Clinical Data MARIO-3 Key Data Updates:

• This data update includes 50 patients enrolled and 44 evaluable as of the October 2, 2021 data cutoff date, with a median duration of follow up of 9.9 months.
• Of evaluable patients, tumor reduction was observed in 92.8% of patients with PD-L1 positive tumors (13/14) and 85.2% of patients with PD-L1 negative tumors (22/27).
• Disease control rate (DCR)
  • 92.8% (13/14) DCR in patients with PD-L1 positive tumors: CR 14.3% (2/14), PR 57.1% (8/14), SD 21.4% (3/14)
  • 81.4% (22/27) DCR in patients with PD-L1 negative tumors: complete response (CR) 0% (0/27), partial response (PR) 48.1% (13/27), stable disease (SD) 33.3% (9/27)
• Progression free survival (PFS)
  • In patients with PD-L1(+) tumors, median PFS in MARIO-3 was 11.0 months, a 47% improvement in mPFS compared to the 7.5 months reported for atezolizumab and nab-paclitaxel alone in IMpassion130
  • In patients with PD-L1(-) tumors, median PFS in MARIO-3 was 7.3 months, a 30% improvement compared to the 5.6 months reported for atezolizumab and nab-paclitaxel alone in IMpassion130
  • 72% of the 32 PD-L1 (+) and PD-L1(-) patients treated since the June 26, 2021 data cut remain on treatment
  • 67% of the PD-L1(-) patients who reached the median PFS of 7.3 months remain on treatment
• Safety
  • MARIO-3 did not demonstrate any new safety signals compared to benchmark trials, and its safety profile was consistent with expectations for the three component drugs. The most common Grade 3 or higher treatment-related TEAEs were hepatic AEs (18%); neutropenia AEs (16%); skin AEs (12%); fatigue, diarrhea and peripheral sensory neuropathy (6% each); and vomiting and weight decreased (2% each). Seven patients (14%) discontinued treatment for treatment-related TEAEs and nine patients (18%) had treatment-related SAEs.
• Quantification across 11 paired tumor biopsies shows increased immune activation and decreased immune suppression including an increase in CD8+ T cells, activated T cells, and anti-tumor M1 macrophages and a decrease in tumor cells and pro-tumor M2 macrophages resulting in an increase in the M1:M2 ratio.
• Paired tumor biopsy data show 5 of 8 patients with PD-L1(-) tumors converting to PD-L1(+) two months after treatment utilizing the same 1% PD-L1 cutoff standard used in the benchmark IMpassion130 study. PD-L1 expression also increased in the three patients with PD-L1(+) tumors who started the study above the 1% cutoff. None of the patients converting to PD-L1(+) or patients with PD-L1(+) tumors who experienced increased PD-L1 expression had disease progression.