Unraveling Long-Term Survival of Multiple Myeloma Patients, the GEM2000 Experience

Introduction

Multiple myeloma (MM) has traditionally been classified as a recurrent disease, with the possibility of a cure generally not considered. This is conditioned by the fact that most of the clinical trials have limited follow-up periods (rarely exceeding 6-7 years) due to the costs of the extended monitoring. Consequently, information about long-term survival is limited. These aspects have been addressed in retrospective series focused on evaluating long-term survivors, who may represent 5-15% of MM patients. Nevertheless, there are limitations such as the heterogeneity in defining long-term survivors (survival longer than 7-10 years, depending on the study), the small sample size or the inclusion of a broad time window for patient selection, increasing variability in management. Despite the existence of these long-term survivors, prior analyses of conditional survival have not been able to demonstrate operational cure due to the ocurrence of long-term relapses and related deaths.

This study includes a large series of patients treated homogeneously with chemotherapy aiming to delve into the factors that characterize the population that achieved a refined long-term survival endpoint (≥15 years).

Methods

We analyzed 1,018 MM patients uniformly treated between 1999-2004 under the GEM2000 treatment protocol (NCT00560053), which includes chemotherapy (VBMCP/VBAD), transplant and 2 years of maintenance with interferon and prednisone. The included subjects were transplant-eligible newly diagnosed MM patients under 70 years old without prior anti-MM therapy. Clinical data and biological variables were retrospectively collected and analyzed with follow-up updated at the time of abstract submission.

Results

With a median follow-up of 180 months (95%CI 156-256), the median overall survival (OS) of the series was 72 months (95%CI: 60-72), highlighting the existence of a 10, 15, and 20-year OS probability of 30%, 20%, and 12%, respectively. Additionally, the probability that patients were alive and disease-free at 10 and 15 years was 16% and 7%.

We compared baseline features of patients defined as long-term survivors (alive at 15 years from the start of the study, regardless of whether they died or progressed thereafter) (n=106) with those who died within the first 5 years (n=481). In a multivariable logistic regression the following factors were independently associated with being a long-term survivor: age <65 years (OR 3.18; 95%CI 1.03-9.76), ECOG <2 (OR 3.45; 95%CI 1.61-7.37), IgG MM isotype (OR 2.12; 95%CI 1.01-4.74) and a MGUS-like immunophenotipic profile (OR 7.81; 95%CI 2.07-29.41) using an algorithm based on the proportion of total and pathological plasma cells by flow cytometry. The same factors showed independent association with OS in a multivariate Cox regression in addition to normal LDH levels (HR 0.69; 95% CI 0.51-0.97).

MRD negativity (10^-4 sensitivity threshold) in the post-transplant assessment proved to be a key predictive factor for both overall survival (HR 0.49; 95% CI 0.33-0.69) and long-term survival (OR 7.03; 95% CI 2.64-18.69), with a greater impact than the other factors when included in the multivariate analysis. In fact, age, MM isotype, and the MGUS profile lost their value as predictors of the likelihood of achieving >15 years of survival when MRD was included in the model.

Of the total patients who died at the last follow-up (n=808), 41% were due to MM progression (n=328) and 22% (n=176) to other MM-related causes, being infection the most relevant. The cause of death is unknown in 26% of cases (n=211) and in 11% death was not related to MM or attributable factors. In 4.7% of cases (n=38) the cause of death was a second primary neoplasm. It is notable that non-MM-related mortality represented 9% of cases in patients who died within the first 5 years compared to 36% in long-term survivors (OR 8.15; 95%CI 3.60-19.2).

Conclusion

This study based on patients treated in the early 21st-century provides one of the longest follow-up series published, allowing for a more stringent definition of long-term survivors (>15 years), for whom other causes of mortality unrelated to MM become more relevant. Identifying baseline factors characterizing long-term survivors may help us adjust treatment intensity and minimize toxicities and therapy costs. However, achieving MRD negativity remains the primary predictive factor, even despite the use of techniques with limited sensitivity.