PFS is the time from the date of randomization to the date of event defined as the first documented confirmed progression or death due to any cause. If a patient has not had an event, progression-free survival is censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1). Progressive disease is at least a 20% increase in the sum of diameter of all measured target lesions, taking as reference the smallest sum of diameter of all target lesions recorded at or after baseline. Number of participants in each category (progression, death, censored) is reported in this record.

All participants to whom study treatment was assigned by randomization.

PFS is the time from the date of randomization to the date of event defined as the first documented confirmed progression or death due to any cause. If a patient has not had an event, progression-free survival is censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1). Progressive disease is at least a 20% increase in the sum of diameter of all measured target lesions, taking as reference the smallest sum of diameter of all target lesions recorded at or after baseline.

All participants to whom study treatment was assigned by randomization.

Overall survival (OS) is defined as the time from date of randomization to date of death due to any cause. If a patient is not known to have died, survival is censored at the date of last known date the patient was alive.

All participants to whom study treatment was assigned by randomization.

ORR is defined as the percentage of participants with a best overall response of Complete Response (CR) or Partial Response (PR). Tumor response was based on central review of overall lesion response according to RECIST 1.1.

For RECIST v1.1, CR=Disappearance of all non-nodal target lesions. In addition, any pathological lymph nodes assigned as target lesions must have a reduction in short axis to < 10 mm; PR= At least a 30% decrease in the sum of diameter of all target lesions, taking as reference the baseline sum of diameters.

All participants to whom study treatment was assigned by randomization.

DOR only applies to subjects for whom best overall response is complete response (CR) or partial response (PR) based on central review of overall lesion response according to RECIST 1.1. DOR is defined as the time between the date of first documented response (confirmed CR or confirmed PR) and the date of first documented disease progression or death due to underlying cancer. If a patient not had an event, duration was censored at the date of last adequate tumor assessment.

All participants for whom best overall response is complete response (CR) or partial response (PR) as per RECIST 1.1 based on central review.

TTP is defined as time from date of randomization to the date of event defined as the first documented progression or death due to underlying cancer. If a subject did not had an event, TTP was censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was RECIST v1.1. Progressive disease is at least a 20% increase in the sum of diameter of all measured target lesions, taking as reference the smallest sum of diameter of all target lesions recorded at or after baseline.

All participants to whom study treatment was assigned by randomization.

irPFS is the time from the date of randomization to the date of event defined as the first documented progression or death due to any cause. If a patient has not had an event, progression-free survival is censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was immune-related Response Criteria (irRC). Immune-related progressive disease is at least a 20% increase in the sum of diameters of all measured target lesions including new measurable lesions.

All participants to whom study treatment was assigned by randomization to spartalizumab. For the chemotherapy arm, tumor scans for efficacy assessment as per irRC were not planned.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. Cmax is defined as the maximum (peak) observed serum concentration following a dose.

All participants who had at least one dose of spartalizumab and had at least one evaluable concentration measurement of spartalizumab. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. Tmax is defined as the time to reach maximum (peak) serum concentration following a dose. Actual recorded sampling times were considered for the calculations.

All participants who had at least one dose of spartalizumab and had at least one evaluable concentration measurement of spartalizumab. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. The linear trapezoidal method was used for AUC calculation. The duration of the dosing interval (tau) was 28 days.

All participants who had at least one dose of spartalizumab and had at least one evaluable concentration measurement of spartalizumab. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. The linear trapezoidal method was used for AUC calculation. The extrapolation of AUC to infinity could be calculated if the percentage of area extrapolated was less than 20% and the regression analysis of the terminal serum elimination phase met a pre-defined criteria of goodness of fit.

All participants who had at least one dose of spartalizumab and the extrapolation of AUC to infinity could be calculated as described in the description of the endpoint. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. Racc was calculated as the ratio between AUCtau Cycle 3 and AUCtau Cycle 1.

All participants who had at least one dose of spartalizumab and had at least one evaluable concentration measurement of spartalizumab. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Pharmacokinetic (PK) parameters were calculated based on spartalizumab serum concentrations by using non-compartmental methods. Elimination half-life (T1/2) values were calculated as 0.693/terminal elimination rate constant.

