Secondary Primary Malignancies in Multiple Myeloma: A Review

Abstract

As survival times of multiple myeloma (MM) patients continue to improve, second primary malignancies (SPM) have become an increasingly relevant long-term risk among MM survivors. Population studies since the 1960s have consistently observed an increased incidence of hematologic SPMs, specifically acute leukemia, among MM survivors. Prolonged treatment with alkylators, especially melphalan, was associated with an increased hematologic SPM risk, likewise autologous stem cell transplantation appeared to minimally increase SPM risk. Immunomodulatory drugs, specifically lenalidomide, was associated with an increased SPM incidence, although most studies concluded that the benefits of therapy outweighed any risk of SPMs. Newer anti-myeloma therapy such as proteasome inhibitors and monoclonal antibodies did not appear to increase SPM risk although robust long-term follow-up is lacking. This review discusses current understanding regarding SPM among survivors of MM, how different host-, disease- and treatment-related factors contribute to SPM incidence and highlights emerging screening guidelines and prognosis for SPMs.

Introduction

Multiple myeloma (MM) is a malignant neoplasm characterized by the proliferation of neoplastic plasma cells and monoclonal immunoglobulin production. It is the second most common hematologic malignancy, constituting 1-2% of neoplasms worldwide and 2% of all cancer deaths [1]. In the United States alone in 2016, 131,262 patients were living with multiple myeloma [2]. Over the last two decades, median survival times have improved, evolving into a chronic, incurable condition with a prognosis of 8 to 10 years [3, 4]. Prolonged survival times have largely been due to the introduction of novel treatments such as autologous stem cell transplantation (ASCT) in the early 1990s, immunomodulatory drugs in the late 1990s, proteasome inhibitors in the early 2000s, along with second and third generation immunomodulatory drugs and proteasome inhibitors, and most recently, monoclonal antibodies in 2015.

The association between MM and secondary primary malignancies (SPM) as a long-term risk has long been recognized. Given the improved life expectancy which translates into a growing prevalence of MM patients, SPMs are increasingly relevant considerations for both patients and clinicians. In this paper, we review and discuss our current understanding of SPMs after MM and how underlying biology and different MM treatment modalities affect SPMs incidence. We also discuss emerging screening guidelines and prognosis of SPMs.

SPM in MM: population-based trends

Population studies conducted without accounting for treatment-related effects reported a decreased or no increased risk of solid SPM among MM patients, but noted an increased incidence of hematologic malignancies (Table 1) [5–11]. For example, an analysis of the Swedish Family Cancer Database between 1958 and 1996 noted an excess of hematologic SPM in MM patients compared to the general population with a standardized incidence ratio (SIR) of 2.19 (95% CI: 1.74-2.71), indicating that the observed rate of hematologic malignancy development was 2.19 times that of the general population [5]. Specifically, an excess of acute myeloid leukemia (AML), with a SIR of 8.19 (95% CI: 5.70-11.74), and non-Hodgkin lymphoma (NHL) with a SIR of 1.74 (95% CI: 1.12-2.57) was observed. Interestingly, a decreased incidence of solid tumors was also observed with a SIR of 0.81 (95% CI: 0.73-0.90). Similarly, another population study of the Surveillance, Epidemiology and End Results (SEER) database analyzing United States patients diagnosed with MM between 1973 and 2008 noted a decreased risk of solid tumor SPM with a SIR of 0.94 (95% CI: 0.89-0.99) and an increased risk of hematologic malignancies with a SIR of 1.68 (95% CI: 1.46-1.92) [8].

Table 1:

Population-based registries evaluating the incidence of secondary primary malignancies in multiple myeloma

Author	Registry	Study period	Patients (n)	All SPM n (%)	Hematologic SPM n (%)	Solid tumor SPM n (%)	All SPM SIR (95% CI)	Hematologic SPM SIR (95% CI)	Solid tumor SPM SIR (95% CI)	Latency (median)
Dong et al. [5]	Swedish Cancer Registry	1958-1996	8656	475 (5.5)	83 (1.0)	392 (4.5)	NR	All 2.19 (1.74-2.71); AML 8.19 (5.70-11.4); NHL 1.74 (1.12-2.57)	All 0.81 (0.70-0.90)	2.9 years
Mailankody et al. [6]	Swedish Cancer Registry	1986-2005	8740	577 (6.6)	69 (0.8)	508 (5.8)	1.26 (1.16-1.36)	All 2.04 (1.59-2.58); AML/MDS 11.51 (8.19-15.74)	All 1.19 (1.09-1.30); GI 1.30 (1.09-1.53); nonmelanoma skin 2.22 (1.74-2.80)	45.3 months (MDS/AML)
Youlden et al. [7]	Queensland Cancer Registry	1982-2001	2174	134 (0.6)	NR	NR	Males 1.04 (0.84-1.27); Females 0.89 (0.64-1.21)	NR	NR	NR
Chakraborty et al. [8]	SEER	1973-2008	3245	1657 (51.1)	214 (6.6)	1394 (43.0)	0.99 (0.95-1.04)	All 1.68 (1.46-1.92); Leukemia 3.07 (2.57-3.64)	All 0.94 (0.89-0.99); Invasive skin 1.43 (1.09-1.85)	NR
Razavi et al. [10]	SEER	1973-2008	36,491	2012 (5.5)	263 (0.7)	1707 (4.7)	0.98 (0.94-1.02)	All 1.63 (1.45-1.84); leukemia 2.94 (2.52-3.43); AML 6.51 (5.42-7.83); NHL 1.28 (1.04-1.57); CLL 0.34 (0.17-0.68)	All 0.92 (0.88-0.97); esophagus 0.49 (0.28-0.87); lung/bronchus 0.88 (0.78-0.99); breast 0.81 (0.69-0.94); melanoma 1.36 (1.07-1.74); bladder 1.22 (1.03-1.44); kidney 1.30 (1.01-1.66); thyroid 1.63 (1.05-2.52)	5.2 years
Tzeng et al. [9]	National Health Research Institutes (Taiwan)	1997-2009	3970	71 (1.8)	35 (0.9)	36 (0.9)	NR	All 13.0 (7.79-21.6); AML 23.9 (10.5-54.5); NHL 7.72 (3.83-15.6)	All 0.57 (0.40-0.79)	1.9 years
Ailawadhi et al. [11]	SEER	1973-2008	NR	2021	NR	NR	Hispanic Whites 0.67 (0.50-0.88); non-Hispanic whites 0.97 (0.92-1.02); African American 1.07 (0.97-1.18); Asians/Pacific Islanders 1.11 (0.87-1.40)	AML: Hispanic Whites 1.54 (0.04-8.58); non-Hispanic Whites 6.85 (5.55-8.38); African Americans 6.24 (3.41-10.47); Asians/Pacific Islanders 6.32 (1.72-16.19)	All: Hispanic Whites 0.66 (0.48-0.89); non-Hispanic Whites: 0.90 (0.85-0.95); African Americans 1.05 (0.94-1.17); Asians/Pacific Islanders 1.03 (0.79-1.33)	4.7 years

