Preliminary results of a first-in-human, dose-escalation study of a cell-based vaccine in HLA A*02+ patients with recurrent, locally advanced, or metastatic HPV16+ solid tumors SQZ-PBMC-HPV-101: Jong Chul Park, Joaquina Baranda, Monica Mita, Wade T Iams, Michael S Gordon, Matthew Taylor, Neesha Dhani, Scott Loughhead, Rui Ru Ji, Ricardo F Zwirtes, Martin Kornacker, Oliver Rosen, Howard Bernstein, and Antonio Jimeno Massachusetts General Hospital, Boston, MA; University of Kansas Cancer Center, Fairway, KS; Cedars-Sinai Medical Center, Los Angeles, CA; Division of Hematology/Oncology, Department of Medicine, Vanderbilt University, Nashville, TN; Pinnacle Oncology Hematology, Arizona Center for Cancer Care, HonorHealth Research Institute Clinical Trials Program, Virginia G. Piper Cancer Center, Scottsdale, AZ; Providence Cancer Institute, Portland, OR; University Health Network Princess Margaret Cancer Centre, Toronto, Canada; SQZ Biotechnologies, Watertown, MA; Hoffmann-La Roche, Basel, Switzerland; University of Colorado Comprehensive Cancer Center, Aurora, CO ESMO-IO December 9th, 2021

Declaration of Interest Consulting and Advisory Role: ALB, Merck, I-Mab, Mito Research funding: ALX (Inst), Inhibrx (Inst), Merck (Inst), Monopteros (Inst), Novartis (Inst), Oncorus (Inst), SQZ (Inst) Jong Chul Park

SQZ-PBMC-HPV-101: Underlying Technology and Mechanism Prior cancer vaccines relied on cross-presentation, resulting in ineffective MHC-I presentation to CD8 T cells, limiting T cell activation and efficacy. Squeezed Antigen MHC-I = CD8 T Cell Activation SQZTM APC Mechanism: Direct Presentation Typical Cancer Vaccine Mechanism: Cross-Presentation Primarily generates MHC-II presentation for CD4 and antibody responses Suited for prophylactic vaccines Prioritizes creation of: Prioritizes creation of: Primarily generates MHC-I presentation to CD8 T cells CD8 T cell responses in the tumor are highly correlated with patient outcomes MHC-II = CD4 + antibodies MHC-I = CD8 Endocytosis CD8 Killer T cells SQZTM vs. Cross Presentation Data Antibodies CD4 helper cells The Cell Squeeze® Technology The Cell Squeeze® technology has demonstrated robust abilities to deliver antigens directly to the cytosol, thereby circumventing the cross-presentation process most vaccines rely on and enabling efficient MHC-I presentation and antigen-specific CD8 T cell activation. 1. Cells and target cargo together in suspension 2. Cells are squeezed through SQZ chip at high speeds 3. Cell membranes are temporarily disrupted 4. Target cargo enters the cytosol of the cells 5. Membranes reseal Hlavaty KA et al. AACR 2019 Abstract #3187.

Study Design and Cell Therapy Production Cell Therapy Production Monotherapy Escalation Combination De-escalation Recurrent or metastatic HPV-driven cancer HPV-16+ HLA-A *02+ Platinum experienced CPI offered Inclusion Criteria Treatment until PD, unacceptable toxicity or 2 years SQZ-PBMC-HPV 5.0e6/kg Q3W Double Prime (n = 6) SQZ-PBMC-HPV 2.5e6/kg Q3W Double Prime (n = 4) SQZ-PBMC-HPV 2.5e6/kg Q3W (n = 5) SQZ-PBMC-HPV 0.5e6/kg Q3W (n = 3) SQZ-PBMC-HPV RP2D + Atezolizumab 1200mg Q3W SQZ-PBMC-HPV RP2D + Ipilimumab 3mg/kg Q3W x 4 SQZ-PBMC-HPV RP2D + Nivolumab 360mg Q3W SQZ-PBMC-HPV RP2D + Nivo 360mg Q3W + Ipi 1mg/kg Q6W The median time for manufacturing was 17 hours and allowed for a vein-to-vein time of about a week. Among all lots produced, the median viability was 91%. In the highest dose cohort, a median of 5 doses were available for each patient. Patient Leukapheresis Day 1 SQZ Process <24-hours Day 2 QC and Batch Release Day 2-8 Administration to Patient ~1 week

SQZ-PBMC-HPV Evaluated in Advanced Patient Population Characteristic 0.5 M cells/kg SP (n=3) 2.5 M cells/kg SP (n=5) 2.5 M cells/kg DP (n=4) 5.0 M cells/kg DP (n=6) Total (n=18) Median Age, years (min, max) 65 (60, 68)  65 (54, 68) 49 (47, 66) 57.5 (52, 78) 60 (47, 78) Female, n (%)  3 (100) 3 (60) 3 (75) 0 9 (50) Caucasian Race, n (%) 3 (100)  5 (100) 3 (75) 5 (83) 16 (89) Baseline ECOG PS of 1, n (%) 2 (67) 3 (60) 4 (100) 3 (50) 12 (67) Phase 1 RMH Score High, n (%) 0 3 (60) 4 (100) 1 (33) 8 (44) Site of primary tumor, n (%) Anus Cervix Head & Neck   2 (66) 1 (33) 0   3 (60) 0 2 (40)   2 (50) 1 (25) 1 (25) 0 0 6 (100) 7 (39) 2 (11) 9 (50) Metastatic Disease, n (%) 3 (100) 5 (100) 4 (100) 6 (100) 18 (100) Median Number of Prior Lines, n (min, max)  4 (2, 5) 3 (1, 7) 4 (3, 4) 3.5 (2, 6) 4 (1,7) Prior Systemic Therapy, n (%) Chemotherapy Checkpoint Inhibitor Refractory to ICI (PD as BOR) 3 (100) 3 (100) 2/3 5 (100) 4 (80) 3/4 4 (100) 4 (100) 3/4 6 (100) 6 (100) 4/6 18 (100) 17 (94) 12/17 Heavily pretreated patient population with advanced disease at study entry. All but one patient (in the 2nd cohort) were treated with anti-PD-(L)1 checkpoint inhibitor, and the majority of them were considered refractory to a PD-1 inhibition. Patient Characteristics M=millions, SP=single-prime, DP=double-prime, ECOG PS=Eastern Cooperative Oncology Group Performance Status, ICI=immune checkpoint inhibitors, PD=Progressive Disease, BOR=Best Overall Response, RMH=Royal Marsden Hospital

