist for the association of ibrutinib with MoAbs.
Ibrutinib is an irreversible inhibitor of Bruton’s tyrosine kinase. Different studies have demonstrated a negative impact of ibrutinib association with type I and II MoAbs [Khort et al. 2014; Da Roit et al. 2015]. In vitro, authors deny the possibility of a synergic action of ibrutinib on direct cell death and CDC induced by anti-CD20 MoAbs and furthermore, demonstrated in vitro and in vivo, a strong ability to inhibit natural killer (NK)-cell activation and ADCC. Idelalisib has been tested as well and showed a similar but weaker inhibition on CDC, NK-cell degranulation, ADCC, polymorphonuclear-cell activation, and macrophage phagocytosis.
These findings however, need to be confirmed clinically as a phase Ib/II study with ibrutinib and ofatumumab in previously treated patients showed a significant benefit with combination treatments in terms of response rate and survival [Jaglowski et al. 2015]. Moreover, other recent preclinical trials seem to refute the findings of ibrutinib ADCC inhibition that seem to be limited to only an initial or minimal activity [Ysebaert et al. 2014; Davis et al. 2014].
Finally, venetoclax, a Bcl2 inhibitor, when evaluated in combination with rituximab in patients with relapsed/refractory CLL, achieved a high 88% response rate including, differently from B-cell receptor antagonists, even MRD negative complete responses [Roberts et al. 2014].
The association between second-generation MoAbs and targeted therapies is currently under investigation in clinical trials and could be extremely promising in future.
Conclusion
Recent data on clinical trials combining chlorambucil and anti-CD20 MoAbs confirms that this approach is effective and well tolerated in less fit patients with CLL. In particular, the addition of ofatumumab to the chlorambucil backbone definitely represents an advantage in the landscape of chemoimmunotherapy in elderly/unfit patients with CLL. Direct comparison between second- generation antibodies is warranted in order to define the best chemoimmunotherapy approach as first line for this population.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
MM received a grant from GSK for participation on the advisory board, as an invited speaker, and for research support, and from Novartis as an invited speaker. AMF, AT, PP, MM, RC report no conflicts of interest.