False-negative rates for MYC fluorescence in situ hybridization probes in B-cell neoplasms

As MYC rearrangements in high-grade B-cell neoplasms are associated with poor prognosis, interphase fluorescence in situ hybridization (FISH) testing for MYC has become part of the routine clinical evaluation of high grade B-cell lymphomas (HGBCL), diffuse large BCL (DLBCL) and plasma cell neoplasms (PCN).51 This study examined BCL and PCN cases in which FISH for MYC rearrangements was evaluated by concurrent MYC breakapart (BAP) and MYC/IGH dual fusion (D-FISH) probe sets (both commercially available), to assess whether a screening strategy using MYC BAP or D-FISH alone is sufficient for detecting MYC rearrangements. Our results indicate high false negative (FN) rates using either approach alone (4.1% MYC BAP and 22.1% MYC-IGH), suggesting that the use of both FISH probe sets is superior for detection of MYC rearrangements in BCL and PCN than either individual probe. Additionally, mate-pair sequencing (MPseq) uncovered multiple cryptic and complex mechanisms providing the underlying genomic architecture and potential mechanisms that may ultimately lead to improved methods of detection.

BCL with MYC and BCL2 and/or BCL6 rearrangements has been included as a distinct entity in the 2017 World Health Organization (WHO) classification (HGBCL with MYC and BCL2 and/or BCL6 rearrangements).4 These patients have a poor prognosis and require more intensive therapy.65 MYC rearrangement alone also portends poor prognosis in some studies.72 In plasma cell myeloma (PCM), secondary MYC rearrangement is associated with aggressive disease.8 Thus, it has become standard-of-care in the workup of DLBCL, HGBCL, and PCM to perform FISH evaluating for MYC rearrangement.

Fluorescence in situ hybridization testing for MYC on formalin-fixed, paraffin-embedded tissue (FFPET) is the preferred and most widely used clinical laboratory method. In approximately 60% of MYC-rearranged BCL cases, MYC is translocated to an immunoglobulin gene (IG), either the heavy chain (IGH) (most common), lambda light chain (IGL), or kappa light chain (IGK).64 Therefore, MYC rearrangements are detectable either by a BAP or D-FISH probe strategy looking specifically for a MYC/IGH (or -IGL, -IGK) translocation. The MYC break-point can vary widely, particularly if the partner gene is not IGH.9 Thus, it is not surprising that MYC BAP probes have been shown to be more sensitive than MYC D-FISH probes and are often used in isolation for screening.1110 Here we conducted a large retrospective analysis of cases in which FISH was performed concurrently using both MYC BAP and MYC/IGH D-FISH to evaluate for rearrangements detectable with one probe but with a normal result for the other probe. We then used MPseq to analyze a subset of cases in which MYC BAP was falsely negative, in order to understand more fully the biological mechanism of these unusual MYC rearrangements.

From January 2006 to June 2011, we performed both the MYC BAP and MYC/IGH D-FISH probe sets on 2518 FFPET and 971 fresh liquid [peripheral blood (PB), bone marrow (BM), and body fluids] specimens, yielding a total of 3489 cases in our study. Commercial MYC BAP and MYC/IGH D-FISH probe sets (Abbott Molecular, Des Plaines, IL, USA) are used in our laboratory. All specimens were subjected to standard specimen-specific laboratory protocols (Online Supplementary Appendix 1).

A total of 266 out of 2518 (10.5%) FFPET and 98 out of 971 (10.1%) liquid specimens were positive for MYC/IGH fusion using the D-FISH probes. Of these, 246 FFPET and 72 liquid specimens also had MYC BAP results, yielding a total of 318 cases that had successful results for both probe sets. MYC BAP FISH was negative in 9 out of 246 (3.7%) FFPET and 4 out of 72 (5.6%) liquid specimens in which MYC/IGH fusion was detected, resulting in a FN rate overall for the MYC BAP of 4.1% (13 out of 318). Of these 13 cases with negative MYC BAP, 10 showed two intact MYC signals, and 3 demonstrated three intact signals due to gain of chromosome 8 or the MYC region when using the MYC BAP. Of the 318 cases, we evaluated 175 cases with a BCL2 BAP probe and 39 (22.3%) demonstrated a rearrangement; 115 cases were evaluated with a BCL6 BAP probe and 8 (7%) demonstrated a rearrangement. One case had a concurrent BCL2 and BCL6 rearrangement.

When the indication for FISH is DLBCL/HGBCL, an Ig gene partner is detected in approximately 60% of MYC-rearranged cases and non-IG partner in 40%.6 Of the MYC-Ig cases, 75% have MYC/IGH fusion, while 25% show fusion with either IGL or IGK.6 Therefore, if 4.1% of MYC/IGH fusion cases are associated with FN MYC BAP probe results, and MYC/IGH fusion accounts for only 45% of MYC rearranged cases, these data suggest there may be additional FN results associated with non-IGH partners.

To address the opposite scenario, in which MYC rearrangements were identified by the MYC BAP probe but had a signal pattern not indicative of fusion using the MYC/IGH D-FISH probe, we identified an additional 131 FFPET cases from 2016. Of these 131 cases with a positive MYC BAP probe result and a negative MYC/IGH fusion result, 102 cases (77.9%) had additional MYC signals in relation to centromere 8 FISH probe signals using the MYC/IGH probe, suggesting a MYC rearrangement with a non-IGH partner. Of these 102 cases, 9 had MYC/IGK fusion, 26 had MYC/IGL fusion, and 67 were interpreted as MYC/non-Ig fusion positive. The remaining 29 cases (5 with MYC/IGK fusion, 1 with MYC/IGL fusion, and 23 with MYC/non-Ig fusion per subsequent analysis) had both a positive MYC BAP result and a “normal” MYC result using the MYC/IGH probe set (no evidence of MYC/IGH/IGH fusion nor additional MYC signals in relation to centromere 8), resulting in a 22.1% (29 out of 131) FN rate for MYC using the MYC/IGH probe set.

