Table 3. Concordance between potentially immune-related AEs events in phase 1 and late-phase trials Question: Were potentially immune-related AEs events seen in both phase and late-phase studies in similar frequencies?

Toxicity	Phase 1	Late-phase trial	P	Odds ratio (95%CI)
Trial level analysis
	n = 10	n = 15
Rash	10 (100%)	13 (87%)	0.50	NA
Pruritus	9 (90%)	12 (80%)	0.63	0.4 (0.04, 5.0)
Vitiligo	4 (40%)	5 (33%)	0.99	0.8 (0.1, 3.9)
Diarrhea	9 (90%)	15 (100%)	0.40	NA
Colitis	1 (10%)	11 (67%)	0.01	18 (1.8, 185)
Hypophysitis	2 (22%)	7 (47%)	0.23	3.5 (0.5, 22.3)
Adrenal insufficiency	0 (0%)	2 (13%)	0.50	NA
Hypothyroidism	7 (70%)	11 (73%)	0.99	1.2 (0.2, 6.9)
Hyperthyroidism	4 (40%)	9 (60%)	0.43	2.3 (0.4, 11.5)
Pneumonitis	7 (70%)	13 (87%)	0.36	2.8 (0.3, 20.8)
Patient level analysis
	n = 1650	n= 4823
Rash	223 (14%)	463 (10%)	0.36	1.0 (0.98, 1.1)
Pruritus	171 (10%)	503 (10%)	0.12	1.1 (0.99, 1.1)
Vitiligo	32 (2%)	134 (3%)	0.30	1.1 (0.95, 1.2)
Diarrhea	150 (9%)	554 (11%)	0.049	1.1 (1.0, 1.2)
Colitis	2 (0%)	41 (1%)	0.045	3.0 (1.02, 9.0)
Hypophysitis	3 (0%)	12 (0%)	0.19	2.1 (0.7, 6.5)
Adrenal insufficiency	0 (0%)	6 (0%)	NA	NA
Hypothyroidism	62 (4%)	240 (5%)	0.16	1.1 (0.98, 1.2)
Hyperthyroidism	17 (1%)	111 (2%)	0.15	1.2 (0.9, 1.5)
Pneumonitis	42 (3%)	114 (2%)	0.26	1.1 (0.9, 1.3)

Data was reported as frequencies and percentages at the study level. For trial-level analysis, P-values were obtained via Fisher's exact test.

For patient-level analysis, odds ratios were obtained via logistic regression. P-values and odds ratios were obtained via logistic regression.