Clinical cohort IO results
A total of 596 patients with advanced solid tumors [breast, ovarian, pancreatic, and prostate cancers] among the participating collaborators were found to have bTMB ≥ 10 mut/Mb (Fig. 1). Patients who received an ICI prior to G360 (n = 21), had an unavailable bTMB or a bTMB of < 10 mut/Mb prior to ICI initiation (n = 6), were excluded. Our final analysis included 48 bTMB high patients treated with ICI therapy (n = 36 prostate, n = 12 breast).
Table 1 summarizes the baseline characteristics of the study cohort (bTMB ≥ 10 mut/Mb and subgroup of patients with bTMB ≥ 16 mut/Mb. In this cohort, liquid biopsy was performed after a median of 3 (1–9) lines of treatment and ICI was administered after a median of 3 (1–9) lines of treatment. Most patients were previously exposed to chemotherapy (52%). All patients except one were ICI-naïve prior to TMB results. Most patients were treated with ICI monotherapy including a PD-1 inhibitor while only three patients were offered a combination regimen of two different immunotherapies (Table 1).
Table 1
Baseline Characteristics of Clinical Cohort (N = 48). NH = non-Hispanic, ADT = androgen deprivation therapy, ARSI = androgen receptor signaling inhibitor, ICI = immune checkpoint inhibitor, BiTE- bispecific T-cell engager.
Table 1: Baseline Characteristics | |
| bTMB ≥ 10, (N = 48) | bTMB ≥ 16 (N = 27) |
Median Age (range) | 72 (37–87) | 72 (37–86) |
Race (%) NH White NH Black Hispanic American Indian Unknown (%) | 37 (77.1) 4 (8.3) 3 (6.3) 2 (4.2) 2 (4.2) | 21 (77.8) 1 (3.7) 3 (11.1) 2 (7.4) 0 (0) |
Primary tumor, (%) Prostate Breast | 36 (75.0) 12 (25.0) | 18 (66.7) 9 (33.3) |
MSI Status, (%) MSI-Detected MSI Not Detected Unknown | 8 (16.7) 32 (66.7) 8 (16.7) | 8 (29.6) 17 (63.0) 2 (7.4) |
MMR deficient | 3 (6.3) | 3 (11.1) |
Histology Adenocarcinoma Small cell | 47 (97.9) 1 (2.1) | 27 (100) 0 (0) |
Location of metastases Bones Lymph nodes Visceral Liver Lung Other – bladder, adrenals, pleura, brain Soft tissue | 36 (75.0) 24 (50.0) 11 (22.9) 7 (14.6) 8 (16.7) 4 (8.3) | 21 (77.8) 14 (51.9) 7 (25.9) 4 (14.8) 7 (25.9) 2 (7.4) |
Prior systemic therapies prior to IO, (range) | 3 (1–9) | 3 (1–9) |
Prior systemic therapies ADT ARSI Aromatase inhibitors Chemotherapy Sipuleucel-T Targeted therapies Investigational agents PD(L)-1 agents | 39 (81.3) 31 (64.6) 7 (14.6) 25 (52.1) 9 (18.8) 13 (27.1) 1 (2.1) 1 (2.1) | 18 (66.7) 14 (51.9) 5 (18.5) 14 (51.9) 2 (7.4) 7 (25.9) 0 (0) 1 (3.7) |
ICI received Pembrolizumab Nivolumab Atezolizumab Investigational anti-PD-1 Ipilimumab/nivolumab Anti-PD-1/CTLA4 bispecific Atezolizumab/BiTE | 42 (87.5) 1 (2.1) 1 (2.1) 1 (2.1) 1 (2.1) 1 (2.1) 1 (2.1) | 26 (96.3) 0 0 0 1 (3.7) 0 0 |
Supplemental Figures |
Table S1
HRR genes by bTMB-H cutoffs and MSI status in patients with prostate cancer: bTMB-16 vs bTMB-10 groups.
