Table 3. Categorical analysis association of anti-TPO antibodies with CBR and toxicity.

Baseline and pre-cycle 3 categories were dichotomized by ≤ to median titers (10.0 IU/mL) versus > median titers. Δ was dichotomized using pre-cycle 3 titers ≥ 1.5x baseline titers vs pre-cycle 3 titers < 1.5x baseline titers. Patients who developed toxicity before pre-cycle 3 blood draw (n=5) were removed from both the CBR and toxicity pre-cycle 3 and Δ analysis.

Event	Time point	Method of dichotomization	Unadjusted			Adjusted for cohort		Multivariable a
			With event – no. (%)	Without event – no. (%)	P value b	OR (95% CI)	P value c	OR (95% CI)	P value d
CBR	Baseline	≤ 10.0 IU/mL	11 (46)	28 (52)	0.81	1.3 (0.5–3.6)	0.65	1.0 (0.3–3.1)	0.98
		> 10.0 IU/mL	13 (54)	26 (48)
	Pre-cycle 3	≤ 10.0 IU/mL	11 (48)	22 (44)	0.80	1.0 (0.3–2.8)	0.97	1.0 (0.3–3.2)	0.94
		> 10.0 IU/mL	12 (52)	28 (56)
	Δ	Pre-cycle 3 < 1.5x baseline	21 (91)	44 (88)	> 0.99	0.5 (0.08–3.1)	0.45	0.7 (0.1–4.8)	0.71
		Pre-cycle 3 ≥ 1.5x baseline	2 (9)	6 (12)
Toxicity	Baseline	≤ 10.0 IU/mL	8 (42)	31 (53)	0.60	1.6 (0.5–4.6)	0.41	1.4 (0.4–4.7)	0.61
		> 10.0 IU/mL	11 (58)	28 (48)
	Pre-cycle 3	≤ 10.0 IU/mL	4 (29)	29 (49)	0.23	2.7 (0.7–10.3)	0.14	2.3 (0.5–10.2)	0.26
		> 10.0 IU/mL	10 (71)	30 (51)
	Δ	Pre-cycle 3 < 1.5x baseline	10 (71)	55 (93)	0.039	6.1 (1.1–32.7)	0.035	5.9 (0.8–42.6)	0.078
		Pre-cycle 3 ≥ 1.5x baseline	4 (29)	4 (7)

CBR = clinical benefit rate; CI = confidence intervals; OR = odds ratio.

^aDetermined using propensity scores with cohort as strata and the following covariates: age, gender, ethnicity and PD-L1 status. One patient was removed due to unknown ethnicity.

^bCalculated using Fisher’s Exact Test.

^cBased on conditional logistic regression with the cohort as strata.

^dBased on conditional logistic regression with the cohort as strata and adjusted for the propensity scores.