NR-not reported, SPM-secondary primary malignancies, SIR-standardized incidence ratio, CI-confidence interval, SEER-Surveillance Epidemiology and End Results

In general, the risk of developing hematologic SPMs rose with longer duration of follow-up time after MM diagnosis; risk appeared to start 12 months after MM diagnosis and increased with time, with highest rates usually seen at 5-10 years after diagnosis [12]. For instance, a recent population study utilizing the SEER 13 registries to analyze three cohorts of MM patients diagnosed during 1995-1999 (pre-thalidomide, limited use of ASCT), 2000-2004 (post-thalidomide, pre-lenalidomide and bortezomib, increased utilization of ASCT) and 2005-2009 (post-lenalidomide and bortezomib, highest utilization of ASCT) showed the 7.5-year cumulative incidences of SPM to be 4.7%, 6.0% and 6.3% respectively [12]. During the first 5 years of MM diagnosis, the risk of SPM was 23% lower than expected in the 1995-99 cohort (SIR 0.77; 95% CI: 0.59-0.99), while no different compared with the background population in the 2005-09 cohort (SIR 1.15; 95% CI: 0.47-2.78). The observed risk discrepancy between the two cohorts is due to an increased incidence of hematologic SPMs, primarily leukemias, from SIR of 1.28 (95% CI: 0.47-2.78) to SIR of 2.17 (95% CI: 1.27-3.48), respectively. During years 6-15 of MM diagnosis, an increased risk of hematologic SPM, specifically of leukemias was noted in both 1995-99 and 2000-04 cohorts with SIRs of 8.76 (95% CI: 3.21-19.06) and 7.30 (95% CI: 2.94-15.05), respectively. Also noted was an increased incidence of lymphoma from 1995-1999 to 2000-2004 with SIRs of 0.59 (95% CI: 0.01-3.29) to 3.31 (95% CI: 1.51-6.27), respectively.

Above studies provide insights into risk of SPM development at the population level, but due to differences in reporting and methodology, direct comparisons are not possible. SIRs do not account for the competing risk, and thus may be overestimating the risk of SPM development, especially when looking at long latency periods between MM and SPM diagnosis as the risk of death due to MM or its other complications rise over time. None of these studies account for treatment-related factors. Period analysis has been used to account for this, relying on the approval of various treatments to infer exposure to them, though misclassification may be present. Despite its limitations, above population studies are likely generalizable to MM patients. While the risk of solid tumor SPMs has varied across studies, reporting both decreased and increased risk, they have consistently reported an increased risk of hematologic SPMs that increases over time, necessitating both patients and clinicians be aware of this risk during follow up in the event that new symptoms arise that are inconsistent with either treatment related toxicity or progressive MM.

SPM in MM: inherent risk factors

Multiple studies have evaluated different risk factors which affect SPM development after a diagnosis of MM. SPM risk is multi-factorial in etiology, likely due to a combination of intrinsic and extrinsic risk factors. Intrinsic risk factors include host-related factors such as sex, age, race/ethnicity, comorbidities, genetic predispositions and disease related-factors such as MM disease characteristics [13]. Extrinsic risk factors include treatment regimen and duration and lifestyle factors known to increase cancer risk such as smoking, sun exposure, obesity etc. [14].

Advanced age has been variously reported as both a risk and a protective factor against SPM development, depending on study methodology. In observational studies, Chakraborty et al. reported advanced age to be a risk factor [8], while Razavi reported advanced age to be protective [10]. Retrospective studies also show older age as adverse risk factors for SPM development [15, 16]. In the Myeloma XI trial, age was a risk factor for SPM development with the highest SPM incidence of 17.3% observed in transplant non-eligible patients aged >74 years receiving lenalidomide maintenance compared to 9.7% in those aged <74 years [17].

Studies have also suggested male sex to be associated with increased SPM risk [8, 9, 18, 19]. Race/ethnicity has also been proposed as a risk factor for SPM development. Two separate analyses of the SEER registry have suggested that African Americans have a higher risk and Hispanics have a lower risk of SPM development compared to Non Hispanic Whites [8, 11]. Specific SPM disease trends also differed among different race/ethnicities.

