Utilizing relative survival to investigate melanoma subtypes across intersectional demographics: a Surveillance, Epidemiology, and End Results (SEER) cohort study Authors: Benjamin D Hu, Brandon R Block, Grace Rabinowitz, Hannah Verma, Raphaella Lambert, Bragi Adalsteinsson, Jón Gunnlaugur Jónasson, Bardur Sigurgeirsson, Shane A Meehan, Jonathan Ungar, Jesse M Lewin, Nicholas Gulati, Jonas A Adalsteinsson Published: 2026 ABSTRACT AbstractBackgroundMelanoma accounts for nearly 8000 skin cancer-related deaths annually in the USA. Its impact on survival across intersectional sociodemographic factors remains unclear.ObjectivesTo characterize relative survival rates (RSRs) in patients with melanoma stratified by various intersecting ­sociodemographic factors.MethodsIn this retrospective cohort study, relative survival analysis was conducted on patients with acral lentiginous (ALM), nodular (NM), superficial spreading (SSM) and lentigo maligna melanoma (LMM). Relative survival is advantageous as it compares survival in patients with cancer with the expected survival of individuals without cancer with matched characteristics to minimize noncancer mortality differences. This study utilized the Surveillance, Epidemiology, and End Results (SEER) database, established on cancer registries covering ∼48% of the US population. Age-standardized 5-year RSRs and 95% log–log confidence intervals (CIs) were calculated for each melanoma subtype for subgroups stratified by intersections of age, race, sex, marital status and Breslow thickness.ResultsAll melanoma subtypes except LMM had significant survival reductions, with NM having the lowest RSR (73.5%, log–log CI 72.7–74.3). ALM impacted survival most in male patients, followed by older populations, with younger male patients having lower RSRs than older female patients but younger female patients having similar survival to their counterparts without cancer. Male patients of all races or ethnicities other than White had lower RSRs than White male patients at every Breslow stage, whereas female patients had similar RSRs across all races or ethnicities. In NM, younger male patients had lower RSRs than younger female patients, but this difference narrowed with age. Asian and Black patients with NM experienced the most impact on survival. Patients with NM and SSM who were married had higher SSRs than patients who were not married. With regard to SSM and LMM, many age groups had RSRs of 100%; younger patients had slight but significant survival reductions. In SSM, Asian male patients had a particularly low RSR (85.9%), while other race–sex subgroups had RSRs >94%, a pattern observed across all Breslow thicknesses.ConclusionsThe impact of melanoma on survival is largest on older male patients with ALM of all races or ethnicities other than White; Asian, Black and older patients with NM; and Black female patients and Asian patients with SSM. White patients with all melanoma subtypes, older patients with SSM and LMM, and younger female patients with ALM and NM were the least affected. Our findings demonstrate the varying survival impact of melanoma on different subpopulations, underscoring the importance of investigating intersectional sociodemographic factors. Melanoma survival varies by subtype, with lentigo maligna melanoma and superficial spreading melanoma having a lower impact on survival, and acral lentiginous melanoma and nodular melanoma having a higher impact on survival. Melanoma survival varies by subtype, with lentigo maligna melanoma and superficial spreading melanoma having a lower impact on survival, and acral lentiginous melanoma and nodular melanoma having a higher impact on survival. Melanoma survival is also known to differ across certain factors, with male patients and patients with skin of colour often experiencing worse outcomes, but not much is known about how the intersection of these demographic factors affect outcomes. Melanoma survival is also known to differ across certain factors, with male patients and patients with skin of colour often experiencing worse outcomes, but not much is known about how the intersection of these demographic factors affect outcomes. Most studies assessing survival in patients with melanoma have employed either an overall survival or a cause-specific survival analysis. Most studies assessing survival in patients with melanoma have employed either an overall survival or a cause-specific survival analysis. This large database study uses relative survival analysis, which has not been thoroughly utilized, to reveal how age, sex and race intersect to influence survival in patients with melanoma while controlling for Breslow depth. This large database study uses relative survival analysis, which has not been thoroughly utilized, to reveal how age, sex and race intersect to influence survival in patients with melanoma while controlling for Breslow depth. Relative survival analysis has the advantage of isolating the effect