ctDNA detection in stage I-III oesophageal adenocarcinoma using a tumour agnostic approach
We analysed plasma samples from a cohort of patients with OAC (n=57) who were treated with curative intent (Figure 2). The clinicopathological characteristics are show in Table 1 and Supplementary Table 1.
To determine whether ctDNA isolated from both baseline (B1) and post-treatment (B2) plasma samples had prognostic significance, we used a tumour agnostic approach (Figure 1). This method gives insight into whether ctDNA can be a useful clinical tool when the patient tumour sample is not available for genomic reference. The median time between the B1 and B2 blood samples was 2.99 months (range 1 to 7 months).
Using a CAPP-seq 77 gene pan-cancer panel (33), we isolated, sequenced and analysed 114 plasma samples to detect tumour derived somatic ctDNA variants. We applied stringent filtering criteria to minimise the false-positive calls in the analysis. To eliminate CHIP variants and germline variants, we excluded variants with VAF >0.4. This filtering method is essential as CHIP variants are present in blood but are not tumour derived.
The median ctDNA concentration was 7.88ng/ml (range 1.67 to 126.51 ng/ml, Table 1) at baseline and 8.02 ng/ml (range 1.71 to 115 ng/ml) in the post-treatment bloods. When we assessed the clinical features, we found no statistical association between ctDNA concentration and either clinical stage or pathological tumour size (data not shown). A pileup approach identified ctDNA variants in 22/57 patients (38%). We detected a total of 90 ctDNA variants (range 0 to 7 per sample) occurring in the pre- and post-therapy plasma samples (Figure 3). Recurrent variants (n=14) between the baseline and post-treatment blood samples were observed in 12/57 (21%) patients.
ctDNA detection in stage I-III oesophageal adenocarcinoma using a tumour informed approach
Studies sequencing peripheral blood mononuclear cells (PBMCs) have shown CHIP variants may be detected in up to 23% of OAC patients (31). This method identifies CHIP variants so that they can be excluded from analysis, however, they cannot verify whether the remaining ctDNA variants are genuinely tumour derived. We used a tumour informed approach, profiling 185 matched plasma and tumour samples to determine whether ctDNA variants originated from the tumour sample.
Pre-treatment ctDNA and primary tumour samples from 39 patients were sequenced using the AVENIO pan-cancer 77 gene panel. We detected a total of 56 variants (median 2 per patient, range 1 to 9) in the primary tumour samples of 26/39 (67%) patients (Figure 3). In a subset of patients (33%, n=13), no variants were detected in the tumour samples. For 17 patients, an additional biopsy from the same tumour was sequenced to assess the macroscopic tumour heterogeneity. The spatially distinct tumour biopsies (1 to 2 cm apart) are referred to as T1, T2, T3 (Figure 1). In 94% (16/17) patients, where 2 tumour biopsies were sequenced with the AVENIO platform, we identified a variant in at least one of the tumour biopsies. In contrast, when only one tumour biopsy was sequenced, variants were observed in only 45% (10/22) of patients. Sequencing 2 tumour biopsies per patient increased the likelihood of identifying variants that might otherwise be missed due to ITH.
WGS/WES data (n=26) was available for 18 patients. WGS data was available for multiple tumour sites in 6/18 patients (Figure 4A, Supplementary Table 1). We focussed on the 77 genes that were represented on the AVENIO panel. For each patient, we performed a sequence pileup to compare the sequence data from the primary tumour with all available plasma sequence output. In total, we detected 68 tumour variants in WGS/WES, 22% (15 variants) were aligned with ctDNA calls from the same patient (Figure 4A-B). Patients with these variants (n=10) are referred to as ‘shedders’ as there is evidence of the tumour shedding DNA into blood.
We assessed the concordance between the AVENIO platform and the WGS/WES data. The AVENIO platform provided data on a select panel of genes sequenced to a great depth (average 1955x for the tumour, average 5454x for ctDNA, SupFigure 1). In contrast, WGS/WES provided sequence calls across the entire tumour genome but to a lower depth (average 50x). When we compared WGS/WES data to the AVENIO calls, 74% of the variants were in alignment (Figure 4A).
