- Top of page
- Patients and methods
Induction chemotherapy followed by high-dose melphalan (HDM) is the standard treatment for fitter patients with myeloma. The place of bortezomib and the thalidomide analogues within this treatment paradigm is yet to be established. We sought to identify patients who may benefit from the introduction of novel agents during their initial management. An intention-to-treat analysis was performed on 383 patients with newly diagnosed myeloma eligible for HDM to determine whether the extent of response to induction therapy and HDM correlated with long-term survival. Early response [complete response (CR) and partial response (PR)] to induction therapy was predictive of overall survival (OS) [median OS, 7·47 years for responders (CR and PR) versus 4·89 years for non-responders; P = 0·035]. The attainment of CR at 3 months post-HDM correlated with a prolonged progression-free survival (PFS) (median PFS, 7·4 years in CR group versus 5·3 years in non-CR group; P = 0·023). This data suggests that, at every stage of treatment, the aim should be to achieve CR. Patients with suboptimal responses could be offered alternative therapy. We propose a multiparametric risk-adapted model that includes response to induction chemotherapy and HDM, for identifying patients who may benefit from novel approaches to treatment.
With the advent of novel chemotherapeutic agents, such as thalidomide and its analogues (Singhal et al, 1999; Richardson et al, 2002; Anderson, 2003; Schey et al, 2004) and more recently the proteosome inhibitor bortezimib (Richardson et al, 2003), responses have been observed in myeloma patients who are refractory to standard modalities of treatment. Given these findings, the potential use of these drugs in first or second-line therapy for patients with myeloma is a natural aim in the effort to augment initial responses. However, the rationale use of these novel options requires a full understanding of the basis for current best therapeutic practice in terms of depth, rate and duration of response with respect to each treatment stage, and its impact on disease-free survival. To date, ‘standard treatment’ has involved infusional chemotherapy followed by high-dose melphalan (HDM) with stem cell support for fitter patients (Bjorkstrand et al, 1995; Fermand et al, 1995; Harousseau et al, 1995; Bensinger et al, 1996; Powles et al, 1997; Raje et al, 1997; Desikan et al, 2000; Harousseau & Attal, 2002). Evaluating the impact of this two-stage approach on clinical outcome is a crucial first step in developing the evidence base for the incorporation of new agents. Of particular importance is the significance of remission status following the induction regime and following HDM, and the long-term outlook as a function of response at both of these stages. Clinical studies thus far have conflicted in their findings (Lahuerta et al, 2000; Rajkumar et al, 2000; Alexanian et al, 2001) with complete response (CR) status post high-dose, translating into a survival advantage in some, but not all, published series. Here, we report on a large series of previously untreated myeloma patients who were treated with infusional chemotherapy followed by HDM. The impact of responses to induction chemotherapy and at 3 months following high-dose treatment with respect to long-term outlook was assessed with a view to develop a predictive model.
- Top of page
- Patients and methods
This analysis underlines the heterogeneity of myeloma in terms of biological behaviour and, with the introduction of novel treatment approaches, emphasizes the importance of identifying discriminators at an early stage to predict those who will respond poorly to standard therapy. Much in the same way that risk stratification has been usefully adopted in acute leukaemia and high grade lymphoma, profiling of patients with myeloma using some of the clinical parameters we have identified in this study, will help to stratify patients, allowing those in a poor risk group to be considered for novel therapeutic agents or more aggressive strategies, such as mini-allogeneic transplantation.
This analysis is one of the few reports of long-term outcome for myeloma patients treated with HDM, spanning just less than two decades. These results are comparable with, if not better than, outcomes in other published series with good long-term outcomes (median survivals of approximately 5 years) and a significant proportion of patients (35%) still alive at 10 years although a plateau in survival was not reached (Harousseau et al, 1995; Alegre et al, 1998). A number of studies have confirmed the superiority of HDM compared with conventional chemotherapy (Blade et al, 1996, 2000, 2001; Vesole et al, 1999) The Intergroupe Francais du Myelome 90 (Attal et al, 1996) was the first randomized trial showing better response rates, which translated into a survival benefit, and was subsequently confirmed in the larger Medical Research Council (MRC) VII study (Child et al, 2003). This current data set suggests that survival following HDM has significantly improved over the last decade, as has the safety of the approach, permitting most patients access to autologous transplantation. However, if this therapeutic approach is to be optimized, with the introduction of newer agents, it is essential to understand the effect of response on outcome of each component of the treatment. In particular, it is important to define whether the attainment of a CR is the desired aim for each stage of therapy. This can only be carried out on data sets like this, which have long follow-up and have paid close attention to the collection of response data prospectively.