of melanoma on survival and not relying on accurate cause-of-death coding, a problem that often affects other large database studies. Relative survival analysis has the advantage of isolating the effect of melanoma on survival and not relying on accurate cause-of-death coding, a problem that often affects other large database studies. Additionally, this study looks at the intersection of various demographic factors rather than in isolation of each other, allowing us to identify high-risk groups, such as Asian male patients with superficial spreading melanoma, that are not obvious when looking at one characteristic alone but who may be the ones to benefit most from targeted interventions. Additionally, this study looks at the intersection of various demographic factors rather than in isolation of each other, allowing us to identify high-risk groups, such as Asian male patients with superficial spreading melanoma, that are not obvious when looking at one characteristic alone but who may be the ones to benefit most from targeted interventions. Melanoma is among the most lethal forms of skin cancer, and its incidence is rising.1,2 Although only representing ∼1% of skin cancers diagnosed in the USA, melanoma causes ∼8000 skin ­cancer-related deaths annually.2,3 Early detection is crucial, as shown in countries such as Iceland, Germany and Australia, where sun-safety campaigns and detection programmes have significantly decreased melanoma rates.4–7 Differences in survival based on sociodemographic factors have been observed in melanoma.8–10 Studies have demonstrated that patients with skin of colour have higher mortality, potentially leading to thicker melanomas at diagnosis.11,12 While survival outcomes have been explored, relative survival analyses remain underutilized in melanoma research. Relative survival analysis was developed by Ederer et al. in the mid-1900s.13 Hakulinen further enhanced the method to adjust for censoring.14 Relative survival rates (RSRs) are advantageous because they compare cancer survival with expected survival of demographic-matched individuals without cancer, helping to mitigate other mortality factors not related to cancer.15–17 They circumvent reliance on potentially misclassified cause-of-death coding, an issue seen in ­cause-specific survival analyses. RSRs account for background mortality and more accurately estimate excess mortality attributable to the disease of interest.18,19 Melanoma is classified into several histopathological subtypes: acral lentiginous melanoma (ALM), nodular melanoma (NM), superficial spreading melanoma (SSM) and lentigo maligna melanoma (LMM), each with variable prognoses.20–22 ALM develops on non-sun-exposed sites such as nails and extremities, leading to later-stage diagnoses with thicker presentations.23–25 NM is characterized by early vertical growth and increased Breslow thickness.26,27 SSM, the most prevalent subtype, is largely confined to the epidermis and often <1 mm thick.28–30 LMM is slow-growing but invasive, developing in chronically sun-damaged skin.30–35 NM possesses the lowest 5-year survival, followed by ALM, SSM and then LMM.20,36,37 ALM incidence is highest in Hispanic populations, while NM, SSM and LMM are all predominantly found in White male patients.26–28,31,38,39 ALM is also the most common melanoma subtype found in Black patients.40 To our knowledge, the RSRs of patients with melanoma stratified by intersecting sociodemographic characteristics have not been characterized. We conducted a population-based study investigating RSRs of melanoma subtypes by age, sex, race, marital status and Breslow thickness to analyse differential survival among subpopulations and identify potential areas for intervention. Patients and methods All patients in Surveillance, Epidemiology, and End Results (SEER) diagnosed with ALM, NM, SSM or LMM from 2000 to 2021 based on International Classification of Diseases for Oncology, 3rd Edition codes were initially included.19,41 Survival data, age at diagnosis (15–44, 45–54, 55–64, 65–74, ≥75 years), race or ethnicity [non-­Hispanic (NH) White, NH Black, NH Asian American/Pacific Islander (AAPI), American Indian, Hispanic], Breslow stage (T1, <1.0 mm; T2, 1.0–2.0 mm; T3, 2.0–4.0 mm; T4, >4.0 mm),42 sex and marital status (never married, married, divorced, widowed) were recorded for each patient. Five-year RSRs and 95% log–log confidence intervals (CIs) were calculated by age, race and sex, as well as by the intersecting subgroups of these characteristics. The RSRs and CIs were calculated using the Hakulinen method, SEER*Stat software and expected survival tables provided by SEER. To combat confounding biases from age distribution, age-standardization was conducted using the International Cancer Survival Standard 2.14,43 Subgroup RSRs stratified by Breslow thickness were additionally calculated. Breslow thickness was recoded in SEER beginning in 2010, so this subanalysis was conducted on patients