Patient SOG062 had two variants, in genes TP53 and KRAS, that were concordant between tumour regions and across WGS and AVENIO platforms (Figure 4A). Of note, only the KRAS variant was detected in the baseline blood. Similarly, SOG104, SOG179, SOG203 and SOG501 had concordant variants across platforms and multiple tumour regions. For SOG069, the AVENIO platform detected two distinct KDR variants which were not seen in the WGS data (Figure 4A) and were not detected in the plasma. Similarly, in SOG143, the AVENIO platform detected three variants, occurring in the tumour and the plasma, which were not observed in the WGS data.
Patient SOG066 and SOG083 had no tumour variant detected despite sequencing 2 tumour biopsies on each platform. The biopsies sequenced using WGS had high tumour content, tumour cellularity ranging from 47% to 60%. The biopsies sequenced using the AVENIO platform had cellularity ranging from 22% to 60%. Focussing on the 77 gene panel, there was an absence of variants in these cancer-related genes on both platforms, showing concordant results. Similarly, SOG506 had no variants detected using WES or the AVENIO platforms.
OAC tumour heterogeneity
To determine whether heterogeneity is an obstacle to the development of ctDNA as a clinical tool, we aligned the variant profile of patients with multiple tumour biopsies sequenced (Figure 4). Five patients were sequenced using the AVENIO capture alone. SOG427 had an RB1 variant which was only detected in one of the biopsies as well as in the baseline blood. In SOG415, T2 had three variants in TP53, PIK3CA and CDKN2A. Only two of these were seen in T1. However, all three variants were detected in the pre- and post-therapy blood samples. SOG443 (Figure 4) contained the highest number of variants and showed heterogeneity across both tumour and blood samples. Three variants were detected in T2 (genes BRCA1, PIK3CA and NTRK1). Only two of these were shared with T1. The pre-therapy blood was found to carry all three variants while the post-therapy blood only contained the NTRK1 variant. For 20 patients, we had multiple tumour biopsies sequenced using a combination of the AVENIO and WGS/WES platforms (Figure 4). 50% of patients (10/20) had concordant variants between the 2 biopsies. 40% of patients (8/20) had individual variant specific to one biopsy. In two patients, no variant was observed.
The tumour from patient SOG143 was heterogeneous as T1 and T2 had a common PTCH1 variant but individual variants in BRCA2 and TP53 genes in T1 and T2, respectively (Figure 4). In the corresponding blood samples, only the PTCH1 variant was detected. The tumour from SOG221 also displayed heterogeneity. While the TP53 variant was seen in all biopsies analysed (across both platforms), the RNF43 variant was only detected in 2 of 3 biopsies. None of these variants were observed in the blood samples.
ctDNA-positivity as a prognostic biomarker
It has been demonstrated that ctDNA from post-treatment bloods has prognostic value in the treatment of oesophageal cancer(31, 32). First, we sought to determine whether ctDNA positivity has prognostic value when applying the tumour agnostic approach. We analysed blood samples taken post-treatment, either after neoadjuvant therapy or surgery. We detected 35 variants (range 0 to 7 variants per patient, Figure 3) in the post-treatment blood samples. Univariate survival analyses demonstrated ctDNA positive patients had significantly worse DSS (median 58.8 months, range 7.4-92.5 months, *p=0.0130, log rank, Figure 5A) and a trend towards worse PFS (median 23 months, range 1.9 to 92.5 months, p=0.1017, log rank, Sup Figure 2) compared with ctDNA negative patients (median DSS not reached, range 5.1 to 80.8 months and median PFS not reached, range 2.1 to 80.8 months). There was no association between baseline blood samples and PFS or DSS (data not shown).
Various ctDNA studies have used a cut-off of 2 variants to be classified ctDNA positive(31, 38). We repeated the survival analysis splitting the cohort into 2 groups: patients with 0 or 1 variant versus patients with > 1 variant and the outcome remained significant (median 13.5, range 7.8 to 90.5 months, *p=0.0205, log rank, Supplementary Figure 2). Considering 3 categories: 0 variant, 1 variant, and >1 variant, a clear separation was observed between the groups (*p=0.0175). In our data, patients with at least 1 ctDNA variant had worse DSS than the 0 variants group, therefore patients with a minimum of 1 variant were considered ctDNA positive.
We assessed whether the number of ctDNA variants was associated with the pathological tumour size. We found that patients with small tumours, <10mm, were less likely to have detectable variants in their post-treatment plasma. However, this did not reach significance (p=0.1211, Sup Figure 2). We next examined whether tumour regression was associated with the number of ctDNA variants. We compared patients that were ctDNA positive to those that were negative. Patients with major histological response (90-100% tumour regression) were more likely to be ctDNA negative (p=0.0556, chi-square test, Figure 5B).
We then sought to determine whether the tumour informed approach to ctDNA detection would have prognostic significance. We analysed blood samples taken during the peri-operative period, either after neoadjuvant therapy or surgery. Patients with detectable variants in their tumour (n=30) were split into 2 groups, shedders and non-shedders, where shedders had ctDNA variants verified by the primary tumour genomic profile. A third group was defined as patients with no variant detected in their tumour, n=16 (Supplementary Figure 3). The non-shedders and the patients with no variant detected were pooled as univariable survival analysis indicated they had the same DSS (median survival not reached, range 5.1-92.6 months, p=NS, log rank). We examined the post-treatment bloods and found that shedders had worse DSS (median survival 10.6 months, range 7.99-37.55 months, ***p=0.0007, log rank, Figure 5C) and worse PFS (median survival 9.27 months, range 2.1-37.55 months, *p=0.0311, log rank, Supplementary Figure 3). There was no association between baseline blood samples and survival (data not shown).
We next assessed whether the presence of the primary tumour at the peri-operative time point influenced the ability to detect ctDNA variants. When patients had a complete response and the tumour was absent at blood collection, there was no significant correlation between shedders and non-shedders/no variant detected (Supplementary Figure 3). However, when the tumour was present, shedders had worse DSS (*p=0.0473, log rank, Figure 5D).
ctDNA detection during following treatment and disease recurrence
For 10 patients, we analysed multiple blood samples, one baseline sample and several post-treatment blood samples (Figure 2). Five patients were alive with no sign of recurrence (NSR). Of this group, 3 patients had no variant detected in any of their blood samples using the AVENIO platform (Figure 6). SOG062, the longest survivor, had no variant above the threshold for calling in the post-treatment blood samples, however, the KRAS variant identified by WGS and AVENIO tumour sequencing was detected with a low number of reads in the plasma. The KRAS variant was higher at the time of diagnosis. Post chemotherapy, the VAF reduced but was still detectable. After surgery, at 60 months from diagnosis, the KRAS variant was present but remained below the threshold for calling, in this long-term survivor. SOG315 had an APC variant at baseline. At all post-treatment time points, this variant was detectable at low levels, below the threshold for calling. In the blood sample taken after surgery, 21 months from diagnosis, a TP53 variant was identified at low levels. Neither variant was confirmed in the primary tumour, however only one tumour biopsy was available for sequencing.
Three patients were alive with disease. SOG425 had 6 variants detected in the blood at baseline. None of these variants increased in the post-treatment samples, however the longest time point available was only 1 month after resection. Similarly, SOG460 had very short follow-up time points, however, a MET variant was detectable in all samples at low levels. In SOG317, we observed a PIK3CA variant present at baseline and post-treatment blood samples. The VAF increased at 21-months post-surgery reflecting the disease progression which was confirmed by CT scan. Two patients died of their disease (SOG490 and SOG529). Both patients presented low levels of ctDNA variants in their baseline blood sample. However, post-treatment, high levels of several ctDNA variants were detected in their blood (Figure 6).