Risk and prognosis of second cutaneous melanoma after radiotherapy for breast cancer: A population-based analysis

Radiation therapy (RT), a primary treatment for breast cancer (BC), may be associated with increased non-BC tumor risk. We aimed to examine second cutaneous melanoma (SCM) risk in BC patients who underwent RT and to assess their survival outcomes. Data from 520,977 BC patients diagnosed between 1973–2018 were collected from the Surveillance, Epidemiology, and End Results (SEER) database. Cumulative SCM incidence was estimated using the Fine-Gray competing risk model. Poisson regression analysis was conducted to calculate the standardized incidence ratio (SIR) and estimate the SCM relative risk (RR) in patients who underwent RT compared to those who did not. Overall survival (OS) and cancer-specific survival (CSS) were assessed using the Kaplan–Meier method. Among the 520,977 BC patients, 243,676 (46.8%) underwent surgery and RT, while 277,301 (53.2%) only underwent surgery. Our results suggest that BC patients receiving RT had a higher SCM risk than those who did not (hazard ratio [HR] 1.40; 95% confidence interval [CI] 1.30–1.51; P < 0.001). SCM incidence was also higher in BC patients treated with RT than in the general US population (SIR 1.12; 95% CI 1.05–1.19; P < 0.05). However, SCM patients who received RT had a significantly higher 10-year survival rate than those who did not receive RT (14.90% vs 5.94%; P < 0.001). No significant difference was found in 10-year OS or 5-year CSS between SCM following RT and only primary cutaneous melanoma (OPCM), but SCM patients who did not receive RT had a significantly lower 10-year OS, with no significant difference in CSS. This study suggests an increased SCM likelihood in BC patients due to RT, although the overall risk is minimal.


Table S1 .
Comparison of clinicopathologic baseline characteristics of patients with BC and SCM by treatment modality

Table S2 .
Univariable and multivariable computing risk regression analysis if there is risk of developing SCM in BC patients

Table S3 .
Radiation-attributed risk of SCM subgroup

Table S4 .
Radiation-attributed risk of second SSM subgroup

Table S5 .
Standardized incidence ratio of SCM

Table S6 .
Subgroup analyses of subhazard ratios of developing SCM in BC patients who received RT versus those who did not

Table S7 .
Comparisons of baseline characteristics of SCM between RT group and NRT group

Table S8 .
Propensity score-matched patients who developed SCM after RT and after no RT

Table S9 .
Patients' characteristics of SCM and matched OPCM

Table S10 .
Studies of the risk of developing SCM after radiation therapy among BC patients

Table S1 . Comparison of clinicopathologic baseline characteristics of patients with BC and SCM by treatment modality.
b P value was calculated using X 2 test for categorical variables; Significant P values are in[bold].BC: Breast cancer; SCM: Second cutaneous melanoma.

Table S2 . Univariable and multivariable computing risk regression analysis if there is risk of developing SCM in BC patients. Characteristics Univariable competing risk regression
Gray competing risk regression analyses are used to calculate the HR and 95 CI for SCM in BC patients treated with RT versus patients not treated with RT.Covariables that are significant in univariable competing risk regression analysis (P < 0.05) are included in the multivariable analysis.The other sites of BC include axillary tail of breast, overlapping lesion of breast.Significant P values are in[bold].sHR: subdistribution hazard ratio; CI: Confidence interval; BC: Breast cancer; SCM: Second cutaneous melanoma; RT: Radiation therapy; UO: Upper outer quadrant of breast; UI: Upper inner quadrant of breast; LI: Lower inner quadrant of breast; LO: Lower outer quadrant of breast; CEN: Central portion quadrant of breast; HR: Hazard ratio.CI: Confidence interval.

Table S3 . Radiation-attributed risk of SCM subgroup.
Poisson regression analyses were used to calculate the RR and 95 CIs of SCM for patients with radiation therapy versus patients without radiation therapy.Adjusted RRs were stratified by age at BC diagnosis and calendar year of BC diagnosis.A determination of statistical significance of RRs was based on P < 0.05 (two sided).Significant P values are in[bold].RR: Radiation-attributed risk; BC: Breast cancer; SCM: Second cutaneous melanoma; RR: Radiation-attributed risk; CI: Confidence interval.

Table S4 . Radiation-attributed risk of second SSM subgroup.
Poisson regression analyses were used to calculate the RR and 95 CIs of SSM for patients with radiation therapy versus patients without radiation therapy.Adjusted RRs were stratified by age at BC diagnosis and calendar year of BC diagnosis.A determination of statistical significance of RRs was based on P<0.05 (two sided).Significant P values are in[bold].RR: Radiation-attributed risk; BC: Breast cancer; SSM: Superficial spreading melanoma; CI: Confidence interval.SIR was defined as the ratio of the number of observed SCM cases among BC survivors to the expected number of cases in the US general population and was stratified by age at BC diagnosis, calendar year of BC diagnosis and latency period calendar.A determination of the statistical significance of SIRs was based on a P < 0.05 (two sided).95confidenceintervals were calculated by Poisson exact methods.The background incidence of SCM was derived from data provided by the SEER database.#indicatesP < 0.05.Significant P values are in[bold].BC: Breast cancer; RT: Radiation therapy; NRT: No radiation therapy; SIR: Standardized incidence ratio; SCM: Second cutaneous melanoma.

Table S8 . Propensity score-matched patients who developed SCM after RT and after no RT.
developed second SCM were matched with patients with OPCM at a PSM ratio of 1:5 for other SCM versus OPCM.BC: Breast cancer; RT: Radiation therapy; NRT: No radiation therapy; SCM: Second cutaneous melanoma; OPCM: Only primary cutaneous melanoma; PSM: Propensity score matching.

Table S10 . Studies of the risk of developing SCM after radiation therapy among BC patients.
IGR: Institut gustave roussy; SEER: Surveillance, Epidemiology and End Results program; TCR: Thames cancer registry; IC: Institut curie; CCCN: Comprehensive cancer center-north netherlands; CCCA: Comprehensive cancer center amsterdam; CCCS: Comprehensive cancer center south.Figure S1.Flow diagram.SEER: Surveillance, Epidemiology and End Results; PSM: Propensity score matching; RT: Radiation therapy; BC: Breast cancer; CM: Cutaneous melanoma.