diagnosed between 2010 and 2021. Due to limited sample sizes in multiple race and age subgroups, stratified Breslow analyses were not conducted in patients with LMM. Patients with unknown demographic or survival data were excluded from relevant RSR analyses. Results A total of 144 696 cases of melanoma were analysed, including 3157 ALM, 21 928 NM, 100 060 SSM and 19 551 LMM. Demographic data for each melanoma subtype are summarized in Table 1. The overall 5-year relative survival differed significantly from expected survival for all melanoma subtypes except LMM, with NM having the lowest RSR (RSR 73.5%, CI 72.7–74.3), followed by ALM (81.8%, CI 79.9–83.6) and SSM (99.0%, CI 98.7–99.3) (Figure 1). Overall, female patients had higher RSRs than male patients. RSRs declined with age in ALM and NM but improved with age in LMM and SSM. As expected, RSRs generally worsened with increased Breslow thickness across all subtypes.44,45 Table 1Demographics of all patients included in study, split by melanoma subtype, including sex, race, age at diagnosis and Breslow depth distributionNM (n = 21 928)ALM (n = 3157)SSM (n = 100 060)LMM (n = 19 551)Total (n = 144 696)Sex Female8626 (39.3)1724 (54.6)47 669 (47.6)6660 (34.1)64 679 (44.7) Male13 302 (60.7)1433 (45.4)52 391 (52.4)12 891 (65.9)80 017 (55.3)Race/ethnicity NH White20 220 (92.2)2054 (65.1)90 141 (90.1)18 062 (92.4)130 477 (90.2) NH Black144 (0.7)233 (7.4)184 (0.2)17 (0.1)578 (0.4) NH AAPI211 (1.0)243 (7.7)469 (0.5)70 (0.4)993 (0.7) American Indian59 (0.3)17 (0.5)232 (0.2)33 (0.2)341 (0.2) Hispanic1048 (4.8)564 (17.9)2991 (3.0)352 (1.8)4955 (3.4) Other/unknown246 (1.1)46 (1.5)6043 (6.0)1017 (5.2)7352 (5.1)Age at diagnosis (years) 15–442876 (13.1)421 (13.3)23 632 (23.6)709 (3.6)27 638 (19.1) 45–543392 (15.5)511 (16.2)20 566 (20.6)1908 (9.8)26 377 (18.2) 55–644836 (22.1)686 (21.7)23 803 (23.8)4338 (22.2)33 664 (23.3) 65–744734 (21.6)729 (23.1)18 847 (18.8)6124 (31.3)30 434 (21.0) ≥756090 (27.8)810 (25.7)13 212 (13.2)6472 (33.1)26 584 (18.4)Marital status Married11 835 (54.0)1720 (54.5)47 868 (47.8)8189 (41.9)69 612 (48.1) Never married3408 (15.5)456 (14.4)11 326 (11.3)1122 (5.7)16 312 (11.3) Divorced1702 (7.8)226 (7.2)4900 (4.9)672 (3.4)7500 (5.2) Widowed2277 (10.4)341 (10.8)3921 (3.9)1241 (6.3)7780 (5.4) Other/unknown2706 (12.3)414 (13.1)32 045 (32.0)8327 (42.6)43 492 (30.0)Breslow depth (mm) T1 (<1.0)1350 (6.2)657 (20.8)48 723 (48.7)10 615 (54.3)61 345 (42.4) T2 (1.0–2.0)3051 (13.9)433 (13.7)10 093 (10.1)939 (4.8)14 516 (10.0) T3 (2.0–4.0)4369 (19.9)489 (15.5)3469 (3.5)348 (1.8)8675 (6.0) T4 (>4.0)5382 (24.5)447 (14.2)1428 (1.4)177 (0.9)7434 (5.1) Unavailablea7776 (35.5)1131 (35.8)36 347 (36.3)7472 (38.2)52 726 (36.4)Data are presented as n (%). AAPI, Asian Americans and Pacific Islanders; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NH, non-Hispanic; NM, nodular melanoma; SSM, superficial spreading melanoma. aBreslow depth was recoded in Surveillance, Epidemiology, and End Results (SEER) starting in 2010, so our Breslow depth analysis was only conducted on patients diagnosed between 2010 and 2021, with all patients diagnosed prior to 2010 labelled as ‘unavailable’. Demographics of all patients included in study, split by melanoma subtype, including sex, race, age at diagnosis and Breslow depth distribution Data are presented as n (%). AAPI, Asian Americans and Pacific Islanders; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NH, non-Hispanic; NM, nodular melanoma; SSM, superficial spreading melanoma. aBreslow depth was recoded in Surveillance, Epidemiology, and End Results (SEER) starting in 2010, so our Breslow depth analysis was only conducted on patients diagnosed between 2010 and 2021, with all patients diagnosed prior to 2010 labelled as ‘unavailable’. Relative 5-year survival (%) by melanoma subtype [nodular melanoma (NM), acral lentiginous melanoma (ALM), superficial spreading melanoma (SSM), lentigo maligna melanoma (LMM)] and by sex, with 95% log–log confidence intervals (CIs). SEER*Stat software does not calculate CIs in instances where relative survival is >100%. *A significant difference in survival from those without the disease. When analysing ALM survival stratified by sex and age, RSRs generally declined with increasing age in male and female patients, with RSRs plateauing for those aged >55 years. Male patients of nearly all ages had RSRs lower than or comparable to those of older female patients; men aged 45–54 years had an RSR of 79.2% (CI 72.0–84.8), and women aged >75 years had an RSR of 79.2% (CI 70.0–85.0) (Figure 2). When stratifying by age and Breslow thickness, older age groups had higher proportions of higher-stage tumours (P < 0.05). At every Breslow stage, male patients generally had worse RSRs than female patients, and RSRs were generally lower in the three older age groups for both sexes (Table 2). Looking at marital status in patients with ALM, unmarried and married patients had similar RSRs [82.3% (CI 