Evolving patterns in systemic treatment utilization and survival among older patients with advanced cutaneous melanoma

Abstract Introduction In the last decade, melanoma treatment has improved significantly. However, data on population‐level treatment utilization and survival trends among older patients is limited. This study aimed to analyze trends in systemic anticancer therapy (Rx), including the uptake of immune checkpoint inhibitors (ICIs), in conjunction with trends in cause‐specific survival among older patients (66+) diagnosed with advanced melanoma (2008–2019). Methods We used the Surveillance, Epidemiology, and End Results (SEER)‐Medicare Condensed Resource to assess any Rx utilization among patients first diagnosed with advanced melanoma in 2008–2010, 2011–2014, and 2015–2019, stratified by stage, and type of first‐line Rx among patients receiving Rx. The SEER dataset was used to evaluate trends in cause‐specific survival by year of diagnosis. Results Rx utilization (any type) almost doubled, from 28.6% (2008–2010) to 55.4% (2015–2019) for stage 3 melanoma, and from 35.5% to 68.0% for stage 4 melanoma. In 2008–2010, the standard first‐line treatment was cytokines/cytotoxic chemotherapy/other. By 2015–2019, only 5.1% (stage 3) and <3.6% (stage 4) of patients receiving Rx received these agents, as ICIs emerged as the dominant treatment. Both 1‐year and 5‐year cause‐specific survival significantly improved since 2010 for stage 4 and since 2013 for stage 3. Conclusions This study shows a significant rise in Rx utilization and a rapid transition from cytokines/cytotoxic chemotherapy to ICIs, reflecting a rapid uptake of highly effective treatment in a previously challenging disease with limited options before 2011. The documented survival improvement aligns with the adoption of these novel treatments, underscoring their significant impact on real‐world patient outcomes.


| INTRODUCTION
For decades, there were limited systemic treatment options for advanced melanoma, with the mainstay of therapy being cytotoxic chemotherapy agents, such as dacarbazine, and cytokine-based immunotherapy, such as interleukin-2 (IL-2).These treatments provided very limited survival benefits and were often accompanied by high toxicity. 1,2][5][6] Since 2011, the treatment landscape for advanced melanoma has evolved rapidly, with the introduction of the first immune checkpoint inhibitor (ICI) and v-raf murine sarcoma viral oncogene homolog B1 (BRAF) inhibitor in 2011, and mitogen-activated protein kinase kinase (MEK) inhibitor in 2013.Programmed death-1 (PD-1) inhibitors were first approved in 2014.(Figure 1).
Given the increased availability of effective treatment options in this new era of melanoma treatment, it is important to understand how patterns of systemic treatment among older patients with advanced melanoma may have changed.It is also imperative to understand whether outcomes, such as survival, have changed over time for older patients with melanoma, alongside recent advances in treatment.
][10][11] However, to our knowledge, no studies have analyzed trends in survival in conjunction with the availability of new treatments.Hence, in this study, we aimed to assess (1) trends in any melanoma systemic treatment receipt, (2) factors associated with any melanoma systemic treatment receipt, (3) changes in the selection of first-line systemic treatment over time among older patients with advanced melanoma, and (4) trends in causespecific survival, examining periods that cover both the pre-and post-novel therapy eras.Patients are also excluded if the tumor is nonmalignant/ malignancy unknown, the month of diagnosis is unknown, or if the cancer is diagnosed at autopsy or through a death certificate only. 13he analytic cohort for this study was further restricted to patients who were included in both ABFFS and D cohorts, i.e., enrolled in Medicare Parts A, B and D, and who had a diagnosis of stage 3 or 4 cutaneous melanoma (ICD-O-3 site: C440-C449, ICD-O-3 histology: 8720-8790; first or only primary) between January 1, 2008 and December 31, 2019.Continuous inclusion in the ABFFS and D cohorts was required for at least 12 months following diagnosis (including the month of diagnosis) or until death, whichever occurred first, to ensure that patients' claim records were complete for treatment assessment.We also excluded patients with unknown or missing baseline patient variables.