HRR Genes | bTMB-10 | bTMB-16 | p Value | bTMB-10 MSI-Detected | bTMB-10 MSI- Not Detected | p Value | bTMB-16 MSI-Detected | bTMB-16 MSI- Not Detected | p Value |
ATM | 26.26% | 32.53% | 0.0018 | 43.17% | 24.28% | ≤ 0.0001 | 43.95% | 26.76% | ≤ 0.0001 |
BRCA1 | 8.25% | 15.36% | ≤ 0.0001 | 23.35% | 6.48% | ≤ 0.0001 | 23.77% | 11.11% | ≤ 0.0001 |
BRCA2 | 17.46% | 31.02% | ≤ 0.0001 | 40.53% | 14.76% | ≤ 0.0001 | 40.81% | 26.08% | ≤ 0.0001 |
CDK12 | 10.09% | 13.25% | 0.0267 | 24.67% | 8.38% | ≤ 0.0001 | 25.11% | 7.26% | ≤ 0.0001 |
CHEK2 | 14.42% | 16.27% | 0.2616 | 14.98% | 14.35% | 0.432 | 15.25% | 16.78% | 0.349 |
PALB2 | 5.02% | 11.14% | ≤ 0.0001 | 20.7% | 3.19% | ≤ 0.0001 | 21.08% | 6.12% | ≤ 0.0001 |
RAD51D | 1.38% | 3.31% | 0.0025 | 5.29% | 0.93% | ≤ 0.0001 | 5.38% | 2.27% | 0.0323 |
The median bTMB was 16.4 (10–186) mut/Mb; no significant difference between the median bTMB in prostate cancer [15.8 (10-147.9)] and breast cancer [24.3 (11.5-174.5)] was observed (Figure S1). Microsatellite instability (MSI) in blood (bMSI) was detected in eight (16%) bTMB high patients treated with ICI (prostate n = 6; breast n = 2). Only four patients had known MMR status known (Table 1): one patient had both MSI-H and MMR defect, one patient had MSI-H with preserved MMR and one patient had MMR defect without MSI-H. The median bTMB among patients with bMSI-H/MMR defect was 91.2 (19.3–186.0) as compared to the median bTMB among patients without MSI-H/ MMR defect, at 16.3 (10.0-147.9). High bMSI was associated with higher bTMB (correlation test, r = 0.66, p < 0.001). Median bTMB was 91.2 (19.3–186.0) mut/Mb in bMSI-H vs 16.3 (10.0-147.9) mut/Mb in bMSI-Not Detected tumors. Patients with high bMSI-H had a mPFS of 10.3 months (CI 95% 0-15.4) compared to a mPFS of 2.8 months (CI 95% 2.1-4.0) in those patients with microsatellite stable tumors (Fig. 2A). With a median follow up of 10.9 (0-34.3) months, the median time on ICI was 2.1 (0-1.7) months (no difference by MSI/MMR status, p = 0.214). Most patients (26) discontinued ICI due to disease progression and/or death, irAES (5 pts: 1 – hepatitis; 1 – ARF, 2 – dermatitis, 1 - colitis), or unknown (4 pts). The median progression free survival was 4.0 months (95% CI, 2.4–5.6) and 37 patients (77%) progressed on therapy. There was no association between median bTMB and mPFS (HR 0.99 CI 0.99-1.0; p = 0.56). Similarly, there was no association between MSI status and mPFS with ICI at different bTMB cut-off of 16, 20 and 50 mut/Mb (Fig. 2A-C).
The median OS was 12.8 months (95% CI, 6.1–19.4) and no difference was observed by MSI/MMR status (16.5 months [CI 95%, 14-19.1] MSI-H/ MMR defect vs 10.3 months [CI 95%, 0-27.2], p = 0.906). For prostate cancer, the median OS was 12.1 months (CI 95%, 6.1–18.1), and for breast cancer the median OS was 16.6 months (CI 95%, 0-39.7). The best overall response rate among eligible patients (n = 36) was 16.7% (complete response n = 1; partial response n = 5); all responses except one were observed in patients with bTMB ≥ 16 mut/Mb. All partial responses were observed in bMSI-H and one complete response was observed in one patient with bMSI-Not Detected.
Genomic analysis
The Guardant Health genomic database (GHGD) was queried for patients with breast (BC; n = 7899) or prostate (PR; n = 6093) cancers who had a bTMB identified by ctDNA NGS (Total n = 13,992; Guardant360, Redwood City, CA) for associations of bTMB with genomic alterations
Previous studies have suggested that a bTMB-high benchmark of ≥ 16 mut/Mb would correlate with tissue TMB-high.15,16 Thus, we sought to analyze breast and prostate tumors within a retrospective ctDNA dataset for bTMB prevalence and genomic differences between bTMB-H and bTMB-L tumors, utilizing two different cut-offs: ≥10 mut/Mb (from tissue approval) which is the cutoff included in this study and ≥ 16 mut/Mb, as suggested in other reports.15,16
Genomic Database –bTMB ≥ 10 mut/Mb
More than a quarter of patients (28.4%, 5327/13992) had a bTMB ≥ 10 mut/Mb (bTMB-10). Over 5% (290/5327) of bTMB-10 patients and only 0.01% (1/8665) of patients with bTMB < 10 mut/Mb had MSI-Detected. Most bTMB-10 patients (94.56%; 5037/5327) were MSI-Not Detected.
Genomic Database –bTMB ≥ 16 mut/Mb
Approximately 15% (2039/13992) of patients had a bTMB ≥ 16 mut/Mb (bTMB-16). Almost 14% (285/2039) of bTMB-16 patients and 0.05% (6/11953) of patients with bTMB < 16 mut/Mb also had MSI-Detected, respectively. Most bTMB-16 patients (86.02%; 1754/2039) were MSI-Not Detected. Altogether, bTMB-16 had a higher percentage of MSI-Detected/TMB-H than patients with bTMB < 16 mut/Mb (13.98% vs 5.4%).
The top 20 most frequently altered genes were significantly higher in both bTMB-10 and bTMB-16 groups compared to patients with bTMB < 10 and < 16 mut/Mb, respectively (Fig. 3A and 3C). Among bTMB-10 and bTMB-16 cohorts, patients with MSI-Detected harbored significantly more frequent alterations in the top assessed genes, with one exception (Fig. <link rid="fig4">3</link>B,<link rid="fig4">3</link>D and S2-3)). In the bTMB-16 group, alterations in PIK3CA were not significantly different in those with MSI-Detected vs. MSI-Not Detected (Fig. 3D).
In breast cancer, most of the genes were significantly higher in bTMB-H patients (MSI-Detected were removed from this cohort) among both the bTMB-10 and bTMB-16 groups (Figure S3A-B). These included TP53, PIK3CA, ESR1, NF1, CDH1, ATM, ARID1A, BRCA2, APC, RB1, BRCA1, ERBB2, MTOR, EGFR, CHEK2, NOTCH1, ROS1, ALK, and PTEN genes. In bTMB-10 GATA3 was significantly higher than in patients with a bTMB < 10, but there was no significance observed in the bTMB-16 group.
Results in prostate cancer were similar to that of the breast cancer cohort, with most genes being significantly higher in bTMB-H patients in both the bTMB-10 and bTMB-16 groups (MSI-Detected were removed in this cohort as well). This included TP53, ATM, APC, BRCA2, CHEK2, NF1, TERT, EGFR, PIK3CA, NOTCH1, ROS1, CDK12, ARID1A, PTEN, RB1, MTOR, BRCA1, and ALK genes (Figure S3C-D). In bTMB-10 AR was significantly higher than in patients with a bTMB < 10, but there was no significance observed in the bTMB-16 group. CTNNB1 showed no significant difference in both bTMB-10 and bTMB-16.
Interestingly, in prostate cancer, the bTMB-16 group has a significantly higher frequency of alterations in the majority of HRR genes than the bTMB-10 group (Table S1), including ATM (32.53% vs. 26.26%; p = 0.0018), BRCA2 (31.02% vs. 17.46%; p ≤ 0.0001), BRCA1 (15.36% vs. 8.25%; p ≤ 0.0001), CDK12 (13.25%vs. 10.09%; p = 0.0267), PALB2 (11.14% vs. 5.02%; p ≤ 0.0001), and RAD51D (3.31% vs 1.38%; p = 0.0025). CHEK2 (16.27% vs. 14.42%; p = 0.2616) was the only HRR gene without a significant difference between groups. Similarly, when MSI status was assessed, alterations of HRR genes were significantly more frequent in MSI-Detected vs MSI-Not Detected, with the exception of the CHEK2 gene which was not significant between groups.