76.2–87.0) and 82.3% (CI 79.7–84.6), respectively]. However, assessing by sex and marital status, unmarried male patients had worse RSRs than married male patients (68.9% vs. 77.8%, respectively), while unmarried female patients had slightly higher RSRs than married female patients (90.0% vs. 86.2%, respectively) (Figure 3). Relative 5-year survival (%) and 95% log–log confidence intervals (CIs) for each melanoma subtype [nodular melanoma (NM), acral lentiginous melanoma (ALM), superficial spreading melanoma (SSM), lentigo maligna melanoma (LMM)], split by age group and sex. SEER*Stat software does not calculate CIs in instances where relative survival range extends beyond 100%. *Significant difference in survival from subgroup counterparts without the disease. Table 2Relative 5-year survival rates and 95% log–log confidence intervals for each melanoma subtypeALMAge at diagnosis (years)T1T2T3T415–4498.60 (88.30–99.80)96.00 (80.30–99.20)84.40 (64.30–93.70)56.40 (31.30–75.40)45–5498.80 (72.70–100.00)85.30 (70.00–93.20)76.90 (60.50–87.20)61.90 (45.20–74.90)55–6495.40 (83.50–98.80)81.90 (69.00–89.80)79.20 (66.60–87.50)44.00 (31.50–55.80)65–7489.10 (75.40–95.40)81.70 (66.20–90.60)73.60 (60.50–82.90)56.40 (41.20–69.10)≥7594.40 (49.50–99.50)89.90 (50.90–98.30)72.30 (53.90–84.40)47.90 (32.60–61.80)Race/ethnicity NH White99.00 (61.20–100.00)89.90 (53.00–94.70)79.90 (71.30–86.20)54.10 (44.10–63.00) NH Black76.90 (44.40–91.80)45.50 (16.70–70.70)74.80 (42.60–90.60)28.50 (7.50–54.30) NH AAPI84.10 (53.40–95.30)88.60 (53.00–97.70)66.90 (43.30–82.40)40.60 (22.30–58.30) American Indian++++ Hispanic81.80 (65.50–90.90)83.00 (64.40–92.40)66.20 (49.40–78.60)52.30 (38.60–64.40) NM Age at diagnosis (years)T1T2T3T4 15–4489.10 (83.00–93.10)92.50 (88.60–95.10)85.70 (81.60–88.90)69.70 (64.50–74.30) 45–5489.10 (82.90–93.10)93.40 (89.80–95.70)81.90 (77.60–85.40)60.20 (55.50–64.50) 55–6489.10 (83.30–92.90)90.70 (87.30–93.30)81.90 (78.50–84.80)60.50 (56.90–63.90) 65–7488.40 (80.60–93.20)90.50 (85.90–93.70)78.10 (73.90–81.60)57.60 (53.60–61.30) ≥7570.30 (58.30–79.40)85.20 (76.10–91.00)74.00 (68.10–79.00)45.70 (41.20–50.00)Race/ethnicity NH White86.40 (83.00–89.20)90.60 (88.40–92.40)80.10 (78.00–82.00)56.90 (54.80–58.90) NH Black++68.40 (31.90–88.10)63.60 (43.30–78.30) NH AAPI++81.20 (53.50–93.30)39.90 (22.40–56.90) American Indian++++ Hispanic77.00 (58.40–88.00)85.60 (73.30–92.50)68.50 (58.20–76.70)59.10 (51.20–66.20) SSM Age at diagnosis (years)T1T2T3T4 15–4499.80 (99.40–100.00)97.10 (95.70–98.00)82.90 (78.20–86.60)75.10 (65.20–82.50) 45–54100.8096.60 (95.10–97.70)87.40 (83.10–90.70)69.20 (60.40–76.40) 55–64101.5096.10 (94.40–97.20)84.40 (80.60–87.50)65.30 (58.30–71.50) 65–74103.8098.50 (94.80–99.60)83.40 (78.80–87.10)59.60 (51.30–67.00) ≥75107.9098.00 (82.80–99.80)76.30 (68.50–82.40)60.80 (49.50–70.40)Race/ethnicity NH White101.7096.80 (95.80–97.60)83.30 (81.00–85.30)65.10 (60.80–69.00) NH Black98.30 (0.00–100.00)86.00 (39.80–97.50)79.30 (0.10–99.30)24.60 (3.80–54.80) NH AAPI98.40 (83.00–99.90)89.20 (69.20–96.50)52.00 (23.20–74.60)42.50 (14.00–68.90) American Indian++++ Hispanic99.60 (83.30–100.00)100.80%80.10 (68.00–88.00)74.20 (58.70–84.60)Data are presented as percentage (confidence interval). Acral lentiginous melanoma (ALM), nodular melanoma (NM) and superficial spreading melanoma (SSM), split by age group and Breslow thickness, as well as by race or ethnicity and Breslow thickness. A ‘+’ indicates subgroups where there was insufficient data or too small of a sample size to calculate an RSR. SEER*Stat software does not calculate CIs in instances where relative survival is >100% or 0%. AAPI, Asian Americans and Pacific Islanders; NH, non-Hispanic. Relative 5-year survival rates and 95% log–log confidence intervals for each melanoma subtype Data are presented as percentage (confidence interval). Acral lentiginous melanoma (ALM), nodular melanoma (NM) and superficial spreading melanoma (SSM), split by age group and Breslow thickness, as well as by race or ethnicity and Breslow thickness. A ‘+’ indicates subgroups where there was insufficient data or too small of a sample size to calculate an RSR. SEER*Stat software does not calculate CIs in instances where relative survival is >100% or 0%. AAPI, Asian Americans and Pacific Islanders; NH, non-Hispanic. Relative 5-year survival (%) and 95% log–log confidence intervals (CIs) for each melanoma subtype [nodular melanoma (NM), acral lentiginous melanoma (ALM), superficial spreading melanoma (SSM), lentigo maligna melanoma (LMM)], split by marital status and sex. SEER*Stat software does not calculate CIs in instances where relative survival range extends beyond 100%. A ‘+’ indicates lack of age-standardization by SEER*Stat software due to insufficient data. *Significant difference in survival from subgroup counterparts without the disease. When examining the intersection of sex and race in ALM, RSRs among female patients were relatively consistent across racial groups, with the lowest