Clinical Characteristics and Prognostic Analysis of Multiple Myeloma with Extramedullary Disease: A SEER-Based Study

Background Extramedullary disease (EMD), an infrequent manifestation of multiple myeloma (MM), can present at diagnosis or develop during the disease course. EMD can be clinically divided into bone-related EMD (EMD-B) and soft tissue-related EMD (EMD-S). The purpose of our study is to investigate the clinical characteristics, survival outcomes, and prognostic factors of MM patients with EMD. Methods A total of 155 MM patients with EMD were ultimately enrolled in our study by retrieving the Surveillance, Epidemiology, and End Results (SEER) database. The Kaplan–Meier survival curves and log-rank test for overall survival (OS) and myeloma-specific survival (MSS) were conducted to compare each potential variable. Variables with a p value <0.1 in the univariate Cox regression were incorporated into the multivariate Cox model to determine the independent prognostic factors, with a hazard ratio (HR) >1 representing adverse prognostic factors. Results The median age at diagnosis was 63 years old. EMD-B occurred in 99 patients (63.90%), while EMD-S occurred in 56 cases (36.10%). Patients with EMD-S had a significant survival disadvantage in MSS (HR = 1.844, 95% CI 1.117–3.042, p = 0.017) and OS (HR = 1.853, 95% CI 1.166–2.942, p = 0.009) compared to those with EMD-B. Patients with EMD interval ≤24 months were at higher risk of death than those with EMD at diagnosis in MSS (HR = 1.885, 95% CI 1.175–3.346, p = 0.042) and in OS (HR = 1.33, 95% CI 1.119–2.529, p = 0.036). Patients with EMD interval >24 months were at a lower risk of death as opposed to those with EMD at diagnosis. Conclusion Age at MM diagnosis, site of EMD, and time interval from diagnosis to EMD occurrence were independent prognostic factors in MM patients with EMD. EMD-B bore a better prognosis than EMD-S.


Introduction
Multiple myeloma (MM) is the proliferative disorder of neoplastic plasma cells, which is usually manifested with hypercalcemia, renal dysfunction, anemia, and bone lesions [1]. Despite wide administration of novel drugs (proteasome inhibitors, immunomodulatory agents, and monoclonal antibodies), MM remains an incurable disease with a high relapse rate and dismal prognosis [2,3]. MM cells are usually confined within the bone marrow. However, they can sometimes escape from the bone marrow microenvironment, thus migrating and infiltrating into other extramedullary organs or even circulating in the blood. [4] e kind of involvement, termed as MM with the extramedullary disease (EMD), may develop at the time of initial diagnosis, at the time of relapse or during follow-up. In this context, the definition of EMD should specifically exclude plasma cell leukemia (PCL) as well as solitary plasmacytomas (SPM), whose biological features and response to radiotherapy differ strikingly from bone marrow-derived MM [5].
ere still exists controversy regarding the precise and uniform definition of MM with EMD. Clinically, EMD can be divided into two broad categories. One is bone-related EMD (EMD-B), in which the tumor mass extends directly from osteolytic skeletons and invades into adjacent tissues. e other one is soft tissue-related EMD (EMD-S), in which tumor mass is situated in visceral organs and soft tissue which do not adjoin bones [6]. EMD-B cells still rely partially on the bone-marrow microenvironment, while EMD-S cells display plasmablastic and more naïve morphology [5,7].
Previously reported EMD incidence in untreated MM ranges from 7% to 18%, while 6%-20% MM patients suffered from EMD later during the disease course [8][9][10][11]. e objective of this study is to explore the clinical characteristics, survival outcomes, and prognostic factors by retrospectively analyzing 155 MM patients with EMD retrieved from the Surveillance, Epidemiology, and End Results (SEER) database.

Materials and Methods
2.1. Data Source. First established in 1973 by the National Cancer Institute of United States (US), the Surveillance, Epidemiology, and End Results (SEER) program has developed into a comprehensively population-based database, which covers approximately 30% of the US population across 18 cancer registries. e SEER database is updated and released annually.
e SEER database provides unidentified individual-level information and is totally available to the public via a formal application. We received permission to gain access to the patient information. SEER * Stat Software (Version 8.3.6, http://www.seer.cancer. gov/seerstat) was utilized to acquire individual patient data.
To avoid any bias that may potentially influence our study especially the survival outcome, the included patients must satisfy the following requirements: (1) MM should be the first primary malignancy with EMD as the second neoplasm, thus restricting the total number of tumors to two; (2) diagnostic confirmation should be performed by positive histology, positive histology and/or immunophenotyping, positive genetic studies; and (3) the follow-up information must be recorded completely; that is to say, patients diagnosed through autopsy should be excluded.
EMD may either present at initial diagnosis or develop during the course of MM. "Interval" in this study was defined as the time in months from MM diagnosis to the occurrence of concurrent (interval � 0) or subsequent EMD.