Systemic treatment
We assessed any melanoma-directed systemic treatment receipt from the date of diagnosis until death, or 12 months (including the month of diagnosis), whichever occurred first, and the first systemic treatment regimen administered or dispensed following diagnosis.The date of diagnosis was set to the first of the month of diagnosis as only the month and year of diagnosis are available in the data.We identified systemic treatments from the CoRe systemic AB and D files.Generic drug names were used to identify treatments (Table S1).We also identified "unspecified" drugs, i.e., claims with an ICD-9/10 diagnosis/procedure code or HCPCS code indicating cancer systemic treatment administration but without a drug-specific HCPCS or NDC code for the same claim ID and date combination.The first systemic treatment regimen was defined as the first systemic agent(s) beginning on or after the date of diagnosis, and any other agents started within 28 days of the first systemic agent.Treatment regimens were categorized as: (a) BRAF inhibitor with or without a MEK inhibitor, (b) ipilimumab (CTLA-4) monotherapy, (c) PD-1 inhibitor monotherapy (which included pembrolizumab or nivolumab), (d) ipilimumab plus PD-1 inhibitor combination therapy, (e) cytokines, cytotoxic chemotherapy agents (e.g., interferon alfa-2b, dacarbazine, temozolomide) and/ or other melanoma systemic agents (e.g., tamoxifen), (f) ICI with any other agent(s), i.e., ICI plus a BRAF/MEK inhibitor, cytokines, cytotoxic chemotherapy or other agent, and (g) unspecified drug(s), i.e., ICD-9/10 diagnosis/procedure code or HCPCS code indicating cancer systemic treatment administration but without a drug-specific code (and no other drug-specific code found within the 28-day window).

Cohort characteristics & trends in any systemic treatment utilization
We first examined the sociodemographic and clinical characteristics of all patients included in the analysis.Characteristics of patients who received any systemic treatment during the first 12 months of diagnosis were described using frequencies and percentages, stratified by stage.We assessed the number and proportion of patients who received any systemic melanoma treatment during each diagnosis period (2008-2010, 2011-2014,  2015-2019), stratified by stage.In addition, we described the characteristics of patients who received any systemic treatment during the first 12 months of diagnosis, stratified by stage and diagnosis period, to gain further insight on the characteristics of patients receiving systemic treatment in each period.

Factors associated with any melanoma systemic treatment utilization
Multivariable logistic regression was used to estimate covariate-adjusted odds ratios (aORs) and 95% confidence intervals (CIs) to identify factors associated with any systemic treatment receipt among patients with stage 3 and 4 melanoma separately.All baseline patient characteristics were included in the regression model except for race and ethnicity, which could not be included due to the small number of patients in the NH Other, NH Unknown, and Hispanic categories.Variables were selected a priori.

Trends in first systemic treatment type
Among patients who received systemic treatment, the type of first systemic treatment was described as frequencies and proportions, stratified by stage and diagnosis period (2008-2010, 2011-2014, 2015-2019).These periods were chosen to capture systemic treatment utilization: (1) prior to the introduction of ICIs and BRAF/ MEK inhibitors (2008-2010), (2) following the approval of the first ICI (ipilimumab; CTLA-4 inhibitor) and BRAF inhibitor (2011-2014), and (3) following the approval of PD-1 inhibitors (2015-2019).In addition, we described the characteristics of patients with stage 4 melanoma who received PD-1 inhibitor monotherapy versus ipilimumab and PD-1 inhibitor combination therapy.

| Survival analysis
We used the November 2022 submission of the SEER-17 registries data to assess for changes in survival for patients diagnosed with stage 3 and 4 malignant melanoma, and aged <66 years and ≥ 66 years old.The survival analysis was not limited to patients in SEER-Medicare.Patients with tumors diagnosed through death certificate/autopsy only, patients with missing/unknown cause of death, and patients lost to follow-up in the month of diagnosis (alive with no survival time) were excluded from the analysis.The analysis was limited to patients with first primary tumors only (sequence number 0 or 1).Using the SEER*Stat software, 15 we calculated up to 5-year causespecific survival.Note that 1-year cause-specific survival was calculated for patients diagnosed between 2004 and 2019, and 5-year cause-specific survival for patients diagnosed between 2004 and 2015 as we had follow-up data only through 2020.
We then used the JPSurv webtool 16,17 to characterize cause-specific survival trends by year of diagnosis.JPSurv estimates the location and number of Joinpoints, i.e., years at diagnosis when changes in survival trends occurred, and the average absolute change (AAC) in survival between Joinpoints to understand how much survival has been increasing or decreasing.The year of diagnosis range was set to 2004-2019, and a maximum of two Joinpoints were tested.All other settings were set to default.The final model for each age and stage combination was selected based on the Bayesian Information Criterion and Akaike Information Criterion.This study did not require Institutional Review Board review as secondary analyses of de-identified data are not considered human subjects research by the National Institutes of Health.

| Cohort characteristics & trends in any systemic treatment utilization
The final cohort included 2012 patients diagnosed with stage 3 and 4 melanoma between 2008 and 2019 (Table 1).Among the 1198 and 814 patients with stage 3 stage 4 melanoma, 43.2% and 56.8% received melanoma-directed systemic therapy in the first year following diagnosis, respectively.Overall, there was an almost two-fold increase in the use of systemic treatment over time, from 28.6% (2008-2010) to 55.4% (2015-2019) among patients with stage 3 melanoma, and from 35.5% (2008-2010) to 68.0% (2015-2019) among patients with stage 4 melanoma (Figure 2).Table S2 describes the characteristics of patients who received any systemic treatment during the first 12 months of diagnosis, stratified by stage and diagnosis period.

| Factors associated with any melanoma systemic treatment utilization
The multivariate logistic regression (Table 2) assessing factors associated with the receipt of any systemic treatment indicated that the odds of receiving systemic treatment were significantly higher for patients diagnosed  ).Among patients with stage 4 melanoma, being married or having a domestic partner was associated with higher odds of systemic treatment receipt, and an NCI comorbidity index of >1 was associated with lower odds of treatment receipt compared to having an index of zero.Among patients with stage 3 melanoma, LIS enrollment status was associated with lower odds of systemic treatment receipt.

| Trends in first systemic treatment type
There were substantial changes in melanoma systemic treatment utilization over the study period.Among patients diagnosed in 2008-2010 receiving systemic treatment, the standard was cytokines, cytotoxic chemotherapy, and/or other agents, e.g., tamoxifen.However, among patients diagnosed in 2015-2019, a small percent received these therapies (stage 3: 5.1%, stage 4: <3.6%;Table 3), with the dominant treatment being PD-1 inhibitors, which was used as monotherapy by 69.9% of patients with stage 3 melanoma and 54.6% of patients with stage 4 melanoma, and in combination with ipilimumab by an additional 4.0% of patients with stage 3 melanoma and 22.7% of patients with stage 4 melanoma.The use of BRAF/MEK inhibitors among patients diagnosed in 2011-2014 compared with 2015-2019 declined from 12.4% to 3.7% (stage 3) and from 15.5% to 8.4% (stage 4).Characteristics of patients with stage 4 melanoma who received PD-1 inhibitor monotherapy versus ipilimumab and PD-1 inhibitor combination therapy are described in Table S3.