being in Black female patients (81.4%, CI 71.5–88.1) and the highest being in White female patients (88.7%, CI 85.8–91.1). With regard to male patients, Black patients demonstrated the lowest RSR (58.3%, CI 38.7–73.5), followed by Hispanic and Asian male patients, whereas White male patients had an RSR of 81.4% (CI 77.5–84.6) (Figure 4). Regarding Breslow depth, a larger proportion of White patients had stage T1 tumours compared with other racial groups (P < 0.05). Overall, White patients had the highest RSRs at every Breslow stage (Table 2). White male patients had higher RSRs at each Breslow thickness compared with male patients with skin of colour, whereas White female patients had relatively more similar RSRs compared with female patients with skin of colour, except for stage T4, where Black and Asian female patients had RSRs roughly 20% points lower than White and Hispanic female patients. Relative 5-year survival (%) and 95% log–log confidence intervals (CIs) for each melanoma subtype [nodular melanoma (NM), acral lentiginous melanoma (ALM), superficial spreading melanoma (SSM), lentigo maligna melanoma (LMM)], split by race/ethnicity and sex. SEER*Stat software does not calculate CIs in instances where relative survival range extends beyond 100%. A ‘+’ indicates lack of age-standardization by SEER*Stat software due to insufficient data. *Significant difference in survival from subgroup counterparts without the disease. When exploring subgroups of patients with NM, the RSR gap between male and female patients for the 15–44 years age group was 12.2% (favouring female patients), while the gap between male and female patients in the ≥75 years age group was 1% (Figure 2). When additionally stratifying by Breslow thickness, this pattern was observed in stage T3 and T4, with the RSR gap between female and male patients decreasing with age. This was also seen in stage T1 in all age groups up to the ≥75 years group, where the RSR difference widened to 20.1% (Table 2). Exploring by marital status, married patients had the highest RSR (77.8%, CI 76.8–78.8), with all other groups having RSRs <66%. This pattern held when split by sex as well, with married male and married female patients having the highest RSRs among their respective groups (Figure 3). Regarding race, Hispanic, Asian and Black patients had the lowest RSRs, regardless of sex (Figure 4). The lowest RSRs were found in Asian female patients (64.8%, CI 52.6–74.5), Asian male patients (59.7%, CI 47.5–70.0) and Black male patients (57.2%, CI 40.8–70.5), while the highest RSRs were found in White female patients (78.6%, CI 77.4–79.4) and White patients (70.3%, CI 69.2–71.4). When stratifying by Breslow depth, Asian, Black and Hispanic patients all had lower RSRs than White patients at stages T1 (ranging from 54.7% to 86.4%) and T2 (77.1% to 90.6%) (Table 2). In stage T3, Black and Hispanic patients had the lowest RSRs, while Asian patients had similar RSRs to White patients. Conversely, at stage T4, Black and Hispanic patients had higher RSRs compared with White patients (63.6% and 59.1% vs. 56.9%, respectively), while Asian had the lowest RSR (39.9%, CI 22.4–56.9). Regarding SSM, RSRs were significantly reduced for younger patients (age groups 15–44, 45–54 and 55–64 years) and all racial groups except American Indian (Figure 2). While these survival differences were considered significant compared with expected rates, most RSRs had high absolute rates, often above 95%. When broken down by sex and age, male patients had slightly lower survival rates than female patients in every age group. Examining Breslow thickness, younger age groups had lower proportions of T3/T4 (P < 0.05). In male and female patients, as age increased, RSRs improved in T1, stayed similar in T2, and increased and then decreased in T3. In T4, the RSRs of male patients increased and decreased with age, while the RSRs of female patients consistently worsened with age. In both sexes, almost every age group had significant RSRs throughout stages T2–T4, declining to the 50–70% range by the latest stage (Table 2). By marital status, married patients had the highest RSR (99.1%, CI 98.7–99.4), with this pattern holding when analysing male and female patients separately, although all groups had RSRs >94% (Figure 3). Stratifying by race and sex revealed relative survival decreases in Asian male patients (85.9%, CI 78.3–91.0), Asian female patients (88.7%, CI 80.8–93.5) and Black female patients (88.6%, CI 76.7–94.6), while White male and female patients had RSRs of 98.1% (CI 97.7–98.4) and 99.6% (CI 99.0–99.9), respectively (Figure 4). When incorporating Breslow thickness, White patients had greater proportions of T1 cases (P < 0.05), but Asian and Black patients had lower RSRs at every stage, with Asian patients having the only significant RSR in T1 (98.4%, CI 83.3–99.9). At higher stages, the difference in survival impact between White and Asian patients was >20% (Table 2). While no significant RSR was found for patients with LMM overall, examining both race and sex revealed significant survival impacts in Black female patients (86.3%, CI 20.6–98.6), Hispanic male patients (98.0%, CI 61.9–99.9) and American Indian male patients (88.2%, CI 17.6–99.1) (Figure 4). Discussion This study identified differential survival impacts of melanoma on patients with various demographics, highlighting the complex interplay of age, race and sex in shaping prognosis. Relative survival is particularly useful for evaluating survival disparities across demographic subgroups, allowing for the assessment of how disease burden differs by age, sex and race without being confounded by differences in overall mortality risks.46 This analysis has been applied elsewhere, such as with the exploration of various cancer survival rates in Iceland, Germany and the Netherlands.47–49 Moreover, relative survival estimates are not dependent on the accuracy of death certificate data, making them robust for analysing cancer registry datasets like SEER. This methodological strength is particularly relevant for malignancies like melanoma, where survival varies significantly across histological subtypes and demographic factors, as demonstrated here. Overall, female patients exhibited higher RSRs across all melanoma subtypes and age groups except one. Additionally, in ALM, although RSRs worsened with age, male patients aged 15–44 years exhibited comparable RSRs to female patients aged ≥75 years. This finding is consistent with prior literature suggesting that male patients have worse survival outcomes, which could potentially be related to hormonal influences on immune response and healthcare-seeking behaviour differences leading to more advanced tumours on diagnosis.50–55 Regarding marital status, married and single patients had similar RSRs when looking at all patients together, but married male patients had higher RSRs than unmarried male patients. This result is consistent with prior research finding that marriage might serve as a more protective role on survival in male patients than in female patients.56 One especially notable finding was that males with skin of colour, particularly Black male patients, had significantly lower RSRs than their White counterparts. In contrast, racial differences in survival were less pronounced among female patients, suggesting a nuanced interaction between race and sex in ALM prognosis. These disparities may reflect differences in healthcare access, tumour biology variations or detection delays, which disproportionately impact on male patients with skin colour.52–55 Given that ALM is frequently diagnosed at advanced stages, these findings highlight the urgent need for targeted outreach and earlier detection strategies in this particularly high-risk population.57 In NM, sex-based relative survival differences were most pronounced in younger patients, with male patients aged 15–44 years having significantly lower RSRs than their female counterparts. However, this gap narrowed with increasing age, with women aged ≥75 years even having slightly lower RSRs than men of the same age, suggesting that sex-based survival differences may be more relevant in early disease progression.58 Additionally, differences by marital status were consistent across sexes, with married patients having higher RSRs. This finding is in agreement with previous studies demonstrating that married patients with cancer have significantly higher survival rates than unmarried patients with cancer.59,60 Moreover, racial disparities were also consistent across sexes, with the minority of patients exhibiting the lowest RSRs. Black male patients, in particular, had a significantly reduced RSR, reinforcing prior research suggesting that melanoma in Black patients is often diagnosed at later stages.61 These findings emphasize the need for heightened awareness and improved early detection, particularly for racial minority groups. SSM was associated with high overall survival, with most RSRs exceeding 95%, although younger patients and patients in minority ethnic groups exhibited significant survival differences. Specifically, Asian male patients, Asian female patients and Black female patients had significantly lower RSRs than their White counterparts, potentially related to disparities in disease progression, access to dermatological care, and/or treatment response.62,63 Interestingly, Black male patients had comparable survival to White male patients, whereas there was a survival disadvantage in Black female patients relative to White female patients, highlighting survival differences on an intersectional level. This pattern contrasts with the trends in NM and ALM, implying that sex-related survival disparities may vary by ­melanoma subtype. These findings also challenge blanket assumptions regarding sex- and race-based survival advantages, ­reinforcing the need for tailored prognostic models. They highlight the disproportionately poor survival outcomes for Asian male and female patients, a population not traditionally recognized as high-risk for SSM-related mortality but who may benefit from closer clinical monitoring. Additionally, the fact that RSRs improved with age highlights that the notion of melanoma outcomes worsening with age cannot be applied universally.64 The reasoning behind why these survival reductions are smaller but still significant could be due to the fact that SSM is a less aggressive tumour compared with other melanoma subtypes, thus leading to slower disease progression while still contributing to excess mortality over time.9,65 In terms of marital status, patients with SSM demonstrated the same pattern as patients with NM, with married patients having higher RSRs across sexes, similarly highlighting how patients without spousal support have worse cancer outcomes. Overall survival differences in LMM were not statistically significant. While relative survival was quite favourable on an absolute scale, Black female patients, Hispanic male patients and American Indian male patients experienced lower RSRs. These results challenge the assumption that female patients universally exhibit better melanoma outcomes compared with males and highlight the necessity of analysing race–sex intersections to fully elucidate survival disparities. However, the minority ethnic group sample sizes in the LMM cohort were relatively small, and additional studies with adequate representation are needed to validate these observations. The impact of Breslow thickness on relative survival varied among the melanoma subtypes. While Breslow stage has been considered the most important prognostic factor in melanoma, its relationship with survival may not be as clear as previously thought.66,67 Consistent with previous literature, in ALM, we see the same trend of decreasing RSRs with increased age, as well as an increased proportion of stage T3/T4 cases in older patients.68,69 Breslow thickness may thus account for some of the disparity in RSRs between age groups, but the finding that RSRs were worse in older patients across all Breslow stages demonstrates that there are also other contributory factors. In NM, we see female patients having notably higher RSRs than male patients at younger ages in stages T3 and T4, but less so in stages T1 and T2, indicating that being male may be associated with worse NM survival at increased Breslow thicknesses but less so in lower-stage disease. In SSM, we see that RSRs generally improve with age in stages T1 and T2, but worsen with age in T3 and T4, highlighting a difference in trend based on Breslow thickness. Younger patients having worse relative survival overall despite a lower proportion of higher-stage tumours, indicates that other factors besides Breslow staging may contribute to these differences in survival. For instance, it is possible that as older adults face other serious health complications, having a low-stage, slowly progressing SSM diagnosis may not affect their relative survival as much as it would healthier younger populations. Moreover, the switch in survival trends between thinner and thicker SSMs demonstrates the interplay of tumour characteristics and patient demographics and the importance of considering both when assessing mortality risk in different populations. RSR differences by race held more consistently when stratified by Breslow thickness as compared with age. In ALM, patients with skin of colour consistently had worse RSRs than White patients at every Breslow stage. Although White patients had larger proportions of T1 tumours, potentially contributing to their overall higher relative survival, these findings indicate that other factors are also associated with patients with skin of colour experiencing more of a survival impact from ALM. When stratifying patients with NM by Breslow thickness, Hispanic patients had worse RSRs than White patients in both T1 and T2 (with insufficient sample sizes for Asian and Black patients). This pattern is more variable in thicker tumours, with Black and Hispanic patients ­having lower RSRs than White patients in T3, and Asian patients having the lowest RSR in T4. These findings suggest that racial differences in survival among patients with NM are heterogeneous and may change across higher Breslow categories. Moreover, although White patients with SSM had higher proportions of T1 tumours, which increases their overall RSRs, Asian and Black patients had worse RSRs than White patients at every Breslow stage, highlighting that Breslow depth cannot completely explain relative survival differences among races. The primary limitation of this study is its retrospective design, which is inherent to SEER analyses. Another limitation is that because relative survival compares observed mortality with expected mortality in a matched general population, RSRs near 100% in SSM and LMM may reflect an absence of detectable excess