Sensitive and Specific Detection of Ewing Sarcoma Minimal Residual Disease in Ovarian and Testicular Tissues in an In Vitro Model

Abstract

Ewing sarcoma (EWS) is a common pediatric solid tumor with high metastatic potential. Due to toxic effects of treatments on reproductive functions, the cryopreservation of ovarian tissue (OT) or testicular tissue (TT) is recommended to preserve fertility. However, the risk of reintroducing residual metastatic tumor cells should be evaluated before fertility restoration. Our goal was to validate a sensitive and specific approach for EWS minimal residual disease (MRD) detection in frozen germinal tissues. Thawed OT (n = 12) and TT (n = 14) were contaminated with tumor RD-ES cells (10, 100, and 1000 cells) and EWS-FLI1 tumor-specific transcript was quantified with RT-qPCR. All contaminated samples were found to be positive, with a strong correlation between RD-ES cell numbers and EWS-FLI1 levels in OT (r = 0.93) and TT (r = 0.96) (p < 0.001). No transcript was detected in uncontaminated control samples. The invasive potential of Ewing cells was evaluated using co-culture techniques. After co-culturing, tumor cells were detected in OT/TT with histology, FISH, and RT-qPCR. In addition, four OT and four TT samples from children with metastatic EWS were tested, and no MRD was found using RT-qPCR and histology. We demonstrated the high sensitivity and specificity of RT-qPCR to detect EWS MRD in OT/TT samples. Clinical trial: NCT02400970.

Keywords: Ewing sarcoma; RT-qPCR; fertility preservation; minimal residual disease detection; ovarian tissue; testicular tissue.

Conflict of interest statement

The authors declare they have no conflict of interest.

Figures

Figure 1

Ewing sarcoma (EWS)-FLI1 transcripts detection in ovarian tissue (n = 12). Relative quantification of EWS-FLI1 transcripts (B2M reference gene) for the contamination with 0, 10, 100 and 1000 cells. Each symbol represents one ovarian fragment (the average of the duplicates for 1000 cells or triplicates for 10 and 100 cells). The symbol ** means there was a significant difference and p < 0.001.

Figure 2

EWS-FLI1 transcripts detection in testicular tissue (n = 14). Relative quantification of EWS-FLI1 transcripts (B2M reference gene) for the contamination with 0, 10, 100, and 1000 cells. Each symbol represents one testicular fragment (the average of the duplicates for 1000 cells or triplicates for 10 and 100 cells). The symbol ** means there was a significant difference and p < 0.001.

Figure 3

Sensitivity (SE) and specificity (SP) of detection to distinguish 10 and 100 Ewing cells, and 100 and 1000 Ewing cells in ovarian tissue: (a) The AUC (area under the curve, ROC curve) was 0.94 CI 95% [0.86–1.00] to distinguish 10 and 100 Ewing cells. The optimal decision threshold, determined using Liu and Youden indexes, to distinguish between 10 and 100 EWS cells was 354 EWS-FLI1 transcripts with a sensitivity (SE) of 95% and a specificity (SP) of 86% (in red). For maximal SE (100%) and SP (100%), the cut-offs were 319 and 1150 EWS-FLI1 transcripts, respectively. (b) The area under the curve (AUC) was 0.97 CI 95% [0.92–1.00] between 100 and 1000 Ewing cells. To distinguish between 100 and 1000 EWS cells, the optimal decision threshold determined using Liu and Youden indexes was 3998 EWS-FLI1 transcripts with a SE of 100% and SP of 86% (in red). For a maximal SP (100%), the cut-off was 5528 EWS-FLI1 transcripts.

Figure 4

Sensitivity and specificity of detection to distinguish 10 and 100 EWS cells, and 100 and 1000 EWS cells in testicular tissue: (a) The AUC was 0.98 CI 95% [0.94–1.00] to characterize 10 and 100 EWS cells. The thresholds to distinguish between 10 and 100 EWS cells were 642 EWS-FLI1 transcripts (Liu and Youden indexes, SE = 92% and SP = 95%) (in red), 521 EWS-FLI1 transcripts for SE = 100% and 749 EWS-FLI1 transcripts for SP = 100%. (b) The AUC was 0.99 CI 95% [0.98–1.00] between 100 and 1000 EWS cells. The cut-offs to distinguish between 100 and 1000 EWS were 3172 EWS-FLI1 transcripts (Liu and Youden indexes, SE = 93% and SP = 100%) (in red) and 2170 EWS-FLI1 transcripts for SE = 100%.

Figure 5

Illustrations of histology: ovarian tissue (OT) after co-culture with RD-ES cells (at day 7); (a) RD-ES cells (arrow) localized in OT after co-culture stained with hematoxylin and eosin (×10); (b) Area with RD-ES cells disseminated in OT (×20); (c) Area with ovarian follicle (×20).

Figure 5

Illustrations of histology: ovarian tissue (OT) after co-culture with RD-ES cells (at day 7); (a) RD-ES cells (arrow) localized in OT after co-culture stained with hematoxylin and eosin (×10); (b) Area with RD-ES cells disseminated in OT (×20); (c) Area with ovarian follicle (×20).

Figure 6

Illustrations of histology and immunohistochemistry: RD-ES cells (black arrow) localized in testicular tissue (TT) after co-culture (at day 14) on insert; (a) TT in insert stained with hematoxylin and eosin (×20); (b) ERG-positive staining of RD-ES cells (×20).

Figure 7

Fluorescence in situ hybridization (FISH) analysis of ovarian tissue using EWSR1 (22q12) dual color break apart rearrangement probe (Vysis) showing RD-ES cells invasion (white arrows) after co-culture (at day 14) (×80). RD-ES cells displayed one fusion (yellow signal), and the simultaneous split pattern of one orange and one green signal (arrows), indicative of a rearrangement of one copy of the EWSR1 gene. The fusion gene is detected by a yellow signal, corresponding to co-localization of the red and green probes.