| Survival
For stage 4 melanoma, Joinpoints were estimated in 2010 for both age groups (<66 and 66+) (Figure 3).Conversely, for stage 3, changes in survival trends were estimated in 2011 (ages <66) and in 2013 (ages 66+).Post-Joinpoints, all groups-stage 3 and 4, and age groups <66 and 66 +showed a steeper increase in both 1-year and 5-year survival compared to the prior period.In some cases, survival was not increasing in the prior period (Table S4).The greatest improvement in 5-year survival was observed for patients aged 66+ years with stage     The increased use of systemic therapies and the shift in the melanoma treatment landscape align with the improvements in cause-specific survival among older patients with melanoma.Whereas before 2010, survival was not increasing, both 1-year and 5-year survival have been increasing since 2010 among patients aged 66+ years with stage 4 melanoma.Among patients aged 66+ years with stage 3 melanoma, increases in survival trends were observed beginning a few years later in 2013, which may be due to slower adoption of ICIs and inhibitors.Among patients with stage 3 melanoma, older individuals had greater improvements in survival than younger individuals, and for stage 4 melanoma, improvements in survival were similar between older and younger patients.These findings indicate that improvements in survival in the new era of melanoma treatment are similar, if not greater, among older patients with advanced melanoma compared with younger patients, and appear to be continuing to improve over time.
In our study, patients aged 85+ years had lower odds of systemic treatment receipt compared with the youngest age group (66-69 years).Older patients may be more frail and there may be greater concerns about treatmentrelated morbidity. 18It is possible that the perceived risks outweigh the perceived benefits of systemic treatment for some older patients, or systemic treatment might not be part of the patient's goals of care.][21] The observed association between comorbidity and treatment utilization may be related to concerns for treatment tolerability among patients with a higher comorbidity burden.Interestingly, patients with stage 4 melanoma who were married or had a domestic partner were more likely to receive systemic treatment than patients who were single or widowed.This result agrees with findings from previous cancer literature. 11,22,23ur analysis revealed a clear shift in the treatment landscape between 2008 and 2019, with an uptake of ICIs as first-line therapy, and a decline in the use of traditional cytotoxic chemotherapy agents and cytokines.The use of ipilimumab in combination with a PD-1 inhibitor was lower overall than the use of PD-1 inhibitor monotherapy, and this is likely due to increased concerns about the toxicities of combination treatment. 24The results showed that ipilimumab plus PD-1 inhibitor combination therapy was higher among patients with stage 4 melanoma compared with patients with stage 3 melanoma.This is not surprising as ICI combination treatment is approved for unresectable or metastatic melanoma and is more likely to be used among patients with more severe disease, such as patients with brain metastases.Additionally, our results indicate that BRAF/MEK inhibitor use has not been as high as the ICIs.This agrees with the findings of the study by Lee et al. which also saw lower use of BRAF/MEK inhibitors compared with ICIs, and a decline in the use of BRAF/MEK inhibitors after 2015 among a cohort of older patients diagnosed with advanced melanoma between 2012 and 2017 who received immunotherapy or targeted therapy. 7Use of BRAF/MEK inhibitors may be lower than the use of ICIs because BRAF/MEK inhibitors are only indicated for patients with BRAF mutations whereas ICIs can be used regardless of patients' mutation status.In the DREAMseq study, 25 first-line combination immunotherapy was found to be superior to first-line targeted therapy (dual BRAF/MEK inhibition) in patients with BRAF mutations.Although this study was published after the observation period that we employed, advanced knowledge of some results presented at an American Society of Clinical Oncology (ASCO) virtual plenary session could have contributed to higher utilization of ICIs than BRAF/MEK inhibitors.In addition, the need for molecular testing for mutations adds another step and may influence treatment selection, especially outside of academic medical centers.A molecular testing rate of 73.8% has been observed in a study by Hill et al, 25 showing that not all patients receive molecular testing.
The findings of this study need to be considered in the context of its limitations.Treatments provided through clinical trials are not captured in the Medicare claims, and hence we may have underestimated the number of patients receiving systemic treatment within our study population.][28] It should be noted that most patients with metastatic melanoma are initially diagnosed with early-stage disease and thus may not be treated with systemic therapy unless they have a recurrence.We were not able to assess systemic therapy among patients with recurrent disease because neither SEER nor Medicare data reliably identify recurrence.Thus, our study represents the subgroup presenting with advanced disease, which may differ from those who develop unresectable or metastatic disease later.Also, some patients with stage 3 melanoma at diagnosis may have had rapid disease progression to metastatic melanoma by the time of initiating systemic treatment, but we were unable to identify and address disease progression using the current dataset.
We did not have BRAF or KIT mutation status; therefore, we could not assess whether treatments were prescribed by mutation status.Resectability is a key factor mentioned in multiple FDA drug labels for melanoma indications but is not a concept that is captured by cancer registries.We used surgery receipt as a proxy for distinguishing between resectable and unresectable melanoma; however, there may have been patients with resectable melanoma who did not receive surgery, especially patients with greater comorbidity burden or poor performance status.
We did not limit the survival analysis to patients in the treatment utilization analysis, as the small sample size would not allow us to examine trends.Instead, we used SEER data for the survival analysis which also allowed for estimation of survival trends for patients younger than 66 years.In addition, it is possible that someone diagnosed in one period switched a new drug approved in the next period, which may have lengthened their survival.Since we did not examine these later switches, we may be underestimating the role of new treatments on survival.
Despite these limitations, this study has important strengths.This study adds to the literature with its use of recent data that include patients diagnosed between 2008 and 2019 and their claims through 2020, as well as updated survival estimates.Also, the study demonstrates the use of SEER-Medicare CoRe, a newly developed dataset designed to facilitate the use of SEER-Medicare data with more ease.

