Prognostic Impact on Survival of Early Relapse after Autologous Stem Cell Transplantation with Non-cryopreserved Stem Cells for Multiple Myeloma in Real Life: A Single-center Cohort Study from Oran (Algeria)

Background: The aim of this retrospective study was to analyze early relapse in multiple myeloma (MM) in real life and to evaluate its impact on overall survival (OS) and progression-free survival (PFS). Methods : Two groups of patients were identified according to the date of occurrence of relapse after autologous transplantation, within less than 24 months, defining early relapse (G1), or after more than 24 months, defining late relapse (G2). Results: A total of 307 patients with MM were enrolled, including 93 patients (30%) who had experienced relapse. There were 56 early relapses (18%) and 37 late relapses (12%). In G1 the median follow-up was 19.5 months (3-93), as compared to59 months (24-117) in G2. The median of PFS was 18 months (14.8-21.14) in G1 and was not attained in G2 (p=0.0001). The median of OS was 29 months (18.2-39.7) in G1 and was not attained in G2 (p=0.0001). In a univariate analysis, age>60 years (p=0.003), performance status>1 (p=0.036), LDH>normal (p=0.002), ISS III (p=0.0002) and an absence of maintenance therapy (p=0.002) were found to be predictive factors for early relapse. In a multivariate analysis, only a delay from the initiation of treatment to ASCT of>12 months (p=0.02) and an absence of maintenance therapy (p=0.002) were predictive of early relapse. Conclusion: The predictive factors identified here should allow us to adapt the therapeutic strategy. (OS) progression-free (PFS)

The aim of this retrospective study was to analyze early as compared to late relapse in autologous transplant with non-cryopreserved stem cells, on the basis of OS and PFS as discriminating factors. Thus, the patients were divided into two groups according to the date of occurrence of relapse following ASCT, within less than 24 months, defining early relapse (G1) or after more than 24 months, defining late relapse (G2).The comparison of the two groups G1 and G2 comprised the analysis of and search for factors predicting early relapse, such as age, sex, performance status (PS), ISS score, serum lactate dehydrogenase (LDH), the time from the initiation of treatment to transplantation and maintenance therapy. The del 17p and t (4,14) are cytogenetic abnormalities with poor prognosis and are responsible for early relapses and resistance to treatment in MM. The del 17p and the t (4, 14) have not been studied because these examinations are not carried out in all patients in Algeria. This is why they were not included in the analysis of prognostic factors for relapses. The closing date of the study was the 12/31/2019.

Statistics
PFS was calculated from ASCT until the first evidence of disease progression or the date of the last follow-up evaluation. OS was calculated from ASCT until death from any cause or the date of the last contact. The follow up after ASCT was 37 months (range, 3-117) for all patients. Univariate and multivariate analyses of predictive factors for PFS and OS (age, sex, myeloma isotype, ISS score, time from diagnosis to transplantation, maintenance post-ASCT or not) were performed using Fisher's exact test and exact logistic regression, respectively.

Results
Over a period of 10 years, 307 patients suffering from MM had been registered and were suitable for the study, including 93 patients (30%) who had experienced relapse. There were 56 early relapses (18%), representing group G1and 37 late relapses (12%), representing group G2, while 61 patients (66%) had received maintenance therapy. The clinical and biological characteristics of the two groups are presented in Table 1. In the overall population, the median follow-up was 37 months (range, 3-117), the median PFS was 59 months (range, 37-80.9) and the median OS was not attained (range, 51% at 117 months) ( Figure 1). In group G1, the median duration of follow-up was 19.5 months (range, 3-93), as compared to 59 months (range, 24-117) in group G2. The median duration of PFS was18 months (range, 14.8-21.14) in group G1 and was not reached in group G2 (p=0.0001) ( Figure 2). The median duration of OS was 29 months (range, 18.2-39.7) in group G1 and was not attained in group G2 (p=0.0001) (Figure 2). In a univariate analysis, age>60 years (p=0.003), PS>1 (p=0.036), LDH>normal (p=0.002), ISS III (p=0.0002) and an absence of maintenance therapy (p=0.002) were found to be predictive factors for early relapse (Table 2). In a multivariate analysis, only a delay between the initiation of treatment   and ASCT of>12 months (p=0.02) and an absence of maintenance therapy (p=0.002) were predictive of early relapse ( Figure 3).

Discussion
There exist few data on MM in Algeria and in particular few therapeutic data with non-cryopreserved stem cells. Our work, although retrospective, shows an overall rate of relapse following ASCT of 30% with a median follow-up of 37 months, close to that of the literature [7], in real life and in the era of novel agents, together with a rate of early relapse similar to that reported in clinical trials, 18% as compared to 16% [8]. These early relapses compared to late relapse (median durations of PFS and OS are not reached) are characterized by median durations of PFS of 18 months and of OS of 29 months, which are short and represent an unfavorable clinical prognostic factor (p=0.0001). Comparable results were obtained by Majithia et al [8], with a PFS of 8 months and an OS of 21 months (p=0.001),as likewise by Kumar   [11] reported a similar rate of early relapse (13.6%), with a short median OS after early relapse (17.8 months) as compared to late relapse (p=0.0001).
Cytogenetic factors were not analyzed in this work as such tests cannot be performed in routine practice. Thus, in a univariate analysis, PS>1 (p=0.036), LDH > normal (p=0.002), a delay from the beginning of treatment to ASCT exceeding 12 months (p=0.02) and an absence of maintenance therapy (p=0.002) were found to be factors predictive of early relapse in our study. Concerning the interval between the initiation of treatment and autologous transplantation, Jain et al [12] observed a short PFS when ASCT was carried out later as compared to within 12 months (p=0.001), but a similar OS. Meanwhile, in the literature, the factors reported to predict relapse are an ISS stage III [13], the time from the initiation of treatment to ASCT and a lack of maintenance therapy [14]. On the other hand, in our multivariate analysis, the common factor predictive of early relapse was an absence of maintenance therapy (p=0.002). In fact, maintenance therapy is today considered to be indispensable in the management of MM in patients eligible for transplantation, affording in particular with lenalidomide, a prolongation of the PFS and a reduction of the rate of relapse [15]. Overall, early relapse is a major independent clinical prognostic factor for PFS and OS [16][17].
Finally, the present work has limitations, such as the retrospective nature of the study, the heterogeneity of the induction therapy, the lack of cytogenetic studies, or the absence of evaluation of residual disease. It is nevertheless of interest in that early relapse was found to be associated with a short OS and PFS and thereby defines a category of patients at high risk, requiring a particular biological approach and specific therapeutic management.
In conclusion, in real life as in academic clinical trials, early relapse has a negative impact on PFS and OS in MM. Apart from the cytogenetic factors of poor prognosis which have not been studied in our work, we found age, performans status, LDH and ISS III as being independent factors of poor prognosis and early relapse. Performing ASCT within 12 months of the initiation of treatment and administering maintenance therapy after transplantation improve the progression-free and overall survival and are responsible for late relapses. The predictive factors identified here should allow us to adapt the therapeutic strategy.