mortality at the population level rather than a complete lack of melanoma-related deaths, potentially masking ­disease-specific mortality. Additionally, the sample sizes for racial minority groups were smaller than their White counterparts, ­particularly in patients with LMM, and Breslow depth values were also only available for patients diagnosed from 2010 onwards. Data on marital status was additionally labelled as other or unknown for a sizeable portion of patients; thus, the findings should be interpreted with caution. The sample size for American Indian patients was especially low, and given the frequent misclassification of American Indians in healthcare, further research must be conducted to better understand melanoma survival in this group.70,71 Furthermore, select race–sex and marital status–sex subgroups could not be age-standardized with the SEER*Stat software due to insufficient data, slightly affecting a few RSR values (Figure 3). Finally, this analysis is based solely on data from patients from the USA, so further research is needed to determine whether these findings are globally generalizable. Our study highlights notable differences in impact on survival of melanoma subtypes based on age, sex, race and marital status, and their intersections. ALM and NM exhibited the most pronounced survival differences, particularly in certain race–sex combinations, while SSM and LMM had relatively high survival rates overall but displayed significant disparities in younger age groups and certain race–sex subgroups. Overall, the results of this study help to identify specific sociodemographic populations with markedly reduced survival. Clinicians and researchers should be encouraged to examine the crossover of multiple demographic characteristics, as opposed to exploring these factors in isolation. Moreover, these findings emphasize the need for public health strategies tailored to specific groups, including efforts at earlier detection, improved access to specialized care and increased awareness for high-risk populations. Further research is warranted to explore the potential biologic, socioeconomic and healthcare-related contributors underlying these disparities, as well as to develop interventions aimed at improving melanoma outcomes across diverse patient populations. Contributor Information Benjamin D Hu, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Brandon R Block, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Grace Rabinowitz, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Hannah Verma, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Raphaella Lambert, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Bragi Adalsteinsson, Department of Medicine, University of Iceland, Reykjavik, Iceland. Jón Gunnlaugur Jónasson, Faculty of Medicine, University of Iceland, Reykjavik, Iceland; Department of Pathology, Landspitali – the National University Hospital of Iceland, Reykjavik, Iceland. Bardur Sigurgeirsson, Faculty of Medicine, University of Iceland, Reykjavik, Iceland. Shane A Meehan, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Jonathan Ungar, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Jesse M Lewin, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Nicholas Gulati, Kimberly and Eric J. Waldman Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, USA. Jonas A Adalsteinsson, Faculty of Medicine, University of Iceland, Reykjavik, Iceland; Department of Pathology, Landspitali – the National University Hospital of Iceland, Reykjavik, Iceland. Author contributions Benjamin D Hu (Conceptualization, Methodology [equal], Formal analysis, Visualization, Writing—original draft [lead]), Brandon R Block (Writing—original draft, Writing—review & editing [supporting]), Grace Rabinowitz (Writing—original draft, Writing—review & editing [supporting]), Hannah Verma (Writing—original draft, Writing—review & editing [supporting]), Raphaella Lambert (Writing—original draft, Writing—review & editing [supporting]), Bragi Adalsteinsson (Writing—review & editing [supporting]), Jón Gunnlaugur Jonasson (Writing—review & editing [supporting]), Bardur Sigurgeirsson (Writing—review & editing [supporting]), Shane A Meehan (Writing—review & editing [supporting]), Jonathan Ungar (Writing—review & editing [supporting]), Jesse M Lewin (Writing—review & editing [supporting]), Nicholas Gulati (Writing—review & editing [supporting]), and Jonas A Adalsteinsson (Conceptualization [lead], Methodology, Writing—review & editing [equal]) Funding sources This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Conflicts of interest The authors declare no conflicts of interest. Data availability All data are accessible in the Surveillance, Epidemiology, and End Results (SEER) database. Ethics statement Exempt from review. Patient consent Not applicable. References Associated Data Data Availability Statement All data are accessible in the Surveillance, Epidemiology, and End Results (SEER) database.