A Swedish population study of 55652 patients with MM reported that patients with monoclonal gammopathy of undetermined significance (MGUS), the precursor to MM, was associated with an 8-fold increased risk of myelodysplastic syndrome (MDS)/AML SPM development while patients with MM had an 11.51-fold (95% CI: 8.19-15.74) increased risk of MDS/AML development compared to the Swedish population [6]. The increased risk was confined to the IgG and IgA isotype MGUS patients. Furthermore, MGUS patients with M protein concentrations more than 1.5 g/dL had higher risk of MDS/AML development compared to those with less than or equal of 1.5 g/dL, suggesting that molecular heterogeneities present in both MGUS and MM, along with a higher proliferative rate or volume of disease may play a role in SPM development. Another analysis of the Connect MM, a US-based multicenter, prospective observational cohort registry study of patients treated for MM, noted a higher proportion of patients with SPM had a reported a history of MGUS [20].

Heterogeneity in MM cytogenetics have also been associated with SPM development. A single center retrospective study reported 77% of MM patients who developed SPM had complex/high-risk cytogenetics [21]. In contrast, long term follow-up of 744 patients with MM noted predominantly favorable cytogenetics in patients acquiring SPMs suggesting that long disease latency due to indolent myeloma may have allowed for manifestation of additional malignancies [22].

A study of 47 newly diagnosed symptomatic and smoldering MM patients used multicolor flow cytometry to evaluate immunophenotypic alterations of bone marrow stem cells to ascertain a baseline risk for MDS and AML [23]. Interestingly, 13% of symptomatic MM patients had multiple phenotypic abnormalities typically seen in MDS. Single immunophenotypic abnormalities were seen in both symptomatic and smoldering MM patients (47% and 33% respectively), implying that baseline aberrancies in the stem cell population contribute to MM’s predisposition to hematological SPMs. Importantly, the smoldering myeloma patients were enrolled on the QUIREDEX trial of lenalidomide and dexamethasone for 8 cycles – and no changes in immunophenotypic patterns were seen after treatment. Another study examining serial bone marrow specimens from 2418 MM patients who had undergone melphalan-based ASCT identified 105 (4.3%) specimens that had cytogenetically defined myelodysplasia [24]. Out of the 105 patients, 21 developed overt clinical MDS and 5 developed AML. While clinically largely silent in terms of MDS or AML manifestation, the presence of cytogenetically defined myelodysplasia was associated with decreased overall survival.

Germinal polymorphisms present across different malignancies may provide a biological predisposition for SPM in MM. For example, the MTHFR gene polymorphism is associated with an increased risk of MM in one study and a decreased risk of prostate cancer in another [25, 26]. This is further supported by an analysis of the Swedish Cancer Registry which showed that patients with a prior malignancy diagnosis had a significantly increased risk of developing a subsequent SPM compared with MM patients without [27]. In addition, more than 1 prior malignancy correlated with even worse survival outcomes. Whole genome studies may be needed to better understand the role of genetics in SPM development and the complex interactions between predisposition to SPMs and treatment.

SPM in MM: association with alkylators

The effect of treatment-related factors, including alkylators such as melphalan and cyclophosphamide, on the development of SPMs has been extensively investigated and studies consistently show that prolonged alkylator treatment further increases the risk of hematologic malignancies, specifically acute leukemia [28, 29]. Melphalan appears to be more leukemogenic than other alkylators such as cyclophosphamide. In addition, the association between alkylator therapy and SPM development was suggested to be time dependent; with increased risk of MDS and acute leukemia seen with more prolonged melphalan use. Prolonged melphalan treatment did not appear to increase the risk of other hematologic or solid malignancies.

In 1970, Kyle et al. suggested a possible etiologic role of melphalan therapy in the development of acute leukemia after reporting a case series of 4 MM patients who developed subsequent rapidly progressive acute leukemia after receiving prolonged courses of melphalan, ranging from 30 to 57 months [28]. Average time from melphalan initiation to acute leukemia diagnosis was 3.4 years. A retrospective study conducted by the Finnish Leukemia Group involving 432 patients with MM diagnosed and treated with an average of 19 cycles of melphalan-based therapy between 1979 to 1985 noted a nearly 50% increased risk of acute leukemia compared to the general population (SIR 1.45; p<0.001) [29]. The highest risk was seen 5-9 years after first chemotherapy treatment for MM. No significant association was noted between duration and cumulative doses of melphalan and AML SPM risk.

In addition, several randomized-controlled trials of MM patients treated with alkylators noted an increased hematologic SPM risk [16, 30, 31]. An analysis of 648 MM patients diagnosed between 1964 and 1975 and randomized to receive either melphalan or cyclophosphamide found a 5-year MDS and AML SPM prevalence of 3% [30]. The majority of SPMs developed more than 5 years after diagnosis and initiation of treatment. In contrast to the population study by the Finnish Leukemia Group, the cumulative dose of melphalan received up to 3 years before leukemia diagnosis was reported to be the most important determinant of SPM risk. Another trial randomizing newly diagnosed MM patients from 1973 and 1977 to receive either sequential, alternating or concurrent melphalan, cyclophosphamide and carmustine therapy noted a greater than expected incidence of acute leukemia, reaching an incidence of 17.4% at 50 months [31]. The risk was higher in the melphalan arm compared to the cyclophosphamide arm. However, another retrospective review conducted in Japan of 491 newly diagnosed MM patients still noted an association between cyclophosphamide use and SPM development [16].

Most studies of alkylators and their relationship to SPM development are 30 to 40 years old and did not take into account the competing risk of death. Thus, more effective therapy may be associated with higher SPM rates due to longer survival times, particularly in eras when median survivals were less than 5 years. Importantly, the latency periods noted in these studies is similar to those described in other malignancies treated with alkylator based therapy, in which AML and MDS develop 5-7 years after treatment [32, 33]. In summary, alkylators likely play a role in SPM development in this patient population, though melphalan in particular, when compared to cyclophosphamide, appears to be the most mutagenic – particularly when given in high cumulative doses over long periods of time.

SPM in MM: association with autologous stem cell transplantation

The use of ASCT is associated with a potential increase in hematologic and some solid tumor, specifically skin and colon cancer, SPM development [6, 10, 18, 19, 34, 35]. However, multiple studies suggest that the increased SPM development is primarily due to alkylator therapy prior to ASCT as opposed to ASCT itself as multiple studies suggest no difference in SPM incidence prior to and after ASCT.

While the analysis of the Swedish registry showed a significantly increased risk of subsequent AML development with a SIR of 11.51 (95% CI: 8.19-15.74), AML incidence was similar before and after 1995, the year ASCT was introduced in Sweden [6]. Likewise, a large population study examining SPM trends in MM from nine state registries between 1973 and 2008 in the United States reported a decreased risk of breast and prostate cancer and an increased risk of colon cancer [10]. Additionally, increased SPMs were noted across all hematologic malignancies, particularly in AML with a SIR of 6.51 (95% CI: 5.42-7.83). No change in SPM risk was noted after the introduction of ASCT. These two studies rely on period analysis, rather than direct measures of exposure to ASCT, thus potentially misclassifying patients as having received or not received ASCT. In addition, by comparing across eras, the effects of changes in supportive measures and other treatment modalities can be conflated with those of ASCT use.

To address these issues, 2 analyses of the Center for International Blood and Marrow Transplant Research (CIBMTR) registry examined the incidence of SPM development among MM patients receiving upfront ASCT between 1990 and 2010; both reported an increased risk of hematologic malignancies and skin cancers compared to the general population [18, 34]. Raval et al. reported an increased risk of MDS with a SIR of 105.63 (99% CI: 43.51-212.54), AML with a SIR of 7.45 (99% CI: 1.25-23.45) and non-melanoma skin cancers with a SIR of 29.39 (99% CI: 4.94-92.54). Mahindra et al. reported similar conclusions with SIRs of 5.19 (99% CI: 1.67-12.04) for AML and 3.58 (99% CI: 1.82-6.29) for melanoma. A third study attempted to compare the incidences of SPM in MM patients after ASCT, reported from CIBMTR, to those seen in SEER registries considered as a control [19]. They found relative risks of 5-10 for AML and MDS within SEER compared to 10-50 for AML and approximately 100 for MDS within CIBMTR. The authors concluded that there was a substantial increased risk of AML and MDS after ASCT. While the CIBMTR collects robust data on ASCT patients, including prior therapy and maintenance therapy use, data is collected differently than by the SEER registries. This cross-registry comparison can be an additional source of bias. Moreover, by comparing rates of SPM development in MM patients undergoing ASCT to malignancy development in the general population, the contribution of factors intrinsic to MM patients are conflated with the effect of ASCT use, potentially biasing these studies.

To address this, a study of the California Cancer Registry linked to statewide hospitalization data compared the cumulative incidence of SPM development, accounting for the competing risk of death, in MM patients undergoing ASCT to MM patients who did not [35]. There was a 1.3% absolute increased cumulative incidence of developing hematologic malignancies in those who underwent ASCT, which corresponded to an adjusted hazard ratio of 1.51 (95% CI: 1.01-2.27), but not solid SPMs at 10 years. When examining cause of death, death due to SPM was 3.0% (95% CI: 2.7%-3.3%) at 10 years after MM diagnosis, compared to 59.1% (95% CI: 58.2%-60.0%) cumulative incidence of myeloma-specific mortality.

Overall, a small but statistically significant risk of SPM development is likely attributable to ASCT use. When compared to the background population, as has been done with analyses comparing the CIBMTR to SEER, the rates of AML and MDS are significantly higher. However, when looking at MM patients by era, the introduction of ASCT did not appear to alter the rates of SPM development at the population level. Finally, when comparing ASCT use among MM patients diagnosed in California, a small increase in the rate of hematologic SPMs was seen, though the rates of death due to SPMs was low when compared to death due to MM.

SPM in MM: association with immunomodulatory drugs

Immunomodulatory drugs were introduced as treatment for multiple myeloma in 1996-1998, leading to significantly improved overall survival [36]. Clinical trials of immunomodulatory drugs have found higher than expected rates of SPMs in newly diagnosed, relapsed/refractory, transplant-eligible and transplant non-eligible populations. Specifically, hematologic SPM rates, especially rates of MDS, acute leukemias and lymphomas, were noted to be significantly higher in lenalidomide treated arms compared to those without lenalidomide [37, 38]. The vast majority of the studies evaluated SPM risk in lenalidomide, the most commonly used immunomodulatory drug in MM due to its more potent and less toxic profile compared to thalidomide [39]. Few studies also evaluated SPM risk in thalidomide, finding a smaller but still increased risk while data regarding SPM in pomalidomide is minimal [17].