SQZ-PBMC-HPV Was Considered Safe and Well-Tolerated Event, n (%) 0.5 M cells/kg SP (n=3) 2.5 M cells/kg SP (n=5) 2.5 M cells/kg DP (n=4) 5.0 M cells/kg DP (n=6) Total (n=18) Related AEs in >1 Patient 3 (100) 4 (80) 2 (50) 5 (83) 14 (78) Fatigue 0 1 (20) 0 4 (67) 5 (28) Flushing 1 (33) 0 0 2 (33) 3 (17) Hypotension 2 (67) 1 (20) 0 0 3 (17) Infusion Related Reaction (IRR) 1 (33) 0 1 (25) 0 2 (11) Nausea 0 0 0 2 (33) 2 (11) Pruritus 0 1 (20) 0 1 (20) 2 (11) Related Grade ≥3 AEs 0 1 (20%)1 0 0 1 (6%) Related Serious AEs 1 (33%)2 0 0 0 1 (6%) AEs of Special Interest3 1 0 1 0 2 (11%) Dose-Limiting Toxicity 0 0 0 0 0 Related AEs leading to d/c 0 0 0 0 0 Fatal Related AEs 0 0 0 0 0 1Grade 3 anemia. 2Grade 2 cytokine release syndrome (CRS). 3AEs of medical concern related to SQZ-PBMC-HPV mechanism of action (CRS, IRR). M=millions, SP=single-prime, DP=double-prime, AE=adverse events Adverse Events Safety profile of SQZ-PBMC-HPV was consistent across all dose levels. Most related AEs were of low grade. One case of grade 2 cytokine release syndrome in a 1st cohort patient, which resolved in <24h, considered a related serious AE. Only grade ≥3 related adverse event was a case of grade 3 anemia in a 2nd cohort patient. No patients met the DLT criteria.

Potential Proof-of-Concept for SQZ-PBMC-HPV in anti-PD-1 Refractory Patient 52-year-old man with squamous cell carcinoma of the oropharynx. Large primary lesion with significant symptoms burden. Initial diagnosis 3.7 years ago, with 6 prior lines of systemic therapy including carbo/5FU/pembro and anti-TGFβ/pembro (BOR=PD) 9 months before SQZ-PBMC-HPV dosing. Received all 7 doses of SQZ-PBMC-HPV, with excellent tolerability (G1 flushing, G1 fatigue). Marked symptomatic improvement (dysphagia and neck swelling) and improvement of the lesion on physical examination. Case Study: Patient 17 % Change from Baseline Study Day Highest dose cohort changes in CD8 TILs Highest dose cohort tumor measurements Cells/mm2 Patient 17 Post Pre Day-28 on treatment biopsy demonstrated an 8-fold increase in tumor infiltrating CD8 cells. Radiographic response, including confirmed CR on target lesion (mediastinal lymph node (RECIST 1.1)) with a new dermal lesion at the last tumor assessment. 7 Patient 17 Treatment Start (n = 5 patients) (n = 5 patients)

SQZ-PBMC-HPV Turns TME into Inflamed Phenotype in Responding Patient Massive influx of CD8 cells into the tumor changed immunophenotype from desert to inflamed. Increase in HLA-I expression likely due to CD8 driven IFN-γ secretion. Significant reduction in frequency of E6 expressing cells as measured by RNA-ISH. E7 expressing cells follow the same pattern (not shown). Increase in PD-L1 expression suggests potential synergy with a combinatory approach. Case Study: Patient 17 Screening Day 28 CD8 Increase CD8 PanCK 8 % of HPV16+ Cells 93% 23% Cells/mm2 HLA-I Increase HLA-I H-Score (0-300) E6 Reduction HPV16 E6 transcript PD-L1 Increase PD-L1 % of PD-L1+ Tumor Cells

Conclusions and Future Development for SQZ-PBMC-HPV Future Development SQZ-PBMC-HPV at 5.0M cells/kg dose was endorsed by study DSMB to move into PD-(L)1 & CTLA-4 combination cohorts. The highest dose cohort is still enrolling to further characterize efficacy of SQZ-PBMC-HPV as monotherapy. Conclusion Safety: SQZ-PBMC-HPV was considered safe and well-tolerated at all dose levels. Safety profile consisted of mostly low grade (grades 1 and 2) non-specific AEs, only one related serious adverse event, and no dose-limiting toxicities observed. Manufacturability: All batches produced under cGMP yielding multiple cryopreserved doses in <24hrs with ~1wk vein-to-vein. Product characterization confirmed antigen presentation and high viability in all patient batches. Clinical Activity: Radiographic and pathological correlation in patients deriving clinical benefit. Patient 17's clinical response consistent with expected SQZ APC mechanism: increased CD8 tumor infiltration, reduction of E6 and E7 expressing cells suggests antigen-specific killing, increased PD-L1 expression consistent with tumor inflammation.

We thank the patients and their families! Additional information may be found at: