In total, 199 patients who met eligibility criteria were included in the study (mean age 59.1 years, 64.3% [n = 128] male, 35.7% female [n = 71], 72.9% white ethnicity [n = 145], 1.5% Asian ethnicity [n  = 3], 0.5% Black ethnicity [n = 1] and 25.1% unknown ethnicity [n = 50]). Of these, 162 (81.4%) patients were being treated for a/mMM, and 37 (18.6%) were being treated for aRCC (Table 1). Median follow-up from index was 14.5 months (IQR: 7.2–26.2). In patients with a/mMM, metastases were present in 99.4% (n = 161), with lung and lymph node being the most frequent sites (53.7% [n = 87] and 52.5% [n = 85], respectively). The majority of patients had an Eastern Cooperative Oncology Group (ECOG) performance score of either 0 (58.6%, n = 95) or 1 (40.1%, n = 65), received ICI therapy as first-line treatment (63.6%, n = 103) and were initiated on NIVO + IPI (78.4%, n = 127). NIVO monotherapy or IPI monotherapy were received by 11.1% (n = 18) and 10.5% (n = 17) of patients, respectively. In patients with aRCC, metastases were present in 91.9% (n = 34) of patients, and the lung was the most frequent site (55.9%, n = 19). The largest proportion of patients had an ECOG performance score of 1 (59.5%, n = 22), and received ICI therapy as either first-line treatment (43.2%, n = 16) or second-line treatment (35.1%, n = 13). In total, 54.1% (n = 20) of patients with aRCC received NIVO monotherapy, while the remaining 45.9% (n = 17) received NIVO + IPI.

Among patients with a/mMM, 75.3% (n = 122) reported an irAE, with 45.9% (n = 56) of these patients reporting a grade 3 or 4 irAE (Table 2). When stratified by ICI therapy received, irAEs were reported by 77.8% (n = 14) of patients receiving NIVO monotherapy, 47.1% (n = 8) receiving IPI monotherapy and 78.7% (n = 100) receiving NIVO + IPI. Grade 3/4 irAEs were reported by 5.6% (n = 1), 11.8% (n = 2) and 41.7% (n = 53) of patients receiving NIVO monotherapy, IPI monotherapy or NIVO + IPI, respectively. More than one irAE was reported by 38.9% (n = 7) of patients receiving NIVO monotherapy, 17.6% (n = 3) receiving IPI monotherapy and 59.8% (n = 76) receiving NIVO + IPI during the study period. When further stratified by irAE type, colitis/diarrhoea and skin reactions were the most frequently reported irAE types across all ICI therapy groups: NIVO monotherapy (44.4% [n = 8] and 22.2% [n = 4], respectively), IPI monotherapy (35.3% [n = 6] and 17.6% [n = 3], respectively) and NIVO + IPI (42.5% [n = 54] and 37.8% [n = 48], respectively; Fig. 1a). Across all patients with a/mMM, independent of treatment received, colitis/diarrhoea, skin reactions, endocrinopathies, rheumatological and neurological toxicities and pneumonitis irAEs were more commonly classified as grade 1/2, whereas nephritis/renal dysfunction was more commonly classified as grade 3/4 (Supplementary Material Online Resource Table S1).

Incidence of irAEs based on the most frequently reported (≥ 3%) irAE types within the a a/mMM and b aRCC cohorts. Data are stratified by ICI treatment and reported as a proportion of patients in that ICI treatment group. Counts correspond to the number of patients experiencing an irAE in that ICI treatment group. a/mMM advanced (unresectable or metastatic) malignant melanoma, aRCC advanced renal cell carcinoma, ICI immune checkpoint inhibitor, irAE immune-related adverse event

A total of 62.2% (n = 23) of patients with aRCC reported an irAE; 30.4% (n = 7) of those reported were grade 3/4. When stratified by ICI therapy received, irAEs were reported by 60.0% (n = 12) of patients receiving NIVO monotherapy and 64.7% (n = 11) of patients receiving NIVO + IPI, with grade 3/4 irAEs reported by 15.0% (n = 3) and 23.5% (n = 4) of patients receiving NIVO monotherapy or NIVO + IPI, respectively. More than one irAE was reported by 35.0% (n = 7) of patients receiving NIVO monotherapy and 35.3% (n = 6) receiving NIVO + IPI. When further stratified by irAE type, patients receiving NIVO monotherapy most frequently reported skin reactions (20.0%, n = 4), and those receiving NIVO + IPI most frequently reported skin reactions and hepatitis, with both reporting 35.3% (n = 6; Fig. 1b). In patients with aRCC, independent of treatment received, all irAEs were more commonly graded as 1/2 than as grade 3/4 (Supplementary Material Online Resource Table S1).