Statistical Analysis.
Overall survival (OS) was calculated from the time of pathologically confirmed diagnosis to death from any cause or the last follow-up. Myeloma-specific survival (MSS) was calculated from diagnosis to the date of death caused by MM or the most recent follow-up date.
e Kaplan-Meier survival curves for OS and MSS were plotted with the purpose of comparing each potential variable. e log-rank test was used to evaluate variables related to prognosis. Cox proportional hazard regression model is based on the assumption that hazard rates are proportional over time. Variables with p value <0.1 in the univariate Cox regression model were incorporated into the multivariate Cox proportional analysis to determine the independent prognostic factors associated with OS and MSS, with a hazard ratio (HR) >1 representing adverse prognostic factors. All tests were two-sided, with a confidence interval set as 95% and p value <0.05 deemed statistically significant. All statistical tests were performed by using SPSS Software (Version 26.0, IBM Corp., USA) and all figures were plotted by using Graphpad Prism (Version 8.3.0, Graphpad Software, LLC).

Results
rough retrieving the SEER database and rigorous identification, a total of 155 MM patients with EMD were ultimately enrolled in our study. e flowchart of the identification and selection process is demonstrated in Figure 1.
e demographic and clinical characteristics of MM patients are summarized in Table 1. e median age at diagnosis was 63 years. Sixty percent of the patients were above 60 years. e Caucasians accounted for 76.80%. Almost all included patients (98.06%) received chemotherapy during the first course of treatment. Of all the 155 patients, 39 cases (25.20%) had EMD at MM diagnosis (interval � 0, 0-Group), while 45 MM cases (29%) developed EMD within 24 months after MM diagnosis (interval ≤24 months, ≤24-Group) and 71 cases (45.80%) developed EMD above 24 months later after initial diagnosis (interval >24 months, >24-Group).
From the Kaplan-Meier analysis using the log-rank test, patients aged <60 years exhibited improved prognosis compared with those aged ≥60 years in MSS (p � 0.023, Figure 4(a)) and OS (p � 0.0045, Figure 5(a)). Patients who suffered from MM with EMD-B exhibited prolonged MSS (p � 0.0252, Figure 4(d)) and OS (p � 0.044, Figure 5(d)) compared with those with EMD-S. e median MSS for MM with EMD-B and EMD-S were 93 months and 61 months, respectively. e median OS for MM with EMD-B and EMD-S were 79 months and 51 months, respectively. Journal of Oncology Significant differences were found between 0-Group, ≤24-Group, and >24-Group in both MSS (p < 0.0001, Figure 4(e)) and OS (p < 0.0001, Figure 5(e)). e median MSS for 0-Group, ≤24-Group, and >24-Group were 63 months, 29 months and 105 months, respectively. e median OS for 0-Group, ≤24-Group, and >24-Group were 57 months, 26 months, and 93 months, respectively. When stratified by gender (Figures 4(b) and 5(b)) and race (Figures 4(c)and 5(c)), there seemed to be no significant difference in MSS and OS.
Covariates with a p value <0.1 in univariate Cox proportional regression (Table 3) were included in the multivariate Cox proportional regression analysis ( Table 4). As revealed in Table 4, age at diagnosis, site of EMD, and time interval from MM diagnosis to the EMD occurrence were independent prognostic factors which greatly influenced the

