Progression-free survival (PFS) (Comparison between arms)

Will compare between participants randomized to the combination of sacituzumab govitecan plus ivonescimab (SG-I) and sacituzumab govitecan (SG) alone. The comparison of PFS will be done using a 1-sided 10% level log-rank test stratified by randomization stratification factors. Binary proportions along with 90% confidence intervals will be estimated for response to be consistent with 1-sided 5% level testing. With 35 participants per arm, binary proportions can be estimated to within 11% with 90% confidence. The distribution PFS will be estimated using the method of Kaplan-Meier. Comparisons of event time distributions between arms will be summarized by hazard ratios and 80% confidence intervals (consistent with 1-sided 10% level testing) from a Cox Proportional hazards model including the stratification factors.

From date of randomization to date of first documentation of progression, symptomatic deterioration, or death due to any cause, assessed up to 3 years

PFS (Comparison between arms)

Will between participants randomized to SG-I and ivonescimab (I) alone. The comparison of PFS will be done using a 1-sided 10% level log-rank test stratified by randomization stratification factors. Binary proportions along with 90% confidence intervals will be estimated for response to be consistent with 1-sided 5% level testing. With 35 participants per arm, binary proportions can be estimated to within 11% with 90% confidence. The distribution PFS will be estimated using the method of Kaplan-Meier. Comparisons of event time distributions between arms will be summarized by hazard ratios and 80% confidence intervals (consistent with 1-sided 10% level testing) from a Cox Proportional hazards model including the stratification factors.

From date of randomization to date of first documentation of progression, symptomatic deterioration, or death due to any cause, assessed up to 3 years

Response rate of SG against historical response rates (Single arm evaluation)

Within each treatment arm, the objective is to evaluate if response rates show activity beyond standard of care, which generally would be single agent chemotherapy or docetaxel combined with ramucirumab. With 35 eligible and evaluable participants per arm, the evaluation has 90% exact power to rule out a 10% response rate using a 1-sided 5% level test (the exact level is 5.1% based on design assumptions), if the true response rate were at least 29%. The observation of at least 7 participants with a response (a 20% response rate) would be considered evidence to rule out a 10% response rate.

Up to 3 years

Response rate of I against historical response rates (Single arm evaluation)

Within each treatment arm, the objective is to evaluate if response rates show activity beyond standard of care, which generally would be single agent chemotherapy or docetaxel combined with ramucirumab. With 35 eligible and evaluable participants per arm, the evaluation has 90% exact power to rule out a 10% response rate using a 1-sided 5% level test (the exact level is 5.1% based on design assumptions), if the true response rate were at least 29%. The observation of at least 7 participants with a response (a 20% response rate) would be considered evidence to rule out a 10% response rate.

Up to 3 years

Response rate of SG-I against historical response rates (Single arm evaluation)

Within each treatment arm, the objective is to evaluate if response rates show activity beyond standard of care, which generally would be single agent chemotherapy or docetaxel combined with ramucirumab. With 35 eligible and evaluable participants per arm, the evaluation has 90% exact power to rule out a 10% response rate using a 1-sided 5% level test (the exact level is 5.1% based on design assumptions), if the true response rate were at least 29%. The observation of at least 7 participants with a response (a 20% response rate) would be considered evidence to rule out a 10% response rate.

Up to 3 years

Rate of dose-limiting toxicities among participants treated with SG-I in the safety run-in analysis population

Defined as treatment-related Grade 3 or higher non hematologic toxicity, treatment-related Grade 4 or higher hematologic toxicity, or any Grade of treatment-related toxicity that leads to drug discontinuation. Toxicities will be graded based on Common Terminology Criteria for Adverse Events (CTCAE).

During the first cycle of treatment (21 days)

Response rates

Will compare response rates between I with or without SG.

Up to 3 years

Response rates

Will compare response rates between SG with or without I.

Up to 3 years

Overall survival (OS)

Will compare between the SG-I and SG arms. The distribution of OS will be estimated using the method of Kaplan-Meier. Comparisons of event time distributions between arms will be summarized by hazard ratios and 80% confidence intervals (consistent with 1-sided 10% level testing) from a Cox Proportional hazards model including the stratification factors.

From date of randomization to date of death due to any cause, assessed up to 3 years

OS

Will compare between the SG-I and I arms. The distribution of OS will be estimated using the method of Kaplan-Meier. Comparisons of event time distributions between arms will be summarized by hazard ratios and 80% confidence intervals (consistent with 1-sided 10% level testing) from a Cox Proportional hazards model including the stratification factors.

From date of randomization to date of death due to any cause, assessed up to 3 years

Duration of response (DoR)

The distribution of DoR will be estimated using the method of Kaplan-Meier.

From date of first documentation of response to date of first documentation of progression, symptomatic deterioration, or death due to any cause among participants who achieve a response, assessed up to 3 years

Central nervous system (CNS) PFS

CNS progression is defined as the development of new CNS metastasis or progression of CNS disease requiring either local CNS therapy or a change in systemic cancer therapy. The distribution of CNS PFS will be estimated using the method of Kaplan-Meier. Comparisons of event time distributions between arms will be summarized by hazard ratios and 80% confidence intervals (consistent with 1-sided 10% level testing) from a Cox Proportional hazards model including the stratification factors.

From date of randomization to date of first documentation of progression, symptomatic deterioration, or death due to any cause, assessed up to 3 years

Frequency and severity of toxicities

Toxicities will be graded based on Common Terminology Criteria for Adverse Events (CTCAE). Toxicities will be summarized as proportions by item and worst grade reported, among those possible, probably, or likely related to treatment and also grouped overall and as hematologic and non-hematologic toxicities. Any toxicity, with at least 10% true prevalence has a 97% chance of being observed.

During the first cycle of treatment (21 days)

Clinical outcomes by EGFR subgroups

Subgroup analyses will be done using a Cox proportional hazards model for time-to-event outcomes and logistic regression for binary endpoints.

Up to 3 years

Clinical outcomes by PD-L1 expression subgroups

Subgroup analyses (PD-L1 expression \[\< 1% vs 1% or greater\]) will be done using a Cox proportional hazards model for time-to-event outcomes and logistic regression for binary endpoints.

Up to 3 years