The Revised Myeloma Comorbidity Index (R-MCI) As a Promising Approach for Predicting Overall (OS)- and Progression-Free (PFS) Survival and Optimizing Therapy Strategies in Multiple Myeloma (MM) Patients (pts) - Comparative Analysis of 5 Comorbidity Indices (CI), Including Retro- and Prospective Applicability

Introduction: MM is considered an incurable disease. Due to the improvement in long-term survival of MM pts with innovative therapy strategies, the necessity of including comorbidities and biological fitness status to treatment options has significantly increased over the last years. However, elderly pts still remain to benefit to lesser extends revealing lower PFS and OS rates, show more treatment-induced side effects and a lower quality of life. Therefore, it is inevitable to include an objective frailty assessment into MM guidelines to avoid under- and overtreatment. Since no CI has been widely established yet, selection process should include the capability of assessing a retro- and prospective baseline for comparing CI results among available data sets. This analysis compared 5 internationally well-discussed CIs regarding OS and PFS prediction and tested if these CIs can be reliably assessed retro- and prospectively.

Methods: This prospective study was performed for 347 consecutive pts treated at our center, analyzing OS and PFS for the R-MCI, IMWG-, CCI, Mayo- and MRP-scores: the factors that are included in each individual score are shown in Table 1. For each CI, pts were divided into 3 risk groups (low-, intermediate-, high-risk), except for the CCI with possible designation into low- and high-risk group only. Based on these risk groups, OS and PFS were estimated by Kaplan Meier Method and compared via log rank test. Additionally we compared this above mentioned prospective cohort with a prior study including 749 pts treated at our center (Haematologica 2017;102:910-21). For this cohort we now performed a retrospective analysis with 4 of the latter CIs to analyze differences in risk group distribution within these two cohorts via Chi Square tests. Since there was missing laboratory data for the UK-MRP-score in the retrospective cohort, it was excluded from the latter analysis.

Results: Pts' characteristics were typical for tertiary centers with a median age of 65 years (yrs). Median Follow-up was 36 months, median OS was not reached and PFS was 34 months. All 5 CIs could divide pts into risk groups with significantly different OS (p<0.05). The difference in 3yr OS for high- and low-risk group using the R-MCI was 43% and via IMWG- and Mayo-score 37% and 70% respectively. Minor distinction for OS prediction was achieved by CCI and MRP with only 25% and 20% difference. For the MRP, the 3-yr-OS rate for frail (59%) exceeded that of intermediate pts (50%), moreover, this group comprised only low numbers (n=8). In analogy, the difference in 3yr PFS amounted to 48% for the R-MCI vs. 40% and 59% for IMWG- and Mayo-scores, respectively. Here again the CCI and MPR showed lowest PFS differences with only 20% and 27% between high- and low-risk groups (Table 1). Comparing the pro- and retrospective analyses via Chi-Square tests, only the R-MCI showed comparable results for the risk group distribution in both cohorts (p=0.26) with 26% vs. 27%, 60% vs. 55% and 14% vs. 18% in low-risk, intermediate-risk and high-risk groups, respectively (Table 1). Respective results for IMWG and CCI scores showed that significantly more patients were defined as low-risk in the retrospective than in prospective cohorts with 41% vs. 30% for the IMWG and 65% vs. 47% for the CCI, respectively (p<0.001). The results for the Mayo-score revealed that 13% of the retrospective cohort were classified as high-risk as compared to 8% of the prospective cohort (p=0.0209).

Conclusions: To our knowledge this is the first large prospective comparative analysis of 5 internationally discussed CIs. Our results show an excellent separation into risk groups with different OS and PFS via R-MCI, IMWG and Mayo-scores, whereas via CCI and Mayo-score to a much lesser extent. The results of this analysis reinforce the multifunctionality and convenience of using one MM-CI, like the robustly tested and repeatedly validated R-MCI. Further unique features of the R-MCI are the pro- and retrospective applicability in daily clinics, a user-friendly homepage and the future perspective of extended use, e.g. for tailoring therapies, which is currently investigated and which results will be shown at the meeting.

View largeDownload slide

View largeDownload slide

Close modal