Even at presentation a subset of patients can be identified who have adverse parameters, including a high B2M and low serum albumin, reflecting more aggressive disease and poorer performance status, and have a significant mortality during the induction phase. This is in keeping with the proposed prognostic factors identified in the ISS, which highlights a B2M value of >5·5 mg/l as a criterion for identifying stage III disease with poor outcome (median survival 29 months), and a serum albumin of <35 g/l as an important factor reflecting a poorer baseline status. Our analysis also identified creatinine, haemoglobin and albumin levels as powerful predictors of survival at outset. It may be argued that the patients with poor presentation prognostic variables should be offered novel therapies at diagnosis as they have a high chance of death during induction therapy, however, our data and previous data from the MRC trials (Maclennan et al, 1992) would suggest that, as many of these patients can be rescued, a change in strategy later during induction therapy but prior to HDM may be a more appropriate time point.
Analysis of the data carried out with the aim of defining the stage at which response should be assessed identified a group of patients who have not responded after three cycles of induction chemotherapy as having a worse outlook. The significance of response to induction therapy becomes more powerful after further courses of therapy have been delivered and maximum response has been reached (P = 0·0085 at maximum response versusP = 0·035 after three cycles). Thus by carefully monitoring response to therapy at an early stage, after the third cycle of chemotherapy (9 weeks of treatment), it is possible to identify a powerful discriminatory factor at an early stage of the disease process when a change in strategy is appropriate. This result also suggests that treatment should be delivered with the aim of getting maximum numbers of responses before HDM and that maximizing response at this stage is important. The addition of novel agents at this stage in the treatment regime may offer a way of salvaging NR early. Interestingly, our results are in keeping with work published by other groups and demonstrate that patients with unresponsive disease at induction can be salvaged with HDM with the achievement of high response rates (Vesole et al, 1999). However, in light of our findings, which underscore the importance of achieving a CR pre-HDM, this is probably not the most appropriate strategy to improve OS and the use of a novel agent, such as a bortezomib combination, to maximize response during the induction phase would seem logical rather than relying on HDM alone.
The response after HDM and achievement of CR at 3 months were also prognostically important, with differences in both PFS and OS for patients in CR. The impact of attaining a CR was independent of the initial response to CVAMP, further suggesting that attainment of a CR after HDM should be a major therapeutic aim. A number of other studies have demonstrated the advantage of achieving CR after high-dose therapy (Davies et al, 2001). For patients not achieving a CR at this time point, it may be argued that further treatment should be offered either as a full treatment schedule or as a maintenance strategy.
This analysis demonstrated that depth of response to induction treatment and HDM were correlated with both OS and PFS, suggesting that a two-stage response assessment will enable risk profiling to identify NR to induction therapy at an early stage and to define patients not in CR following HDM. Both of these groups may be suitable candidates for novel agents. Our data would argue that, in addition to the more traditional prognostic markers, such as B2M, albumin and haemoglobin levels, a prognostic system to identify patients that would benefit most from novel therapies should incorporate treatment response, given that many patients destined to relapse early after the first high dose cannot always be identified at presentation using the traditional markers. Although cytogenetical data was not assessed in this study, several publications have demonstrated the adverse prognostic impact of chromosome 13 abnormalities, particularly deletion of 13q14, which would be an additive factor in risk stratification (Tricot et al, 1997; Zojer et al, 2000; Facon et al, 2001; Kaufmann et al, 2003; Shaughnessy et al, 2003). A temporal profile could therefore be established, enabling subgroups of patients to be streamlined according to their perceived risk of relapse at each time point along a standard protocol. We are now using the treatment model outlined in Fig 5, which identifies significant temporal windows for the incorporation of novel agents in the treatment of new patients with myeloma. The most beneficial treatment strategy for these patients has yet to be decided; however, we have chosen to evaluate drugs with specific-targeted effects, such as bortezomib and immunomodulatory drugs either alone or in combination, with the aim of increasing CR rate at each stage and improving overall survival.
Figure 5. Proposed risk stratification model for incorporating novel therapies. MINI ALLO, mini allogeneic transplant.
Download figure to PowerPoint