Circulating Plasma Tumor DNA Is Superior to Plasma Tumor RNA Detection in Ewing Sarcoma Patients: ptDNA and ptRNA in Ewing Sarcoma

Abstract

The detection of tumor-specific nucleic acids from blood increasingly is being used as a method of liquid biopsy and minimal residual disease detection. However, achieving high sensitivity and high specificity remains a challenge. Here, we perform a direct comparison of two droplet digital PCR (ddPCR)-based detection methods, circulating plasma tumor RNA and circulating plasma tumor DNA (ptDNA), in blood samples from newly diagnosed Ewing sarcoma patients. First, we developed three specific ddPCR-based assays to detect EWS-FLI1 or EWS-ERG fusion transcripts, which naturally showed superior sensitivity to DNA detection on in vitro control samples. Next, we identified the patient-specific EWS-FLI1 or EWS-ERG breakpoint from five patient tumor samples and designed ddPCR-based, patient-specific ptDNA assays for each patient. These patient-specific assays show that although plasma tumor RNA can be detected in select newly diagnosed patients, positive results are low and statistically unreliable compared with ptDNA assays, which reproducibly detect robust positive results across most patients. Furthermore, the unique disease biology of Ewing sarcoma enabled us to show that most cell-free RNA is not tumor-derived, although cell-free-DNA burden is affected strongly by tumor-derived DNA burden. Here, we conclude that, even with optimized highly sensitive and specific assays, tumor DNA detection is superior to RNA detection in Ewing sarcoma patients.

Figure 1

Collection methods significantly impact the sensitivity of cell-free RNA detection. The black dots represent individual results Plasma cell-free RNA investigated by droplet digital PCR (ddPCR) for the housekeeping genes ACTB and B2M shows that the use of RNase-containing collection tubes (Streck) significantly increases the yield compared with EDTA tubes. Cell-free RNA extracted by the Circulating Nucleic Acid Kit (CNA) significantly increased positive droplets compared with other extraction methods. ∗P < 0.0001.

Figure 2

Droplet digital PCR (ddPCR) can detect trace amounts of RNA and DNA from Ewing sarcoma cells. The black dots represent individual replicates within the experiment. RNA and genomic DNA were extracted from TC-71 and SKES1 cells and subjected to ddPCR analysis in serial dilution. A: Housekeeping genes ACTB and B2M can be detected using as little as input RNA as 1 pg to 100 fg. B: EWS-FLI1 fusion transcripts can be detected using as little as an input RNA as 10 pg, whereas the TC-71–specific EWS-FLI1 probes fail to produce reliable positive droplets at 50 pg.

Figure 3

Plasma tumor DNA (ptDNA) assays are significantly more sensitive than plasma tumor RNA (ptRNA) assays in newly diagnosed Ewing sarcoma patients. The black dots represent individual replicates. Circulating cell-free DNA and RNA were extracted from 1 mL plasma from five patients before starting therapy as baseline samples. For patient ST001, samples also were available from before surgery (after completion of neoadjuvant chemotherapy) and after surgery (after complete resection of primary disease). A: Cell-free DNA content was assessed by an AP3B1 assay, showing satisfactory DNA content in droplets from all five patients at diagnosis with very little variability among replicates. B: Custom designed EWS-FLI1 or EWS-ERG breakpoint ptDNA assays detect significant positive droplet events in the baseline samples for all five patients. C: Cell-free RNA content assessed by housekeeping gene (ACTB, B2M) transcript counts show a satisfactory yield of cell-free RNA. D: EWS-FLI1 or EWS-ERG fusion gene detection shows significant variation of positive signal among all replicates, with statistical significance only in patient ST005. ∗P < 0.0001. CTRL, control; Neg, negative.