Genomic cfDNA Analysis of Aqueous Humor in Retinoblastoma Predicts Eye Salvage: The Surrogate Tumor Biopsy for Retinoblastoma

Abstract

Tumor-derived cell-free DNA (cfDNA) has biomarker potential; therefore, this study aimed to identify cfDNA in the aqueous humor (AH) of retinoblastoma eyes and correlate somatic chromosomal copy-number alterations (SCNA) with clinical outcomes, specifically eye salvage. AH was extracted via paracentesis during intravitreal injection of chemotherapy or enucleation. Shallow whole-genome sequencing was performed using isolated cfDNA to assess for highly recurrent SCNAs in retinoblastoma including gain of 1q, 2p, 6p, loss of 13q, 16q, and focal MYCN amplification. Sixty-three clinical specimens of AH from 29 eyes of 26 patients were evaluated; 13 eyes were enucleated and 16 were salvaged (e.g., saved). The presence of detectable SCNAs was 92% in enucleated eyes versus 38% in salvaged eyes (P = 0.006). Gain of chromosome 6p was the most common SCNA found in 77% of enucleated eyes, compared with 25% of salvaged eyes (P = 0.0092), and associated with a 10-fold increased odds of enucleation (OR, 10; 95% CI, 1.8-55.6). The median amplitude of 6p gain was 1.47 in enucleated versus 1.07 in salvaged eyes (P = 0.001). The presence of AH SCNAs was correlated retrospectively with eye salvage. The probability of ocular salvage was higher in eyes without detectable SCNAs in the AH (P = 0.0028), specifically 6p gain. This is the first study to correlate clinical outcomes with SCNAs in the AH from retinoblastoma eyes, as such these findings indicate that 6p gain in the aqueous humor is a potential prognostic biomarker for poor clinical response to therapy.Implications: The correlation of clinical outcomes and SCNAs in the AH identified in the current study requires prospective studies to validate these finding before SCNAs, like 6p gain, can be used to predict clinical outcomes at diagnosis. Mol Cancer Res; 16(11); 1701-12. ©2018 AACR.

Conflict of interest statement

Conflict of Interest: No author has any conflict of interest to disclose.

©2018 American Association for Cancer Research.

Figures

Figure 1:

Histopathology from an eye enucleated after intravitreal injection. (a) paracentesis site in the cornea on gross examination (b) at 40x and (c) 100x. No tumor was seen in the needle tracks from the paracentesis in the cornea. The cornea tracks were well healed and without tumor. (d) The pars plana entrance of the injection site for intravitreal melphalan at 40x and (e) at 100x also showed healed tissue without evidence of tumor cells.

Figure 2:

Chromosomal CNA profiles from 13 eyes that required enucleation with available tumor tissue for comparison. The profiles demonstrate the similar genomic alterations: concordance of DNA profiles between tumor and the AH ranged from 84.3-100%. Eleven of the 13 eyes showed a near exact match of chromosomal gains and losses between tumor and AH (>90% concordance).

Figure 3:

Pearson’s hierarchical clustering matrix based on the SCNA profiles of the 58 AH and tumor samples from 21 eyes that had more than one sample available for correlation. Samples are listed as Case number_# based on the chronological order of AH sampling (e.g. 1, 2, 3). Tumor samples correlate most closely with the matched AH samples from the same eye (with the notable exception of Case 1, described in text). The majority of longitudinal AH samples also group together with few exceptions. Samples that correlate within the same eye are shown by the grey bars on the right, the black bars indicate samples that did not fall adjacent other samples from the same eye. Samples from eyes that were enucleated (e.g. surgically removed) are indicated by the red bar adjacent the dentogram, those that are salvaged (e.g. saved) are indicated in blue.

Figure 4:

(a) Composite somatic copy number alteration (CNA) profile from cell-free DNA in the Aqueous Humor (AH) samples from enucleated eyes (Enuc, red) and salvaged eyes (Salv, blue). (b). Box plot demonstrating the range of amplitude changes for the enucleated (Enuc) vs. salvaged (Salv) eyes; the black bar represents the median while the green bar represents the mean (of the ratio to median). The sample with focal MYCN gain is shown as a red asterisk in the Chr 2p plot. The median of the ratio to median amplitude of Chr 6p gain is significantly greater in enucleated eyes (p=0.001), which may be both from the increased copy number of the amplified region and an increase in the total fraction of tumor-derived DNA in the AH of enucleated eyes.

Figure 5:

Kaplan-Meier curves of eye salvage for treated eyes (e.g. no primary enucleations) at 800 days by ALL EYES (a) all eyes +/− presence of genomic instability >300 sum deviation from the median (with time from sample to event or last follow-up), regardless of clinical classification WITHIN IIRC (b) all eyes +/− the presence of RB SCNAs in the AH (with time from sample to event or last follow-up), regardless of clinical classification (c) all eyes +/− presence of gain of 6p in the AH (with time from sample to event or last follow-up), regardless of clinical classification WITHIN IIRC (d) IIRC Group (with time from diagnosis to event or last follow-up) />(e) Group D eyes +/− gain of 6p in the AH (with time from sample to event or last follow-up) (f) Group E eyes +/− gain of 6p in the AH (with time from sample to event or last follow-up) WITHIN TNM (g) TNM Group (with time from diagnosis to event or last follow-up). (h) cT2b eyes +/− gain of 6p in the AH (with time from sample to event or last follow-up) (i) cT3 eyes +/− gain of 6p in the AH (with time from sample to event or last follow-up) This demonstrates that the presence of any RB SCNA aids in prediction of globe salvage more accurately than Group Classification alone. Within the RB SCNAs, 6p gain was most predictive of risk of tumor recurrence requiring enucleation.

Figure 6:

Copy Number Alteration (CNA) profile and histogram from two cases demonstrating changes in amplitude of alterations that correlate with clinical tumor response. The CNA profiles for Case 6 (a) demonstrates increased chromosomal alterations in chromosomes 1q, 2p, 6p and 16q; additionally, 7q, 11q and 12q were altered and are shown. AH samples 1–5 were taken longitudinally separated by at least 1 week between sample. Case 6 demonstrates decreased CNA magnitude at AH sample 2 relative sample 1 which correlated with clinical response to therapy; however, these alterations then increase steadily with persistent tumor activity and this eye eventually required enucleation. The CNA profile from the tumor (shown in Figure 2) mimics the AH profile. Case 22 (b) demonstrates an opposite finding: as the tumor responded to therapy the CNA magnitude from the AH declined. This suggest both a smaller concentration of tumor-derived DNA as the tumor responds to therapy