Clinicopathologic characteristics of the training cohort
There were 40 patients with EOC in the training group, and their clinical characteristics are shown in Table 1. The median age was 56.5 years old, and the median pretreatment CA-125 value was 889 U/ml. All patients had advanced-stage (FIGO stages III and IV) high-grade serous EOC. All patients underwent primary debulking surgery, with R0 resection in 13 and R1 resection in 27. All patients received adjuvant platinum-based and paclitaxel chemotherapy, and 23 (57.5%) patients were platinum-sensitive. The median follow-up duration was 60 months. The Tumor recurred in 31 (77.5%) patients, and 21 (52.5%) patients died due to EOC.
Establish the of WES-based HRD
The HRD scores and deleterious gene mutations detected in these two tests are shown in Figure 2A. In the WES-based scarHRD test, the HRD scores of 40 patients ranged from 17 to 90. In the Myriad MyChoice CDx test, the HRD scores of 40 patients ranged from 3 to 84. A linear regression model was applied to analyze the correlation between the WES-based scarHRD score and the Myriad MyChoice CDx HRD score (Figure 2B). The WES-based scarHRD score was strongly correlated with the Myriad MyChoice CDx × HRD score (correlation coefficient (r): 0.82, p<0.001). Based on the regression model, we defined positive HRD status as BRCA gene mutation or a score of ≥50 in our WES-based scarHRD test, which is equal to a score of 42 in the Myriad MyChoice® CDx test.
Overall, 32 patients had a positive HRD status (score >50) in the WES-based test and 30 patients had a positive HRD status according to the Myriad MyChoice® CDx test (score >42). For DDR gene mutations, BRCA1 mutation was noted in two patients, BRCA2 in two patients, ATM in one patient, CHEK2 in one patient, RAD51C in two patients, FANCG in one patient, and MSH6 in one patient (Supplementary Table S2). All patients with DDR mutations had a positive HRD status. In addition, we confirmed that the sequencing outcome of BRCA mutations in the cohort found by the Myriad MyChoice® CDx test was recapitulated in our WES-based test.
Compared with the positive HRD status in the Myriad MyChoice® CDx test, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the WES-based HRD test were 93.5% (29/31), 77.8% (7/9), 93.5% (29/31), and 77.8% (7/9), respectively (Supplementary Table S3).
Correlation of HRD status to clinical outcomes in the training group
In WES-based test, a higher percentage of patients with a platinum-sensitive response had a positive HRD status than that of patients with a platinum-resistant response (95.7% [22/23] vs. 52.9% [9/17], p=0.002, Fisher's exact test, Table 2). Similarly, a higher percentage of patients with a platinum-sensitive response had a positive HRD by Myriad MyChoice® CDx test than that of patients with a platinum-resistant response (95.7% [22/23] versus 52.9% [9/17], p=0.002, Fisher’s exact test, Table 2). There was no significant difference in the percentage of EOC patients with positive HRD status as defined by the two tests according to cancer recurrence and cancer-related death.
Patients with EOC who underwent debulking surgery with R0 resection had a longer PFS (p=0.032, log-rank test; Figure 2C) and OS (p=0.013, log-rank test; Supplementary Figure 1A) than those who underwent R1 resection. Patients with a positive HRD status, either by the WES-based test or the Myriad MyChoice® CDx test, had a longer PFS (both p=0.002, log-rank test; Figure 2D & 2E). The predictive value of OS for the two HRD tests was unsatisfactory (Supplementary Figure 1B, C).
Debulking surgery with R0 resection (hazard ratio [HR] 0.42, 95% CI 0.18-0.98, p=0.045) and HRD positive status test (WES based HRD: HR 0.29, 95% CI 0.12-0.68; , Myriad MyChoice CDx test: HR 0.28, 95% CI 0.12-0.66) were associated with disease recurrence in univariate analysis. The multivariate Cox regression model showed that a positive HRD status, which was defined by either the WES-based test or the Myriad MyChoice® CDx test, was an independent factor for disease progression after adjustment for R0/R1 resection. However, a positive HRD status in both tests was associated with a better OS trend, which was not statistically significant. The multivariate Cox regression model for the risk of cancer-related death revealed that only debulking surgery with R0 resection was an independent risk factor in the multivariate analysis.
Evaluation of WES-based scarHRD test in validation group
There were 23 patients in the validation group, all of whom had advanced-stage high-grade serous EOC. The follow-up period was 24 months. The clinical characteristics are shown in Table 4. Twelve patients were positive for HRD status, as defined by the WES-based test. The median age was 57.5 years old in patients with a positive HRD status and 64 years in those with a negative HRD status. The median pretreatment CA-125 level was 1211 U/ml in patients with positive HRD and 1206 U/ml in those with negative HRD. All patients underwent primary debulking surgery, with R0 resection in three patients, R1 in six patients, and R2 in three patients with positive HRD. Among the negative HRD patients, R0 was observed in six patients, R1 in three patients, and R2 in two patients. All patients received adjuvant platinum-based and paclitaxel chemotherapy, and platinum-sensitive response was noted in 12 (100%) patients with positive HRD patients and three (27.2%) of negative HD patients. As shown in Figure 3, six positive and five negative HRD patients received maintenance PARPi. The median interval of PARPi was 17 months in patients with positive HRD and 3 months in those with negative HRD results. Eight patients with positive HRD and five with negative HRD had cancer recurrence, and the median PFS was 14.5 months in patients with positive HRD and 4 months in patients with negative HRD.