Prognostic impact of second primary malignancies in multiple myeloma: An ASCO CancerLinQ discovery analysis of the modern era.

Multiple myeloma (MM) is the second most common hematologic malignancy diagnosed in adults in the United States (U.S.). Current reports suggest that approximately 5%–7% of all MM patients develop a second primary malignancy (SPM) at some point during their disease course.1 As novel MM therapies translate into improved long-term outcomes, it is conceivable that extended survival will lead to more time to develop SPMs. Although there is significant literature regarding the incidence of SPMs in MM,1 there is a relative paucity of data regarding prognosis and mortality after SPM development. Available data on mortality after SPM development are either from European countries that lack the racial-ethnic diversity of the U.S.,2, 3 are from retrospective single center studies, or do not include the modern era of anti-cancer therapeutics or International Myeloma Working Group (IMWG) imaging modalities.4 Therefore, critical prognostic information on SPMs in MM patients within the U.S. in the modern era is not available. The American Society of Clinical Oncology's (ASCO) CancerLinQ Discovery® (CLQD) MM data set was utilized to determine outcomes of real-world MM patients in the U.S. who developed SPMs in the modern treatment era. CancerLinQ® is a health technology platform developed and implemented by ASCO that provides comprehensive real-world individual patient-level data from a diverse group of practice settings within the U.S. Aggregated, de-identified patient data from the CLQD MM data set were included for patients diagnosed with MM (ICD-9 and -10 codes 203.0 and C90.0, respectively) between 2009 and 2021. SPMs were considered any malignancy diagnosed at least 30 days after MM that did not include the word “secondary” (which would indicate a metastatic lesion, not a new primary cancer), and were classified as synchronous (occurring within 6 months after MM diagnosis) or metachronous (occurring greater than 6 months after MM diagnosis). Patients were excluded if their date of MM diagnosis was missing or incalculable. ICD-9 and -10 codes were utilized to identify the top 10 reported SPMs. Primary endpoint was overall survival (OS) of patients with MM who later developed an SPM, calculated from the date of initial MM diagnosis as well as the date of SPM diagnosis to the date of death or last follow up. To minimize the risk of guarantee time bias, patients with SPMs were matched with up to three patients without the documented SPM by sex, presumed MM diagnosis date (±1 year), age (±3 years), and race (White, Black, Other), as per previously reported methods.3 In addition, each patient without a documented SPM had to be alive at the time their matched patient was diagnosed with an SPM. Analyses were conducted with STATA Version 17.0 (College Station, TX); critical alpha level was set at 0.05 (two-tailed). Kaplan–Meier curves were generated to estimate survival and compared using the log rank test. Cox proportional hazards models were constructed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). At the end of follow-up from CLQD (October 2021), the median follow-up time was 22 months (range: 0–151 months). There were 51 105 patients in the MM CLQD data set who were diagnosed with MM between 2009 and 2021. Among these patients, 8.2% (n = 4188) developed an SPM. The median latency from MM diagnosis to any SPM development was 15 months (range, 1–145 months). The top SPMs identified in decreasing order included non-melanoma skin cancer, that is, basal cell and squamous cell carcinomas (1.6%, n = 826), breast cancer (1.3%, n = 684), non-Hodgkin lymphoma (NHL, 1.3%, n = 667), prostate cancer (1%, n = 510), acute myeloid leukemia (AML, 1%, n = 485), lung cancer (0.9%, n = 468), melanoma (0.8%, n = 425), colorectal cancer (0.6%, n = 320), diffuse large B-cell lymphoma (DLBCL, 0.4%, n = 203) and acute lymphoblastic leukemia (ALL, 0.3%, n = 174). Most SPMs were metachronous (72% vs. 28%), and this trend was true for each individual SPM. Former or current smokers were nearly 50% more likely to develop an SPM than nonsmokers (RR 1.49, 95% CI 1.38–1.60, p < .001). Women were slightly less likely to develop an SPM than men (RR 0.94, 95% CI 0.88–0.998, p = .043). Patients with documented intravenous (IV) melphalan administration, which is a proxy for autologous stem cell transplant (n = 1773) had an increased risk of AML (RR 1.58, p = .023, 95% CI 1.06–2.33). Survival results are shown in Figure 1. The median OS of the entire cohort from the time of MM diagnosis was 107 months. Development of any SPM was associated with an inferior OS (HR 1.30, 95% CI 1.18–1.43, p < .001), even when separated by synchronous (HR 1.47, 95% CI 1.23–1.75, p < .001) vs. metachronous (HR 1.25, 95% CI 1.12–1.40, p < .001) SPM development. SPMs that adversely affected OS include AML (HR 2.58, 95% CI 2.0–3.32, p < .001), ALL (HR 2.31, 95% CI 1.49–3.55, p < .001), lung cancer (HR 2.27, 95% CI 1.73–2.96, p < .001), colorectal cancer (HR 2.1, 95% CI 1.48–2.99, p < .001), DLBCL (HR 1.63, 95% CI 1.04–2.57, p = .035), and melanoma (HR 1.41, 95% CI 1.07–1.85, p = .015). When separated by synchronous versus metachronous, an SPM of synchronous lung cancer was associated with the worst survival (HR 3.43, 95% CI 2.12–5.05, p < .001), followed by synchronous AML (HR 2.43, 95% CI 1.37–4.31, p = .002), metachronous ALL (HR 2.26, 95% CI 1.41–3.63, p = .001), metachronous AML (HR 2.24, 95% CI 1.69–2.96, p < .001), metachronous colorectal cancer (HR 2.00, 95% CI 1.37–2.92, p < .001), and metachronous lung cancer (HR 1.75, 95% CI 1.27–2.41, p = .001). Synchronous ALL and colorectal cancer did not affect OS, likely because of the small number of cases which limited statistical power. SPMs of NHL, breast, non-melanoma cutaneous, and prostate cancers did not affect OS, regardless of whether they were synchronous or metachronous. No statistically significant differences were found in patients who developed SPMs as a function of whether they had received IV melphalan or not (median OS 124 vs. 131 months, HR 0.87, p = .488, 95% CI 0.58–1.29); however, this could have been due to the small number of patients available for this analysis and the sparse data available on patients receiving IV melphalan. With respect to race, Black patients had a lower risk of developing any SPM than white patients (RR 0.81, 95% CI 0.74–0.87, p < .001), exhibiting a lower risk of non-melanoma skin cancer (RR 0.08, 95% CI 0.05–0.13, p < .001), melanoma (RR 0.46, 95% CI 0.34–0.63, p < .001), and NHL (RR 0.76, 95% CI 0.61–0.94, p = .011). However, Black patients had a higher risk of breast cancer (RR 1.50, 95% CI 1.25–1.79, p < .001) and prostate cancer (RR 1.39, 95% CI 1.12–1.71, p = .003). No significant difference in risk of ALL, AML, colorectal, DLBCL or lung cancer were noted by race. There was no statistically significant difference in OS after SPM development in Black versus White patients (median OS 121 vs. 128 months, HR 0.95, 95% CI 0.77–1.16, p = .597). This study utilized a large data set of real-world MM patients treated in diverse settings across the U.S. to characterize the prognostic impact of SPM development in the modern era. Historical studies assessing mortality after SPM development in patients with MM have been carried out in European countries that lack the racial and ethnic diversity of the U.S., in single institutions, or have not focused on the modern era.2, 3 Current IMWG imaging guidelines recommend use of PET scans, whole body low-dose CT scans, and whole-body MRI scans,4 which could theoretically detect SPMs at earlier stages. This might be the case for SPMs of NHL, breast and prostate cancers which did not adversely impact OS. In addition, it is possible that SPM detection at early disease stages could obviate the need for tumor-directed systemic therapy and subsequent interruption of anti-myeloma therapy. Last, there have been a multitude of U.S. Food and Drug Administration approvals of novel agents and immunotherapies not only for MM but across the full spectrum of malignancies, which could translate to better long-term outcomes. Despite improvements in treatment and outcomes of MM and other tumor types, SPMs continue to carry an adverse prognosis, especially AML, ALL, lung, and colorectal cancers. Because of large amounts of missing data on treatments, it was difficult to characterize SPMs according to therapies received. While retrospective designs have their limitations, some observations in this study are on par with what is reported in the literature, serving as a good internal control, and validating the accuracy of the CLQD MM data. For example, the incidence of SPMs found in this data set is consistent with prior reports.1 In addition, the median OS of the entire cohort was 107 months, which is comparable to reports of long-term survival among newly diagnosed MM patients receiving frontline lenalidomide, bortezomib, and dexamethasone induction.5 Lastly, the findings that black MM patients had a lower risk of non-melanoma skin cancer, melanoma and NHL and a higher risk of prostate cancer are not unexpected.6 A better understanding of cause of death after SPM development in the context of SPM stage and MM depth of response could help clinicians better understand competing risks of death between MM and SPMs, thereby guiding difficult treatment decisions. Finally, it is possible that a larger patient population would have the power to show that other SPMs and patient characteristics such as race also significantly affect OS. SG performed protocol development, result interpretation, and manuscript preparation; JT performed statistical analysis, result interpretations and manuscript preparation; KP performed protocol development, result interpretation, and manuscript preparation; JC interpreted results and performed manuscript preparation; KS performed statistical and bioinformatics analyses, results interpretation and manuscript preparation; CBD performed protocol development, results interpretation, and manuscript preparation. None of the authors have any conflicts of interest to disclose. The data that support the findings of this study are available athttps://cancerlinq.org/solutions/researchers.