**3. Conclusions**

*Multiple Myeloma*

which included only 40% of all patients. Prediction matrix models based upon those created for cardiovascular disease and rheumatoid arthritis [8, 9] which can calculate the risk of specific outcomes such as mortality allowing the differential weighting of risk factors. We can identify patients at risk with NDMM for EM to provide insight in applying different treatment approaches. Prediction tools have been applied in other hematologic malignancies to predict EM. In AML, prediction factors have improved treatment paradigms [10]. In diffuse large B-cell lymphoma, cell of origin and molecular markers along with PET scans have provided earlier

Real World patients often differ from those enrolled in clinical trials. These patients tend to be older and less fit, have more co-morbidities and less often SCT candidates [13]. An observational patient registry allows broad patient characteristics and treatment outcomes while assessing NDMM patient characteristics, biology, co-morbidities and treatments for progression-free survival (PFS) and

The Connect MM Registry reported on more than 3000 NDMM to identify and characterize EM. The first cohort included the first 1500 patients. Data was collected from an unselected patient population from routine clinical practices (81% community and 18% academic). Here a prognostic tool to assess the risk of EM based upon weighting of risk factors in elderly, SCT and non-SCT eligible patients was created

to construct an Early Mortality Prediction Matrix (EMPM). See **Figure 1.**

For the 102 NDMM patients with EM, 39.2% (2.7% of total enrolled) were due to MM progression and 32.9% were related to non-causes. Common causes of death included heart failure, pneumonia, infections, and renal failure. The other

*A color-coded guide for the clinician identifies which patient (in red) who are at highest risk for EM within six* 

*months of diagnosis compared to green and yellow patients who are lowest risk.*

treatment interventions to improve outcomes [11, 12].

**2. Early mortality in the Connect MM Registry**

overall survival (OS) [13–16].

**10**

**Figure 1.**

Prior to the era of novel agents, the incidence of EM in NDMM was 10–14% [1, 5, 17, 18]. Novel agents, supportive care, and SCT have improved PFS and OS. The promise of CAR-T therapy, monoclonal antibodies and unique agent BCMA directed against tumor necrosis super family member 17 suggest ongoing improvement for NDMM patients. Key management issues and controversies in EM patients in NDMM patients were passionately presented by Gonsalves [19]. Here the authors defined EM occurring in phase III trials and outlined key management issue strategies for NDMM to mitigate EM and summarizing those patients most at risk. The EMPM here describes parameters to identify NDMM patients at risk for EM, pitfalls in treatment and opportunities to formally address EM in clinical trials.

The prognosis of NDMM patients depends upon staging, patient features, disease biology and treatment outcomes [3]. Risk stratification utilizes the Revised-International Staging System (R-ISS) as devised by the IMWG. The R-ISS is applicable for long-term prognosis but cannot identify those at risk for EM with NDMM [20]. Issues with the R-ISS include a point-based system which is disease specific factors which cannot assess the relative individual of each factor and does not account for patient-specific risk factors. The frailty score, as in the R-ISS, is a pointbased system that combines age, functional status and co-morbiditites to predict long-term survival and treatment feasibility in elderly patients with NDMM [20]. Combining the frailty score with the R-ISS stage improves the prognostic value for each score to predict long-term survival. However, neither score alone or when combined has been used to predict NDMM patients at highest risk for EM.

Prognostic studies provide clinicians with a better understanding of the relationship in NDMM patients between the aggressiveness of disease and survival. There are significant gaps in our understanding the optimum ways to risk-stratify NDMM patients when incorporating patient and disease-specific risk factors along with combining the relative contributions of individual risk factors. Existing point-based systems make it difficult to accurately predict outcomes in patients who have a combination of standard and high-risk characteristics [20–23]. Additionally, pointbased models are primarily based upon data from interventional clinical trials that may not be representative of Real World NDMM patient populations. The EMPM model here allows differential weighting of the impact of the individual patients and disease-specific risk-factors [24–29].

Patient co-morbidities have been associated with higher mortality in various clinical trials of patients with MM [4, 30–37]. For some NDMM patients, comorbiditites are both a direct cause of death and places patients at risk for early disease-related mortality by limiting their ability to tolerate therapy [4, 31, 33, 34]. Though the decline the EM to 6.8% reflects the benefit of novel agents and supportive care there are other considerations here. NDMM patients with EM tend to be older and poorer in health with higher rates of co-morbidities (especially diabetes), greater burden of disease, and high-risk features cytogenetics and Stage III disease. The EMPM demonstrates that a lower mobility score, age > 75, history of hypertension, thrombocytopenia, higher ECOG performance status, high ISS disease stage and renal insufficiency were associated with a higher likelihood of EM. Multivariate

analysis did not independently find anemia, del(17q) mutation, low self-care score from EQ-SD, hypercalcemia, diabetes and R-ISS score, beta-2 microglobulin and albumin did not predict EM. NDMM patients with EM received more radiation therapy which delayed the initiation of therapy and limited their ability to receive triplet therapy. The EMPM has been validated by bootstrapping for internal crossvalidation [38–40]. A high degree of concordance was observed when applying the model using data from patients form the Phase 3 MM-015 trial and the phase 3 FIRST trial despite more rigorous eligibility criteria [41].

This matrix has the potential to be a clinically useful tool for NDMM patients who are at risk for EM and for analyses of specific patient populations, selection of therapy, identification of new targets for treatment and standardized comparisons between trials.

High-quality systematic research to identify patients at risk for EM have not been studied prospectively. The EMPM is the first weight-based model that accounts for both patient and disease-specific risk factors. This model can facilitate early recognition for NDMM patients at high risk for EM to assist physician selection of personalized treatments, avoidance of nephrotoxic agents, monitoring of steroid dosing in diabetic patients, prompt initiation of doublet or triplet therapy and limiting radiation fields when applicable and the use of prophylactic antibiotics. The EMPM can be applied in routine clinical practice and considered in a risk-adaptive approach. Clinical trials of reducing EM patients in NDMM can be designed for new areas of research.