Immunodulatory drugs: newly diagnosed MM

A population study of the SEER-Medicare database conducted by Giri et al. of treated MM patients diagnosed between 2007-2015 noted a 5-year SPM cumulative incidence of 4.3% [41]. In a multivariable competing risk model, there was no association between first-line lenalidomide containing therapy and SPM development overall (sub-hazard ratio [SHR]: 1.06, 95% CI: 0.88-1.29), in solid tumors or hematologic cancers (SHR: 1.13, 95% CI: 0.92-1.40 and SHR: 0.72, 95% CI: 0.42-1.23), respectively. An analysis of the Connect MM, a US-based multicenter, prospective observational cohort registry study designed to observe the management and outcomes of patients with newly diagnosed MM found the 3-year incidence of invasive SPM to be 4% with a SIR of 1.61 (95% CI: 1.25-2.09) [20]. Consistent with the SEER-Medicare database analysis by Giri and colleagues, lenalidomide was not found to be associated with an increased risk of SPM. Both of these studies collected data prospectively, though the former was a retrospective analysis. However, median follow up times were short at 5 and <3 years respectively. Thus changes in the results once longer follow up is achieved may be reported in future analyses.

These results are in contrast with an analysis of 2732 MM patients enrolled in the Myeloma XI trial which randomized newly diagnosed transplant-eligible and ineligible MM patients to either upfront cyclophosphamide, thalidomide and dexamethasone (CTD) or cyclophosphamide, lenalidomide and dexamethasone (RCD) [17]. Those that did not achieve ≥ Very Good Partial Response were further randomized to cyclophosphamide, bortezomib and dexamethasone (VCD) or proceeded directly to the next phase of treatment. Transplant-eligible patients went on to ASCT followed be a subsequent randomization to maintenance therapy with either lenalidomide, lenalidomide with vorinostat or observation. Transplant-ineligible patients proceeded directly to the maintenance randomization. This trial reported a 3-year overall cumulative SPM incidence of 3.8% (95% CI: 2.9-4.6%). 3-year SPM rates were similar with RCD and CTD induction in both transplant-eligible and transplant-ineligible patients, with adjusted hazard ratios of 1.74 (95% CI: 0.88–3.45) and 0.86 (95% CI: 0.52–1.42) respectively, after accounting for the competing risk of death. In contrast, lenalidomide maintenance was associated with increased SPM incidence compared to observation in both transplant-eligible and ineligible patients.

A meta-analysis of seven clinical trial conducted between 2000 and 2012 found an overall SPM incidence of 6.9% (95% CI: 5.3%-8.5%) with lenalidomide versus 4.8% (95% CI: 2.0-7.6) without lenalidomide at 5 years, after accounting for the competing risk of death [40]. Solid SPM incidences were similar between the two groups, but hematologic SPM incidences were higher with lenalidomide at 3.1% (95% CI: 1.9-4.3) compared to 1.4% (95% CI: 0-3.6) without lenalidomide. This study concluded that although higher SPM incidence was noted in lenalidomide maintenance therapy and warrants ongoing monitoring, lenalidomide benefit on survival still likely outweighed its risk. However, in contrast to the previous study [20], this study also reported that exposure to lenalidomide plus oral low-dose melphalan increased the risk of hematologic SPM compared to oral low-dose melphalan alone.

Although some population studies suggest otherwise, trials which randomized patients to upfront lenalidomide treatment found increased risks of hematologic SPM development compared to non-lenalidomide treatment [17, 40]. This increased risk is seen in both transplant-eligible and ineligible patients.

Immunodulatory drugs: relapsed/refractory treatment

Studies consistently report an increased risk of hematologic SPM with lenalidomide in the relapsed/refractory setting. An institutional retrospective review of 195 patients with relapsed/refractory MM treated with lenalidomide and dexamethasone reported an incidence ratio of 2.37 compared to the general population, with MDS being the most common SPM noted [15].

The POLLUX trial which compared lenalidomide and dexamethasone either alone or in combination with daratumumab in MM patients who have received one or more previous lines of therapy noted low rates of SPMs in both groups with 2.8% in the daratumumab group and 3.6% in the control group [42]. In addition, a meta-analysis of 11 clinical trials of relapsed/refractory MM patients treated with lenalidomide showed an increased invasive SPM risk with SIR of 2.08 (95% CI: 1.60-2.60) compared to the general population [43]. A subgroup analysis reported SIRs of 3.98 (95% CI: 2.51-6.31) with lenalidomide/dexamethasone therapy and 1.38 (95% CI: 0.44-4.27) with dexamethasone monotherapy. The paper concluded that the benefit/risk profile of lenalidomide/dexamethasone therapy remained positive.

Immunodulatory drugs: post-ASCT maintenance

Although some smaller studies suggest no association between SPM development and lenalidomide in the post-ASCT maintenance setting [44,45], 4 major randomized-controlled trials with robust long-term follow-up suggest SPM risk to be significantly increased with lenalidomide in this setting (Table 2) [46–50]. In addition, long-term follow up from these trials not only reported an increased risk of hematologic SPMs but also observed an increased risk for solid tumor SPMs such as breast, colon cancer, melanoma and non-invasive skin cancers in this setting as well.