Discussion
EMD, an infrequent manifestation and an aggressive subentity of MM, is described as the infiltration of monoclonal MM cells into organs or tissues anatomically distant from the bone marrow [12]. Despite the extensive use of various novel agents, MM with EMD is associated with a much worse prognosis than MM without EMD. A previous study reported that MM patients with EMD-B had a significant survival advantage over those with EMD-S [13]. is is concordant with our findings. We may deduce from the results that the biological behavior and feature of EMD-B cells and EMD-S cells are completely divergent.
TP53 mutations might be correlated with the presence of EMD at diagnosis [14]. Qu et al. revealed that MM with EMD was closely linked with drug resistance even in the era of novel agents. A higher frequency of adverse cytogenetic abnormalities including del(17p13) and amplification (1q21) was observed in MM patients with EMD than those without EMD [15]. Deng et al. have also reported that Chinese MM patients with EMD showed more p53 deletion than those without EMD (34.5% vs.11.9%) [16]. e above results suggested TP53 gene aberration might play a pivotal role in extramedullary transformation. In a retrospective study of 19 MM cases who subsequently progressed with EMD, del(13q) was detected in 11 cases at diagnosis [17]. FAK (focal adhesion kinase) can inhibit MM cell apoptosis and promote their migration and invasion through interacting with phosphatase and tensin homolog (PTEN). Significant upregulation of FAK protein was detected in MM patients with EMD in comparison to those without EMD [18]. Rasmussen et al. have disclosed that RAS mutations were detected in over half of the extramedullary tumor specimens while they were not present in the corresponding bonemarrow specimens [19]. Next generation sequencing analysis revealed that MM cases with subsequent EMD bore a high frequency of RAS mutations (69%) in their bonemarrow specimens at diagnosis [20]. e above-mentioned indicated that FAK/RAS mutation might be responsible for the extramedullary spread and great clinical efficacy can be achieved in MM with EMD by targeting FAK and/or RAS signaling pathways. Disruption of crosstalk between various immune cells and bone marrow microenvironment prevents MM cells from immune surveillance, thus facilitating angiogenesis and MM cell dissemination [21]. As a rare extramedullary manifestation of MM, infiltration of the central nervous system (CNS) can be diagnosed in <1% of patients and is correlated with unfavorable high-risk cytogenetics. Acquisition of CIC (capicua transcriptional repressor) gene mutation was found in MM patients with CNS involvement. CIC gene mutation can also lead to drug resistance in BRAF-MEK inhibitor [22]. Up till now, neither standard regimen nor consensus for the treatment of MM with EMD has been established. Previous cases reported the low efficacy of thalidomide in the treatment of EMD [23,24]. EMD is more sensitive to bortezomib-based regimen than thalidomide [25]. Lenalidomide is a novel immunomodulatory drug which has achieved a remarkable clinical efficacy in MM with EMD [26]. Bortezomib-based induction therapy with subsequent autologous stem cell transplantation (ASCT) or high-dose therapy would be the optimum therapeutic regimen without increasing the EMD recurrence rate. ASCT benefited MM with EMD-S more than MM with EMD-B [27]. A recent study verified that tandem ASCT showed clinical superiority over single ASCT in the newly diagnosed MM with EMD [28]. Consolidation therapy combined with following maintenance or tandem ASCT were also essential especially for MM with EMD-S in that cases with soft tissue plasmacytoma bear bleaker prognosis as opposed to those with skeletal plasmacytoma [29].
ere exist several inevitable defects in our present research mostly due to the inherent drawbacks of the SEER database. Firstly, all enrolled patients were retrieved from the SEER database which lacks the important laboratory tests of individual patients, such as β2-microglobulin, hemoglobin, albumin, tumor size, lactate dehydrogenase, creatinine, free light chain, MM isotype, Durie-Salmon stage, and cytogenetic abnormalities. We were unable to conduct risk-stratification analysis according to the revised- international staging systems (R-ISS) and genetic risk. Secondly, information regarding disease progression, relapse, and comorbidities were not documented. Furthermore, the specific therapeutic regimens, drug dosage, radiation dose, and administration frequency were neither recorded in the SEER database. Whether MM patients received stem cell transplantation was unknown from the database. Lastly, the high heterogeneity of MM cannot be overlooked.
Despite the above-mentioned limitations, our present study, to the best of our knowledge, is the first study to explore the influence of time interval (from the initial diagnosis to EMD occurrence) on the survival outcome of MM with EMD. We demonstrated that time interval is an independent prognostic factor. In our study, EMD occurrence within 24 months after initial diagnosis bore bleaker prognosis in comparison to EMD presentation at diagnosis and beyond 24 months after diagnosis. is can be explained by the fact that EMD occurrence within 24 months after diagnosis may be attributed to rapid drug resistance and high heterogeneity among different subclones of MM cells.

Conclusions and Perspectives
In conclusion, EMD results from the migration of neoplastic plasma cells from the bone marrow microenvironment into extramedullary organs. EMD may present at diagnosis or develop during the course of the disease. e sites of predilection for EMD are the spine. EMD-B bears a better prognosis than EMD-S. Multicenter studies oriented to MM with EMD are warranted so that we can get a better understanding of the biological nature, genetic attributes,  Journal of Oncology 7 prognostic factors, and standardized therapeutic modalities. Future clinical trials should also be conducted to investigate the efficacy and safety of novel agents including Selinexor, CAR-T cells, immune checkpoint inhibitors, and other monoclonal antibodies for the treatment of MM with EMD.