Evaluation to Prognostic Staging System of Multiple Myeloma in Novel Agent Era


 Purpose This study was to evaluate existing staging system of multiple myeloma (MM) in the real world. Methods We retrospectively analyzed 886 newly diagnosed MM from two institutions. Results The overall survival (OS) of eligible patients was 61.0 months. R-ISS held a larger receiver operating characteristic curve (ROC) area (0.603) than that of ISS (0.573) and DS staging system (0.567). In the group of immunomodulatory agents-based regimens, the median OS was 92.0 months in R-ISS I, 63.0 months in R-ISS II and 18.0 months in R-ISS III (p<0.0001). In the group of proteasome inhibitors-based regimens, the median OS was 102.0 months in R-ISS I, 63.0 months in R-ISS II and 22.0 months in R-ISS III (p<0.0001). In different subgroups grouped according to Age, HGB, CREA and Ca, R-ISS also had a good stratification effect. Patients in R-ISS II were further analyzed, which accounted for 69.9% of all R-ISS patients. Using univariable and multivariable Cox analysis, Age>65 years (p=0.001), HGB<100g/L (p<0.001), elevated LDH (p=0.001) and Ca (p=0.010) were independent factors indicating worse prognosis for R-ISS II. Conclusion R-ISS remains a valuable staging system in the real world of new drug era. But patients classified in R-ISS II still have large heterogeneity.


Introduction
Multiple myeloma (MM) was a kind of malignant disorder with abnormal proliferation of clonal plasma cells, which 7 with ROC area of ISS 0.573(95%CI, 0.517-0.629), R-ISS 0.603(95%CI, 0.549-0.657), and D-S staging system 0.567 in R-ISS II and 18.0 months in R-ISS III (p<0.001). In the group of Proteasome inhibitors-based regimens, the median OS 122 was 102.0 months in R-ISS I, 63.0 months in R-ISS II and 22.0 months in R-ISS III (p<0.001).

124
We then included factors that were statistically significant in univariate analysis into multivariate analysis for OS.
Since CA has approximately 2/3 of the censored data, it was not included in the multivariate analysis. When R-ISS was 126 included multivariate Cox analysis, Age, PLT, HGB, CREA and Ca were incorporated multivariable model. Sβ2, ALB and 127 LDH were not in the multivariable model because they were included in R-ISS. The results showed that the risk of death 128 was related to five independent prognostic variables: Age>65 years (p=0.001), HGB<100g/L(p=0.010), CREA>177umol/L 129 (p=0.012), Ca>2.5 mmol/L (p=0.007), R-ISS III vs I (p=0.001) and II (p=0.040). These results were listed in Table 3. We that neither ISS nor DS staging system was strongly predictive of outcomes, and ISS and DS staging system had low 145 concordance in stage assignment (Hari et al. 2009). It was unlikely that any one clinical staging system could fully

149
In this study, we evaluated any correlation between baseline clinical findings and the length of OS or PFS in 859 150 patients. R-ISS significantly classified OS and PFS of patients into different risk groups (p<0.001, Figure 3). R-ISS held 151 larger ROC area (AUC=0.603) than that of ISS and DS staging system in present data. Besides, in the new drug era with 152 different treatment groups, R-ISS can still stratify patients with MM well (Figure 4). R-ISS also had predictive ability on 153 OS in different subgroups based on clinical characteristics ( Figure 4). So, to sum up, R-ISS had a strong ability to judge 154 prognosis. However, most patients were classified as R-ISS II, and the survival outcome of patients in this stage was highly 155 heterogeneous in R-ISS II, which was also consistent with previous reports (Jung et al. 2018). This is to be expected, as the 156 stage consisted of patients with various characteristics, i.e. with or without CA, normal or elevated LDH, all ranges of 9 prognostic value (Kuiper et al. 2012). However, further heterogeneity may be in the population of the same CA, and the co-163 occurrence of multiple genetic lesions may have greater significance for predicting outcome than any single 164 abnormality (Hebraud et al. 2015;Kumar et al. 2012;Shah et al. 2018). Even within groups with the same genetic prognostic

167
Besides, some studies indicated that patients classified as high risk based on these FISH classifiers had actually favorable

193
There are some limitations to this research. Firstly, because it was a retrospective study, treatment regimens and cycles 194 varied widely, which lead to the results that the OS between patients in different treatment regimens had no statistical 195 difference. Secondly, due to the low acceptance of transplantation among Chinese patients, the transplantation rate of 196 patients is very low, so most of the patients in this study did not undergo hematopoietic stem cell transplantation. Thirdly, 197 cytogenetic testing has only been performed on some patients, and we were not able to perform R-ISS staging for all 198 patients. In this study, the R-ISS II accounted for approximately 2/3 of the R-ISS, which maybe was an overestimation.

199
Since only patients with defined low-risk cytogenetic changes in ISS I were classified as R-ISS I, and those with unclear 200 cytogenetic types were discarded, and those with unclear cytogenetic types were discarded when the LDH was normal in 201 ISS III, which may partially reduce the ratio of R-ISS III.

202
In conclusion, in this study, we retrospectively analyzed patients with newly diagnosed MM from two institutions in 203 China, to conclude that R-ISS remains a valuable staging system in the real world of the new drug era. But patients 204 classified in R-ISS II still have large heterogeneity. Age, HGB, elevated LDH and serum Ca were independent factors chimeric antigen receptors (CARs) targeting B-cell maturation antigen or CAR-T cells against CD19, which have shown 207 encouraging results in patients with relapsed refractory MM and hold great promise in further improving patient outcomes 208 in MM, a more refined staging system also needs to be developed in the future.

223
The data that support the findings of this study are available from the corresponding author upon reasonable request.

225
There is no conflict of interest.