Table 2:

Randomized-controlled trials evaluating the incidence of secondary primary malignancies with lenalidomide maintenance after autologous stem cell transplantation in multiple myeloma

Author	Study	Study period	Patients (n)	Intervention	Follow-up	All SPM n (%)	Hematologic SPM n (%)	Solid tumor SPM n (%)	Latency (median)
McCarthy et al. [46,48]	CALGB 100104	2005-2009	460	lenalidomide maintenance vs placebo after ASCT	91 months	Lenalidomide 32 (7.0); Placebo 12 (2.6)	Lenalidomide 18 (7.8); Placebo 3 (1.3)	Lenalidomide 14 (6.1); Placebo 9 (3.9)	Hematologic 49.8 months; Solid 21.7 months
Attal et al. [47,50]	IFM 2005-02	2006-2008	614	lenalidomide maintenance vs placebo after 2 cycles of lenalidomide consolidation post ASCT	60 months	Lenalidomide 44 (7.2); Placebo 28 (4.6)	Lenalidomide 20 (6.6); Placebo 6 (1.9)	Lenalidomide 24 (7.8); Placebo 11 (4.8)	NR
Palumbo et al. [40]	RV-MM-PI-209	2007-2009	402 ; 231 randomized to maintenance vs no maintenance	lenalidomide maintenance vs observation after ASCT vs melphalan, prednisone and lenalidomide randomization	51 months	During maintenance, Lenalidomide 5 (4.3); Placebo 5 (4.3)	Lenalidomide 0; Placebo 1 (0.9)	Lenalidomide 5 (4.3); Placebo 4 (3.4)	NR
Jones et al. [17,52]	Myeloma XI	2011-2017	4420 overall; 1971 randomized to maintenance treatment	lenalidomide maintenance vs observation after intensive vs non-intensive induction treatment	36 months	Lenalidomide (5.3) Placebo (3.1)	NR	NR	NR

NR-not reported; SPM-secondary primary malignancies

An institutional study which examined 313 patients with MM who received ASCT showed no significant difference of SPM incidence between those who received thalidomide maintenance versus interferon maintenance or no maintenance therapy [44]. Similarly, the use of lenalidomide, in combination with bortezomib and dexamethasone, did not change SPM rates when compared with either thalidomide or interferon maintenance in the Total Therapy program [45].

In contrast, three large prospective randomized trials observed significantly increased SPM risk associated with lenalidomide maintenance after ASCT (Table 2) [46–50]. The increased SPM risk was observed to be so significant in this setting that the IFM 2005-02 study was prematurely stopped due to concerns regarding high SPM rates at a median time of 2 years [47], while the CALGB study completed accrual. Long-term follow up for both IFM 2005-02 and CALGB 100104 trials allow for comparisons between limited duration maintenance versus indefinite maintenance.

CALGB 100104, which randomized 460 MM patients assigned to either lenalidomide or placebo maintenance after ASCT, showed hematologic SPM cumulative incidences of 7.8% with lenalidomide and 1.3% with placebo and solid tumor SPM cumulative incidences of 6.1% with lenalidomide and 3.9% with placebo after a median follow-up time of 91 months [46, 51]. Of the SPMs in the placebo arm, all hematological and over 50% of solid tumor SPMs were in the crossover group, implying that the exposure to lenalidomide need not be immediately following high dose chemotherapy to increase the risk of SPMs. Median latency time to hematologic SPM was 49.8 months and 21.7 months for solid tumor SPM. Similarly, the Intergroupe Francophone du Myelome (IFM) 2005-02 trial of 614 MM patients randomized to either lenalidomide maintenance vs observation after ASCT showed a 5-year cumulative hematologic SPM incidence of 6.6% with lenalidomide and 1.9% with observation after ASCT and lenalidomide consolidation [47, 50]. 5-year solid tumor SPM incidences were 7.8% with lenalidomide and 4.8% with observation. The majority of the hematological SPMs were MDS and AML. However, a few cases of B-cell acute lymphoblastic leukemia were also reported. A variety of invasive solid tumor SPMs such as breast, colon cancer and melanoma were observed. Non-invasive skin cancer such as basal cell carcinoma and squamous cell carcinoma were also noted. While these two trials had similar designs, it is important to note differences in induction therapy which could have affected outcomes. In the IFM2005-02 trial, all patients received either vincristine-doxorubicin-dexamethasone or bortezomib-dexamethasone as induction. In CALGB100104, the induction regimens varied greatly, with 80% of patients receiving an immunomodulatory agent (lenalidomide or thalidomide) during induction. Various inductions were well balanced in both trials. Despite these differences, the SPM rates in the two trials are remarkably similar, implying that ongoing lenalidomide exposure after ASCT is the driving force behind SPM development, particularly regarding hematologic SPMs. Regional differences in screening rates for solid tumors are likely different in the two studies, and may possibly account in part for the fact that increase solid tumor SPMs were seen in the lenalidomide arm of CALGB100104 but not in IFM2005-02.

In addition, the GIMEMA (RV-MM-PI-209) trial randomized 273 patients with newly diagnosed MM to either high-dose melphalan followed by ASCT or melphalan, prednisone, lenalidomide consolidation therapy after induction [40]. Two-hundred and fifty-one patients were subsequently randomized to lenalidomide maintenance therapy or observation. Eleven (2.8%) patients had SPMs throughout induction, consolidation and maintenance. Specifically, during the maintenance phase, hematologic SPM rates of 0 with lenalidomide maintenance vs 0.9% with observation and solid tumor SPM rates of 4.3% with lenalidomide maintenance vs 3.4% with observation were noted after a follow-up time of 51 months. Follow up time in this study was short, thus the SPM rates on final report are likely to be higher, and may still find a difference in the maintenance therapy arm.

A meta-analysis of these three trials found a cumulative incidence rate of SPM to be higher with lenalidomide maintenance versus placebo or observation before disease progression at 5.3% and 0.8% respectively and after disease progression at 6.1% and 2.8% respectively. However, the cumulative incidence of disease progression and death was higher with placebo versus lenalidomide maintenance [48].

Lastly, the Myeloma XI trial’s transplant-eligible arm was randomized post-ASCT to either lenalidomide maintenance vs observation. A higher 3-year cumulative SPM incidence with lenalidomide compared to observation was noted, with rates of 5.3% (95% CI: 3.6-7.1) vs. 3.1% (95% CI: 1.8-4.5), respectively (HR 1.85) [17, 52].

