In this study, we evaluated the prognostic value of BI score in patients with newly diagnosed MM. Low BIs were recorded in 36.8% patients and we found that the low BI was an independent poor prognostic factor for OS and PFS in patients with newly diagnosed MM. The propensity score matching analysis also showed that patients with low BI had shorter OS and PFS. However, patients with low BI had similar response rate compared with patients high BI.
Multiple myeloma is a hematological malignancy characterized by abnormal proliferation of neoplastic plasma cells in bone marrow, which may lead to systemic bone destruction and various clinical manifestations. In the past 20 years, with the development of ASCT, proteasome inhibitors, immunomodulatories and immunotherapy, the survival of MM patients has been significantly prolonged [10–12]. However, due to the survival heterogeneity of MM patients, some patients can survive for more than 10 years, while the median survival time of some patients is only 1–2 years. Therefore, accurate evaluation of patients' prognosis is an important issue in the diagnosis and treatment of MM.
At present, the common clinical prognosis assessment system is mainly based on tumor factors and host factors. The clinical characteristics, physical status, tumor load and other indicators of MM patients play important impacts on the prognosis [13–16]. The age is an important prognostic factor for MM patients, studies showed that patients aged < 65 years had longer survival time than patients aged ≥ 65 years. Therefore, IMWG proposed the GA scoring system, which was based on the daily basic activity scale, instrumental activities of daily living ability assessment scale and Charlson comorbidity index scale, and gave comprehensive scores (0 ~ 5 points) according to patients' age, cognition, comorbidity and physical condition. According to the score, 869 newly diagnosed MM patients were divided into fit group (0 points), intermediate fitness group (1 point) and frail group (≥ 2 points). The 3-year OS rates of the 3 groups were 84%, 76% and 57%, respectively [5]. In addition, GA score system could also indicate the incidence of treatment-related toxic reactions, so it could provide certain reference information for the choice of treatment regimen for newly diagnosed MM patients [13, 17–19].
Since the GA score incorporates age into the system, patients over 75 were at most assigned to the moderate healthy group, which will limit treatment options for older patients with better physical performance. Murillo et al. [6] compared GA score with Fried score regardless of age in 98 elderly patients newly diagnosed with MM (median age 79 years) and found that frail patients screened by the latter had a greater decrease in 1-year total OS rate and a higher risk of death. It indicates that the model has better prognostic value and may be used to assess the physical status of MM patients.
It is also controversial that the GA score assumes that patient self-assessment is more reliable than physician assessment. Facon et al. [7] analyzed 1623 newly diagnosed MM patients and developed a scoring system that combined physician evaluation with patient self-evaluation. This system included age, Charlson Comorbidity Index (CCI), and Eastern Cooperative Oncology Group performance status (ECOG PS) and patients were divided into frail and nonfrail groups. Overall response rates were 72% and 79%, respectively, and PFS were 19.4 months and 24.0 months, respectively. Compared with GA score, Facon model had more significant prognostic evaluation advantages, and its method of dividing MM patients into two groups was more convenient for clinical practice than that of dividing MM patients into three groups. This model was also used to assess the general condition of MM patients to develop appropriate treatment [20–23].
Basic activities of daily life have a great impact on the prognosis of MM patients, which can be used as a prognostic factor of MM patients. The inability of patients to perform basic activities without assistance is an important indicator to evaluate host factors. Barthel index (BI) is currently the most widely used tool to evaluate basic activities of daily life, but few studies have evaluated its prognostic value in MM. In our study, OS and PFS of patients with low BI was significantly shorter than that of other patients. Univariate and multivariate Cox proportional hazards regression analyses also suggested that low BI was a poor prognostic factor for newly diagnosed MM. Moreover, we used propensity-score matched analysis to balance covariate distributions between patients with low and high BI. It showed that the distribution of main factors which could affect outcome of MM patient had no significant difference between the two matched groups. Patients with low BI had significant shorter OS compared with other patients after balancing main factors. PFS was also shorter, though it had no statistical significance. As a result, low BI may be considered an independent poor prognostic factor for patients with newly diagnosed MM.
This study has several limitations. This study was a retrospective analysis with data from a single MM diagnosis and treatment center and the results need to be confirmed by more research centers. The new treatment significantly improved the survival of patients with MM, and the median follow-up time in this study was short, requiring long-term follow-up to verify the results. After the initial treatment in our center, some patients were transferred to other centers for further treatment, leading to the loss of follow-up of some patients, which may affect the results of the study.
In conclusion, our study demonstrated that low BI was an independent poor prognostic factor for patients with newly diagnosed multiple myeloma. It needs further study to confirm its prognostic value of BI for newly diagnosed MM.