The Discriminatory Ability of the R-ISS Is Equivalent to the ISS in a Large Cohort of Newly Diagnosed Multiple Myeloma (NDMM) Patients

Background

In the era of personalized treatment in multiple myeloma, high-risk (HR) patients must be defined accurately more than ever. The International Myeloma Working Group (IMWG) recommends to use the Revised International Staging System (R-ISS) to identify HR patients. This score combines ISS, abnormal serum LDH level and 3 high risk chromosomal abnormalities (CA): del(17p), t(4;14) and t(14;16). However, with the advent of new tools in genomics, assessing only 3 abnormalities seems to be limited. Moreover, LDH level is impacted by various medical conditions; its relevance in the score is questionable.

Aims

The main purpose of our work was to assess the R-ISS on a multi-center cohort of transplant-eligible patients (1180 patients). To our knowledge, this is the first large scale study in Europe.

Methods

Data were collected from NDMM patients enrolled in 3 clinical trials implemented by the Intergroupe Francophone du Myélome. All patients were eligible to an intensive treatment.

The overall survival (OS) and progression-free survival (PFS) curves were estimated using the Kaplan-Meier method and compared using the stratified log-rank test. The hazard ratio (HR) for progression or death were estimated by a multivariate Cox regression analysis adjusted for age, sex and therapy. Discrimination was assessed by the Harrell’s concordance index (C-index) which estimates the proportion of all pairs of patients in whom prediction and outcome are concordant and takes values from 0.5 (no discrimination) to 1.0 (perfect discrimination).

Results

Altogether, 1180 patients with MM were analysed. Median age of our cohort was 58 years. The majority of patients (78%) received an intensive treatment followed by an autologous stem-cell transplantation (ASCT). Median follow-up was 94 months for OS.

Forty-three percent of patients had ISS stage I, 39% had ISS stage II and 18% had ISS stage III. In the multivariable Cox model, the risk of death was increased for ISS stage II versus I (HR, 1.8; P < 0.001), as well for R-ISS stage III versus I (HR, 2.1; P < 0.001).

Thirty percent of patients had R-ISS stage I, 62% had R-ISS stage II and 8% had R-ISS stage III. In the multivariable Cox model, the risk of death was 1.8 times higher for R-ISS stage II versus I and 3.0 times higher for R-ISS stage III versus I.

Then we compared patients between their couple ISS/R-ISS. Thirty-one percent of the patients from ISS I were upgraded in R-ISS II; 55% of ISS III patients were reclassified in R-ISS II. Patients from the ISS I/R-ISS II couple didn’t have a higher risk of progression (HR, 1.02; P = 0.893) or death (HR, 1.36; P = 0.115) than patients in the ISS I/R-ISS I subgroup.

We also focused on the patients classified in R-ISS stage II (736 patients). We compared several subgroups: high risk CA (defined by R-ISS) versus standard risk, del(17p) vs. no del(17p), t(4;14) vs. no t(4;14) and high LDH vs. normal LDH. In the multivariable Cox model for OS, the risk of death was increased for patients with del(17p) (HR, 2.14; P < 0.001), t(4;14) (HR, 2.06; P < 0.001) and any of the high risk CA (HR, 2.15; P < 0.001). Conversely, high LDH didn’t have an impact neither on PFS (HR, 1.10; P = 0.333) nor OS (HR, 1.09; P = 0.540).

Finally, we assessed the performance of the different prognostic models for discriminating patients who progressed from those who didn’t (PFS) and patients who died from those who survived (OS) with the Harrell’s C-index. The C-index for the R-ISS was the same than the ISS one for both PFS (0.56) and OS (0.61). R-ISS didn’t give an additional prognostic value to the ISS. For every models, C-index were the same with or without LDH level; a LDH level upper than the upper limit of normal range didn’t bring predictive gain on PFS or OS. C-index was better when we assessed each criteria of the R-ISS independently; this system presents the advantage of considering different prognostic weights and also different associations.

Conclusion

Our study confirms a significant difference for both PFS and OS between R-ISS stages I, II and III, but importantly, we show that the discriminatory ability of the R-ISS, assessed by the Harrell C-index, is equivalent to the ISS. Moreover, patients in stage R-ISS II have different prognosis depending on their cytogenetic while LDH level doesn’t give any difference.

Combining ISS, high risk CA and LDH level in only 3 categories induces a loss in the prognosis assessment. A model which assesses all the parameters in an independent way would be better.