All participants who had at least one dose of spartalizumab and had at least one evaluable concentration measurement of spartalizumab. PK samples were only collected in participants to whom spartalizumab was assigned by randomization.

Validated immunoassays were used for screening and confirmation of the presence of anti-spartalizumab antibodies (ADA, anti-drug antibodies) in serum. Number of participants with each ADA status is reported in this record.

All participants who had at least one dose of spartalizumab and had a determinant baseline immunogenicity (IG) sample and at least one determinant post-baseline IG sample. A determinant IG sample is neither ADA-inconclusive nor unevaluable. IG samples were only collected in participants to whom spartalizumab was assigned by randomization.

The expression of programmed cell death-ligand 1 (PD-L1) was measured in tumor samples by immunohistochemical methods. This record summarizes the PD-L1 positivity percentage in tumor samples.

All participants to whom study treatment was assigned by randomization to spartalizumab and had a valid assessment for the outcome measure. Biomarker samples were not collected in participants randomized to chemotherapy.

The expression of CD8 was measured in tumor samples by immunohistochemical methods. This record summarizes the percent marker area for CD8 expression in tumor samples.

All participants to whom study treatment was assigned by randomization to spartalizumab and had a valid assessment for the outcome measure. Biomarker samples were not collected in participants randomized to chemotherapy.

The count of tumor-infiltrating lymphocytes (TILs) was measured in baseline (screening) and post-baseline paired tumor samples by immunohistochemical methods. This record summarizes the TIL count in tumor samples.

All participants to whom study treatment was assigned by randomization to spartalizumab and had paired tumor biopsies. Biomarker samples were not collected in participants randomized to chemotherapy.

The levels of interferon-gamma (IFN-gamma) were measured in plasma samples by immunoassay methods. This record summarizes the fold change from baseline in IFN-gamma levels in plasma.

All participants to whom study treatment was assigned by randomization to spartalizumab. Biomarker samples were not collected in participants randomized to chemotherapy.

The levels of interleukin-6 (IL-6) were measured in plasma samples by immunoassay methods. This record summarizes the fold change from baseline in IL-6 levels in plasma.

All participants to whom study treatment was assigned by randomization to spartalizumab. Biomarker samples were not collected in participants randomized to chemotherapy.

The levels of tumor necrosis factor-alpha (TNF-alfa) were measured in plasma samples by immunoassay methods. This record summarizes the fold change from baseline in TNF-alfa levels in plasma.

All participants to whom study treatment was assigned by randomization to spartalizumab. Biomarker samples were not collected in participants randomized to chemotherapy.

PFS is the time from the date of first dose of spartalizumab to the date of event defined as the first documented confirmed progression or death due to any cause. If a patient has not had an event, progression-free survival is censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1). Progressive disease is at least a 20% increase in the sum of diameter of all measured target lesions, taking as reference the smallest sum of diameter of all target lesions recorded at or after baseline. Number of participants in each category (progression, death, censored) is reported in this record.

All participants who crossed over from chemotherapy to spartalizumab.

PFS is the time from the date of first dose of spartalizumab to the date of event defined as the first documented confirmed progression or death due to any cause. If a patient has not had an event, progression-free survival is censored at the date of last adequate tumor assessment.

Tumor response was based on central review of tumor scan and the assessment criteria was Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1). Progressive disease is at least a 20% increase in the sum of diameter of all measured target lesions, taking as reference the smallest sum of diameter of all target lesions recorded at or after baseline.

All participants who crossed over from chemotherapy to spartalizumab.

On-treatment deaths were collected from the start of treatment up to 30 days after study drug discontinuation, for a maximum duration of 4.0 years. Post-treatment deaths were collected after the on-treatment period (starting at day 31 after last dose of study treatment), up to 4.8 years.

For the crossover patients, the deaths collected before crossing over are summarized in the chemotherapy arm and the deaths collected after the crossover are summarized in the crossover arm.

All participants to whom study treatment was assigned by randomization.