In patients with a/mMM, the median time to first irAE onset was 4.9 weeks (IQR: 2.9–9.4; Table 2). Median time to onset for grade 3/4 irAEs was 7.7 weeks (IQR: 3.4–13.1). When stratified by ICI therapy received, median time to onset was 12.6 weeks (IQR: 5.4–25.4) for patients receiving NIVO monotherapy, 4.2 weeks (IQR: 2.8–7.5) for those receiving IPI monotherapy and 4.6 weeks (IQR: 2.8–9.1) for those receiving NIVO + IPI. When further stratified by irAE type, the shortest median time to irAE onset was reported for skin reactions across all ICI therapy groups: NIVO monotherapy (3.1 weeks [IQR: 2.5–10.6]), IPI monotherapy (3.0 weeks [IQR: 1.8–3.9]) and NIVO + IPI (3.1 weeks [IQR: 2.1–10.1]; Fig. 2a). The median time to onset of first irAEs stratified by irAE type is presented in Supplementary Material Online Resource Table S2.

Median time to first irAE onset based on the most frequently reported (≥ 3%) irAE types within the a a/mMM and b aRCC cohorts. Data are stratified by ICI treatment group. a/mMM advanced (unresectable or metastatic) malignant melanoma, aRCC advanced renal cell carcinoma, ICI immune checkpoint inhibitor, irAEs immune-related adverse events

Among patients with aRCC, median time to first irAE onset was 4.9 weeks (IQR: 2.9–9.4), and 7.7 weeks (IQR: 3.4–13.9) for grade 3/4 irAEs. When stratified by ICI therapy received, median time to onset was 13.5 weeks (IQR: 7.9–37.7) for patients receiving NIVO monotherapy, and 3.7 weeks (IQR: 3.0–15.2) for those receiving NIVO + IPI. The irAE type with the shortest reported median time to onset was colitis/diarrhoea for those receiving NIVO monotherapy (7.0 weeks [IQR 4.6–26.7]) and rheumatological toxicity for those receiving NIVO + IPI (0.1 weeks [IQR 0.1–0.1]; Fig. 2b). Median time to onset stratified by irAE type is presented in Supplementary Material Online Resource Table S3.

Among patients with a/mMM reporting irAEs (n = 122), immunomodulating drugs were prescribed to 69.7% (n = 85) of patients, with the largest proportion of these patients receiving corticosteroids alone (81.2%, n = 69; Table 3). Of those patients who developed a grade 3/4 irAE (n = 56), 62.5% (n = 35) were prescribed corticosteroids alone, and 26.7% (n = 15) were prescribed one or more additional immunomodulating drug, including methotrexate (n = 1), mycophenolate (n = 9), tacrolimus (n = 1) or the tumour necrosis factor (TNF) inhibitor infliximab (n = 7). When stratified by the type of irAE reported, patients reporting hepatitis were most frequently prescribed immunomodulating drugs (75.0%, n = 33). The proportion of patients who were prescribed immunomodulating drugs stratified by the type of irAE reported is presented in Supplementary Material Online Resource Table S4.

Of those patients with aRCC reporting any irAE (n = 23), 56.5% (n = 13) were prescribed immunomodulating drugs for irAE management, with 92.3% (n = 12) of these patients receiving corticosteroids alone. Of those patients who developed a grade 3/4 irAE (n = 7), 57.1% (n = 4) were prescribed corticosteroids alone. Data stratified by irAE type is presented in Supplementary Material Online Resource Table S5.

In both the a/mMM and aRCC cohorts, median OS was numerically longer for patients who had experienced an irAE compared with those who did not (Fig. 3). In the a/mMM cohort, the median OS for patients who reported any irAE was 31.1 months (95% confidence interval [CI]: 20.3, not reached), with 12-, 24- and 36-month OS probabilities of 71.4%, 57.5% and 46.2%, respectively (Fig. 3a). For those patients who did not report an irAE, median OS was 14.8 months (95% CI: 10.6, not reached), with 12-, 24- and 36-month OS probabilities of 54.4%, 41.1% and 37.4%, respectively. Median OS in patients who experienced a grade 3/4 irAE was similar to that observed in patients who experienced a grade 1/2 irAE at 39.4 months (95% CI: 20.3, not reached) and 35.6 months (95% CI: 20.0, not reached), respectively (Supplementary Material Online Resource Fig. 1a).

Overall survival stratified by whether an irAE was or was not reported by patients in the a a/mMM and b aRCC cohorts. a/mMM advanced (unresectable or metastatic) malignant melanoma, aRCC advanced renal cell carcinoma, CI confidence interval, irAE immune-related adverse event, NR not reached

In the aRCC cohort, median OS for any reported irAE was 38.4 months (95% CI: 19.1, not reached), with 12-, 24- and 36-month OS probabilities of 80.5%, 56.3% and 56.3%, respectively (Fig. 3b). Where no irAE was reported, median OS was 10.7 months (95% CI: 9.1, not reached), with 12- and 24-month OS probabilities of 49.1% and 24.6%, respectively. Median OS in patients who experienced a grade 1/2 irAE was 38.4 months (95% CI 19.1, not reached) whereas median OS was not reached in patients who experienced a grade 3/4 irAE (Supplementary Material Online Resource Fig. 1b).