Immunodulatory drugs: non-ASCT candidates

Amongst transplant-ineligible patients the data on SPM development related to immunomodulatory therapy is mixed. The MM-015 trial randomized transplant ineligible patients to one of three arms: melphalan prednisone (MP), melphalan, prednisone and lenalidomide (MPR), each given for 9 months, or MPR followed by lenalidomide maintenance (MPR-R) [53]. The 3-year incidence of invasive SPM in patients was 3%, 7% and 7% respectively, suggesting no increased SPM risk with lenalidomide maintenance compared to fixed duration lenalidomide treatment, but that lenalidomide in combination with prolonged exposure to oral melphalan increased the risk of SPMs.

The FIRST trial randomized transplant-ineligible newly diagnosed MM patients to either continuous lenalidomide and low-dose dexamethasone (Rd continuous), lenalidomide and low-dose dexamethasone for 18 cycles (Rd18) or melphalan, prednisone and thalidomide (MPT) for 18 cycles and showed higher rates of hematologic SPM in the MPT arm at 3% compared to Rd continuous (1%) or Rd18 (<1%) at a median follow-up duration of 5.6 years [54]. Incidences of solid tumor SPMs were similar across treatment arms.

The Myeloma XI trial’s transplant-ineligible arm randomized transplant-ineligible patients to either upfront attenuated cyclophosphamide, thalidomide and dexamethasone (CTDa) or attenuated cyclophosphamide, lenalidomide and dexamethasone (RCDa) [17]. Those that did not achieve ≥ Very Good Partial Response were further randomized to attenuated cyclophosphamide, bortezomib and dexamethasone (VCDa) or proceeded directly to maintenance randomization. This randomized-controlled trial reported an overall 3- year SPM incidence of 5.2% among transplant-ineligible patients. 3-year cumulative SPM incidence was similar with both CTDa and RCDa at 5.9%. Lenalidomide maintenance was associated with increased SPM incidence compared to observation at 12.9% and 6.3%, respectively.

In summary, multiple studies report a small but significant SPM risk with lenalidomide exposure when used as upfront therapy, maintenance therapy post ASCT, for relapsed/refractory disease and in both transplant eligible and ineligible patients. However, mixed results are seen with different studies evaluating lenalidomide use in different contexts. In the upfront setting, both meta-analyses conducted by Palumbo and Jones et al. observe an increased risk of hematologic SPM with lenalidomide induction. In the relapsed/refractory setting, the POLLUX trial and a meta-analysis by Dimoupolos et al. continue to show an increased risk of hematologic SPM with lenalidomide treatment. In the maintenance setting after ASCT, significant SPM risk was seen with lenalidomide maintenance after ASCT and the IFM 2005-02 trial was prematurely stopped due to concerns regarding high SPM rate. In addition, both IFM 2005-02 and CALGB 100104 trials also reported an increased risk of solid tumor SPMs such as breast, colon cancer, melanoma and non-invasive skin cancers with lenalidomide maintenance after ASCT as well. Lastly, in the transplant-ineligible setting, both the MM-015 and the FIRST trial suggest the highest risk of SPM with lenalidomide was in combination with low-dose oral melphalan and the MM XI trial demonstrated a higher SPM risk with lenalidomide maintenance regardless of melphalan exposure. Especially in the transplant ineligible population, geographic differences in screening rates for solid tumors, and differences in comorbidities in the different populations may account for some of the disparities in results. Regardless, SPM risk remain a concern across all populations with prolonged exposure. Despite the increased SPM risk, PFS and OS benefits of lenalidomide benefit still likely outweighed its risk and lenalidomide treatment in both upfront and as maintenance therapy post ASCT continues to be listed as category 1 on NCCN.

SPM in MM: with proteasome inhibitors

Proteasome inhibitors such as bortezomib, carfilzomib and Ixazomib are approved for the treatment of untreated and relapsed/refractory MM in combination with other drugs. Randomized-controlled trials consistently show no increased SPM risk with the addition of proteasome inhibitors, however, robust long-term follow up has not been reported [55–57]. The VISTA trial which randomized MM patients to either bortezomib, melphalan and prednisone (VMP) or melphalan, prednisone (MP) showed no difference in incidence of SPMs between the two group [55]. Likewise, the ASPIRE trial which randomized relapsed/refractory MM patients to either carfilzomib, lenalidomide and dexamethasone (KRd) or lenalidomide and dexamethasone (Rd) noted similar SPM rates of 2.8% with KRd and 3.3% with Rd [58]. Most common SPMs were hematologic in nature and included MDS and AML. The ENDEAVOR trial which evaluated carfilzomib and dexamethasone (Kd) versus bortezomib and dexamethasone (Vd) in relapsed/refractory MM patients noted 1 case of AML with carfilzomib versus 0 with bortezomib [56]. The TOURMALINE-MM1 trial which randomized 722 patients with relapsed/refractory MM to Ixazomib with lenalidomide and dexamethasone or placebo plus lenalidomide and dexamethasone showed no difference in SPM incidence between the two groups.[59] Similarly, the MM3 trial which randomized MM patients who had achieved at least a partial response after undergoing standard of care induction therapy followed by high dose melphalan conditioning and single ASCT to either oral Ixazomib or placebo for 2 years showed no similar SPM incidences between the two groups [57].

SPM in MM: with Daratumumab

Daratumumab is a monoclonal antibody used to treat relapsed/refractory or transplant-ineligible MM both in combination with other drugs or as monotherapy. Studies thus far do not appear to show an increased SPM risk with the addition of daratumumab, however, robust long-term follow up is minimal at this time.