We sought to resolve the genomic rearrangements of four discordant FISH cases (normal MYC BAP/abnormal MYC/IGH) using MPseq, a novel next-generation sequencing-based technique that allows the detection of structural rearrangements with higher resolution compared to FISH, thus enabling characterization of precise rearrangement breakpoints1312 (Online Supplementary Appendix 1).

Cases 1 (Figure 1A) and 2 (Figure 1B) uncovered similar reciprocal translocations placing the MYC gene distal to the IGH region on the derivative chromosome 14. While the chromosome 8 breakpoints for both cases were centromeric to the 5’MYC BAP footprint resulting in a FN MYC result using the MYC BAP probe, a single fusion signal was observed using the MYC/IGH D-FISH probes due to the proximity of the IGH and MYC genes. In case 3 (Figure 2A), MPseq characterized a 212 Kb genomic segment of the IGH variable and diversity regions inserted in an inverted orientation within 8q24.21 immediately distal to the PVT1 gene downstream of the MYC gene. Because the inverted segment of IGH was inserted distal to MYC, the 5’ and 3’ MYC BAP footprints remained intact on both the normal and derivative chromosome 8. This insertional mechanism would be cytogenetically “cryptic” and also explains the presence of a single MYC/IGH fusion signal. Case 4 (Figure 2B) also demonstrated a translocation between the variable and diversity segments of the IGH gene and a region telomeric to the MYC gene within the PVT1 gene. A tandem duplication of a 245 kb portion of 8q24.21 containing a portion of CASC11, MYC and a portion of the PVT1 genes was also observed. These results provide detailed visualization of three different genomic mechanisms resulting in the translocation of portions of IGH in close proximity to MYC, resulting in MYC/IGH fusion, yet eluding detection with a MYC BAP probe. Common breakpoints near or within long non-coding RNA sequences surrounding the MYC gene8 were observed in each case (Online Supplementary Appendix 2).

Figure 1.Chromosomal rearrangement mechanisms for patients 1 and 2 with discordant MYC BAP and MYC/IGH D-fluorescence in situ hybridization (FISH) results. Reciprocal translocations in case 1 (A) and case 2 (B) (breakpoints indicated by horizontal dashed red lines) had atypical breakpoints located centromeric to the MYC BAP footprints (striped red and green rectangles located on chromosome 8). However, a single MYC/IGH fusion on the derivative chromosome 14 was observed by the MYC/IGH D-FISH probe set (solid red rectangle located on chromosome 8 and solid green rectangle located on chromosome 14) in both cases. The normal MYC BAP FISH result is explained in both cases since the chromosome 8 breakpoint is centromeric to the FISH probe. Variable (V), diversity (D), joining (J) and constant (C) regions of the IGH locus are indicated.

Figure 2.Chromosomal rearrangement mechanisms for patients 3 and 4 with discordant MYC BAP and MYC/IGH D-fluorescence in situ hybridization (FISH) results. (A) In case 3, a 212 Kb genomic segment of the IGH constant region was inserted in an inverted orientation distal to the PVT1 gene downstream of MYC, thus producing a single fusion by the MYC/IGH D-FISH probe set. (B) In case 4, a 245 Kb genomic segment of the MYC gene and a portion of the PVT1 gene is duplicated and inserted into chromosome 14 distal to the IGH constant region and diversity and joining gene segments. Numerous additional portions of chromosome 8 were translocated into the derivative chromosome 14 distal to the MYC gene. Due to the size and/or complexity of the rearrangement, and the lack of clone coverage at the rearrangement breakpoints, neither of these structural abnormalities would be detected by the MYC BAP FISH probe. Variable (V), diversity (D), joining (J) and constant (C) regions of the IGH locus are indicated.

Factors such as cost, efficiency, and a desire for information about the translocation partner all impact the MYC FISH strategy chosen by a given pathologist or clinician. Literature review, as well as anecdotal experience, suggest that screening with a MYC BAP alone is the most widely used method in both BCL and PCM, and that IG D-FISH probes, if evaluated, are performed only if MYC BAP reveals a positive result.5421 Precise FN rates using the MYC BAP alone are not widely publicized, although our results are similar to previous reports.1514 As our laboratory uses the most sensitive commercial MYC BAP, these estimates of FN rates may actually be conservative.10 Furthermore, determining the significance of extra MYC signals seen with a MYC/IGH D-FISH probe used in isolation (MYC rearrangement in which the partner is not IGH vs. extra intact copies of MYC) requires the concurrent use of a robust MYC BAP probe.10

In summary, given the importance of detecting MYC rearrangements in BCL, awareness of the FN rate of the commonly used detection method is of critical importance to practising hematologists, pathologists, and geneticists. Our results indicate an at least 4% FN rate when screening with MYC BAP alone, which can be mitigated by performing MYC/IGH D-FISH. MYC/IGH D-FISH alone is also inadequate due to high FN rates (22.1%) and the significant proportion of non-IGH MYC translocations. Although only available on a small subset of cases, and limited by the lack of clinical information available to us, MPseq reveals a genomic complexity, and this fresh insight could lead to improved methods of detection.