A Randomized, Open-label, Phase II Open Platform Study Evaluating the Efficacy and Safety of Novel Spartalizumab (PDR001) Combinations in Previously Treated Unresectable or Metastatic Melanoma

• Histologically confirmed unresectable or metastatic stage IIIB/C/D or IV melanoma using AJCC edition 8.

• Previously treated for unresectable or metastatic melanoma:

‣ Subjects with V600BRAF wild-type disease had to have received prior systemic therapy for unresectable or metastatic melanoma with anti-PD-1/PD-L1. Additionally, subjects may have received anti-CTLA-4 as a single agent or in combination with anti-PD-1/PD-L1, irrespective of the sequence. No additional systemic treatment was allowed for advanced or metastatic melanoma.

∙ A maximum of two prior lines of systemic therapies for unresectable or metastatic melanoma were allowed.

∙ The last dose of prior therapy (anti-PD-1, anti-PD-L1, or anti-CTLA-4) had to have been received more than four weeks before randomization.

• Subjects with V600BRAF mutant disease had to have received prior systemic therapy for unresectable or metastatic melanoma with anti-PD-1/PD-L1 and V600BRAF inhibitor. Additionally, subjects may have received anti-CTLA-4 as a single agent or in combination with anti-PD-1/PD-L1, or MEK inhibitor (in combination with V600BRAF inhibitor or as a single agent), irrespective of the sequence. No additional systemic treatment was allowed for advanced or metastatic melanoma.

• A maximum of three prior lines of systemic therapies for unresectable or metastatic melanoma were allowed.

• The last dose of prior therapy had to have been received more than 4 weeks (for anti-PD-1, anti-PD-L1, or anti-CTLA-4) or more than 2 weeks (for V600BRAF or MEK inhibitor) prior to randomization.

• All subjects (with V600BRAF wild-type disease and with V600BRAF mutant disease) had to have documented disease progression as per RECIST v1.1 while on/after the last therapy received prior to study entry and while on/after treatment with anti-PD1/PD-L1. The last progression had to have occurred within 12 weeks prior to randomization in the study.

‣ ECOG performance status 0-2.

⁃ At least one measurable lesion per RECIST v1.1.

⁃ At least one lesion, suitable for sequential mandatory tumor biopsies (screening and on-treatment) in accordance with the biopsy guidelines specified in the protocol. The same lesion had to be biopsied sequentially.

⁃ Screening tumor biopsy had to fulfill the tissue quality criteria outlined in the protocol, as assessed by a local pathologist.

• Histologically confirmed unresectable or metastatic stage IIIB/C/D or IV melanoma according to AJCC Edition 8.

• Previously treated for unresectable or metastatic melanoma:

‣ All subjects had to have received anti-PD-1 checkpoint inhibitor therapy (i.e., pembrolizumab or nivolumab) either as monotherapy or in combination with ipilimumab as the last systemic therapy prior to enrollment and had to have confirmed disease progression as per RECIST v1.1 (confirmed on a subsequent scan, which could be the scan performed during screening) while on or after this therapy prior to enrollment.

⁃ Subjects with V600BRAF wild-type disease had to have received no more than 2 prior systemic therapies, including prior anti-PD-1/PD-L1 (as monotherapy or in combination with ipilimumab).

⁃ Subjects with V600BRAF mutant disease had to have received no more than 3 prior systemic therapies, including anti-PD-1/PD-L1 (as monotherapy or in combination with ipilimumab) and V600BRAF inhibitor (as monotherapy or in combination with a MEK inhibitor).

⁃ The last dose of anti-PD-1-based therapy had to have been received more than four weeks prior to the first dose of study treatment.

⁃ The last documented disease progression had to have occurred within 12 weeks prior to the first dose of study treatment.

⁃ No additional systemic treatment was allowed for advanced or metastatic melanoma (this included, for example, tumor-infiltrating lymphocyte therapy).

• ECOG performance status 0-1.

• At least one measurable lesion per RECIST v1.1.

• Subjects had to have a baseline tumor sample that was positive for LAG-3 per central assessment.