| CONCLUSIONS
This study shows a significant increase in the utilization of systemic therapy among older patients with de novo advanced melanoma, and a rapid uptake of highly effective systemic agents, notably ICIs, in a deadly disease that had few effective treatment options before 2011.The substantial uptake of ICIs as well as increased use of systemic treatment overall align with the improvement in survival observed among patients with advanced melanoma.

2 | METHODS 2 . 1 |
Treatment utilization analysis 2.1.1 | Data source The Surveillance, Epidemiology and End Results (SEER)-Medicare Condensed Resource (CoRe) (https:// healt hcare deliv ery.cancer.gov/ seerm edica re/ medic are/ summa rized -data.html) was used for this cohort study to examine treatment utilization.The CoRe data files are a condensed version of the original SEER-Medicare files 12 which have been simplified to facilitate use of SEER-Medicare data. 13Files used for the study included the SEER-Medicare Cancer File, which has information related to the cancer diagnosis and demographics, the CoRe Enrollment File, which includes information on monthly enrollment in Medicare Parts A, B, C (Medicare Advantage) and/or D, the CoRe Prediagnosis Comorbidity File, which has flags for prediagnosis comorbid conditions and the Charlson and NCI Comorbidity Indices, and the CoRe Systemic Treatment AB and D Files, which summarize cancer systemic therapy provided under Medicare Parts A and B, and Part D, respectively. 2.1.2| Study population The SEER-Medicare CoRe files are limited to patients who have a cancer diagnosis included in the SEER-Medicare F I G U R E 1 FDA approval of melanoma treatments (2011-2019).

| 3 of 11 HONG
et al.Cancer File and are continuously enrolled in Medicare Parts A and B fee-for-service (FFS) plans from 12 months prior to cancer diagnosis through at least 1 month after diagnosis (for inclusion in the Part ABFFS cohort) and/or Part D from 4 months prior to cancer diagnosis through at least 1 month after diagnosis (for inclusion in the Part D cohort).Patients remain in the cohort(s) until: (a) subsequent cancer diagnosis, (b) end of ABFFS/D coverage, (c) death, or (d) end of claims (for this analysis December 31, 2020), whichever occurs first.Patients are excluded from the SEER-Medicare CoRe ABFFS cohort if diagnosed prior to 2000 and from the Part D cohort if diagnosed prior to 2008.As age-related eligibility for Medicare begins at the age of 65 years and 12-month prediagnosis data are used for baseline information, patients are excluded from both cohorts if age at diagnosis is <66 years or unknown.

T A B L E 3
and from 35.5% to 68.0% among patients with stage 4 melanoma.The increased use of systemic therapies over time is likely due to the introduction of PD-1 inhibitors which are more effective and less toxic than the traditional cytotoxic chemotherapy agents and cytokines which were the standard of care prior to 2011.Type of first systemic treatment among patients with stage 3/4 melanoma who received any systemic treatment within 12 months of diagnosis (diagnosis years 2008-2010, 2011-2014, 2015-2019), by stage, SEER-Medicare CoRe.
includes ipilimumab and/or PD-1 inhibitor.b "Any Other" includes BRAF/MEK inhibitor(s)/Cytokines/Cytotoxic chemotherapy/Other agents.c Per SEER-Medicare confidentiality policies, cell sizes <11 have been masked.F I G U R E 3 1-year and 5-year cause-specific survival by year of diagnosis.Figures shown by age (≥66 vs. <66 years) and stage (stage 3 vs.4).| 9of 11 HONG et al.
Characteristics of patients diagnosed with stage 3/4 melanoma in 2008-2019 and receipt of any systemic treatment within 12 months of diagnosis, by stage, SEER-Medicare CoRe.
T A B L E 1 Systemic therapy includes any systemic agent(s) listed in Supplemental TableS1and any "unspecified" agent(s).