The POLLUX trial which compared lenalidomide and dexamethasone either alone or in combination with daratumumab noted low rates of SPMs in both groups with 2.8% in the daratumumab group and 3.6% in the control group. In addition, 10 out of 18 patients with SPMs had noninvasive, continuous SPM such as squamous cell carcinoma or basal cell carcinoma suggesting no increased risk of SPM with the addition of daratumumab [42]. Another meta-analysis of 5 phase III randomized-controlled trials looking at several daratumumab regimens such as bortezomib, melphalan and prednisone with and without daratumumab, lenalidomide and dexamethasone with and without daratumumab, bortezomib and dexamethasone with and without daratumumab and bortezomib, thalidomide and dexamethasone with and without daratumumab showed similar SPM incidences of 4.3% between both treatment and control groups suggesting no increased SPM risk with addition of daratumumab [60].

Screening for SPMs

Although multiple studies highlight the significance of SPM in MM, minimal prospective data and few screening guidelines exist for this patient population. As multiple studies show no increased risk and some even suggest a decreased trend of solid tumor SPMs among MM survivors, the International Myeloma Foundation’s Nurse Leadership Board (NLB) and the International Myeloma Working Group consensus recommend no changes from age-appropriate screening guidelines for the general population for breast, cervical, colorectal and prostate cancer [61,62]. Instead, diagnostic measures that would aid in the detection of suspected SPMs during daily clinical work-up should be considered on a cases-by case basis. Given a slight increase of skin cancer (both melanoma and non-melanoma) SPM among MM survivors seen in several population studies [18, 34, 46, 50], the NLB recommends patient education regarding the ongoing need for routine self-skin examination [61].

Given the increased risk of hematologic SPMs, especially MDS and acute leukemia, seen among survivors of MM, the International Myeloma Working Group highlighted the need for heightened awareness and recommended bone marrow examination at baseline and in the event of any unexplained blood count abnormalities [62]. In addition, every SPM case should be carefully reviewed to accurately assess the true impact of treatment on SPM development and to prevent false inflation of reported SPM rates.

Conclusion and Future Considerations

Due to significant improvement in survival trends among patients with MM, SPM among MM survivors become increasingly relevant. Different anti-myeloma therapies pose variable risks to SPM incidence, however an increased SPM risk, especially hematologic SPMs, were noted overall. Prolonged treatment with alkylators, especially oral melphalan, was associated with an increased hematologic SPM risk. Likewise, ASCT also appeared to minimally increase SPM risk, while immunomodulatory drugs, specifically lenalidomide, was consistently associated with a hematologic SPM risk when used in multiple contexts such including induction, maintenance after ASCT, relapsed/refractory and transplant-ineligible setting. A trend for increased solid tumor SPM risk with lenalidomide was also noted in the maintenance setting. However, the benefits of lenalidomide therapy clearly outweighed any risk of SPMs. Newer anti-myeloma therapy such as proteasome inhibitors and monoclonal antibodies did not appear to increase SPM risk although robust long term follow-up is needed. There continues to be multiple emerging novel anti-myeloma therapy for which we have minimal data regarding SPM risk. Therefore, clinical trials to evaluate emerging anti-myeloma therapy should be designed to include enhanced monitoring, precise measurements, and well-defined endpoints for secondary cancers.

Although studies consistently showed an increased risk of hematologic SPM risk among MM survivors, there is a dearth of recommendations and guidelines for screening and this should be developed and validated. In addition, there should be a heightened risk of awareness and a low threshold for suspicion for SPM and appropriate comprehensive work-up for any clinical changes should be completed before concluding a diagnosis of SPM versus assuming MM disease relapse. In addition, although studies overall did not show an increased risk for solid tumor SPMs among MM survivors, most studies also did not show a significantly decreased risk. Therefore, MM survivors should continue to receive age-appropriate oncologic screening that is recommended for the general population.

While the relative risk of SPM among MM survivors is increased, especially with different emerging anti-myeloma treatments, the absolute risk of SPM development remains low overall and is in part related to the lengthening survival of patients with MM. With every emerging treatment, multiple randomized-controlled trials and meta-analyses continue to consistently conclude a benefit over risk ratio in regards to SPM development. Therefore, based on available evidence, the potential risk of SPMs in MM should not generally alter the current therapeutic decision-making process.

As the pathogenesis of SPMs in MM has shown to be multifactorial which includes age, sex, race/ethnicity, disease cytogenetics and genetic polymorphisms, biologic samples from MM patients included in clinical trials and, when possible, encountered in clinical practice, should be collected and stored for genetic analysis and gene expression profiling. In addition, collection of germline and tumor-related material and re-banking of biologic samples during the course of the disease will provide valuable insight regarding the effect of treatment on SPM risk.

Practice Points

• Melphalan is associated with increased hematologic SPM risk and prolonged use should be avoided if possible

• ASCT in MM appears to potentially increase SPM risk. However, risk is minimal and therefore, all eligible MM patients should still be evaluated for ASCT with high dose melphalan myeloablative therapy

• Lenalidomide is associated with increased hematologic SPM risk although the survival benefit outweighs the SPM risk. SPM risk is highest in the maintenance setting and in addition with melphalan, therefore lenalidomide in combination with oral low-dose melphalan should generally be avoided in this setting.

• Proteasome inhibitors and monoclonal antibodies do not appear to increase SPM risk although robust long-term follow-up is lacking

Research Agenda

• SPM risk with newer emerging anti-myeloma therapy

• Guidelines for SPM screening among MM survivors

• Prospective studies evaluating screening