Could Clinical-Laboratory Features Recognize Functional High Risk Multiple Myeloma Patients? a Real-World Analysis

Background. High-risk MM patients, according to Revised International Staging System (R-ISS), have a dismal outcome in terms of PFS and OS. Nevertheless, approximately 10-20% of patients not included in the definition of high-risk have PFS shorter than 12-18 months, despite optimal initial therapy. This group of patients are termed functional (dynamic) high-risk (FHR). Really, FHR patients could be well identified by genomics or dynamic MRD assessment but these methods are not routinely applied in clinical practice due to several reasons. Therefore, a clinical-laboratory features based model could be useful to simplify their acknowledgement. Aim. The aim of this study was to describe and compare FHR MM patients features with those of patients with longer PFS and to build a simple score to assess FHR probability. Methods. FHR was defined as PFS cut-off value ≤ 18 months for transplant eligible (TE) and ≤ 12 months for not-transplant eligible (NTE) patients. Factors associated with FHR status were searched by logistic regression univariate and multivariate analysis, assigning a hazard ratio-based score for each significant factor, fitting a weight Cox model to assess FHR cumulative incidence. Survival curve was performed by Kaplan-Meier methods and compared by log-rank test. Results. In our database, 448 MM patients underwent optimal first line therapy and they were included in this study. Median age was 70 years (range 30-82), 244 (54.5%) had IgG, 113 (25%) IgA and 91 (21.5%) light chain MM. One hundred eighty patients received autologous stem cell transplantation (ASCT) and 268 standard therapy. Therapy was bortezomib-based in 134 (30%), bortezomib/thalidomide-based in 133 (30%) lenalidomide-based 97 (21.5%), carfilzomib-based in 24 (5.5%) and anti-CD38 combo-based in 60 (13%) patients. Median follow-up was 82 months (95%CI: 36-156). Totally FHR patients were 90, 15 (8.5%) in ASCT group and 75 (28%) in standard therapy one. FHR group significantly differed from not-FHR one for ISS 2-3 (89% vs 57%; p<0.001) R-ISS 2-3 (91% vs 69%; p<0.001), platelets count < 150.000/mcl (30% vs 15%; p<0.001), no maintenance/continuous therapy (94% vs 62%; p<0.001), response to therapy < VGPR (71% vs 22%; p<0.001). Logistic multivariate analysis selected ISS 2-3 (OR: 5.5; 95%CI: 2.6-12), presence of 1q abnormalities (OR: 2.9; 95%CI: 1.2-6.9), platelets count < 150.000/mcl (OR: 2.9; 95%CI: 1.4-5.8), response < VGPR (OR: 8.4; 95%CI: 4.7-15) and no maintenance/continuous therapy (OR: 9.1; 95%CI: 3.4-25) as factors significantly associated with FHR status. Scoring the patients according to the abovementioned criteria, FHR patients were 5 (5%) in the LR group (score 0-2.5 points), 30 (34%) in the IR (score 3-4.5 points) and 55 (61%) in the HR one (score 5-6). The cumulative incidence of FHR at 12 months was 7.5%, 18% and 42% and 12%, 28% and 60% at 18 months in LR, IR and HR groups, respectively. Median OS was significantly shorter in FHR vs non-FHR patients (19 vs 104 months; p< 0.001). Moreover, according to our score, probability of not achieving 3 and 5 years OS was 2%, 30%, 56% and 16%, 50%, 69% in LR, IR and HR groups, respectively. Conclusions. Our results suggest that patients potentially becoming FHR can be recognized with clinical features available before (static assessment) and after (dynamic assessment) therapy without necessarily resorting to more sophisticated methods. By these parameters it was possible to build a prognostic score useful for recognize this group of patients and for planning therapeutic intensity accordingly.