Improving minimal residual disease detection in precursor B-ALL based on immunoglobulin-kappa and heavy-chain gene rearrangements.

Patient-specific rearrangements of immunoglobulin and T-cell receptor genes (Ig/TCR) are currently used as markers of minimal residual disease (MRD) in patients with acute lymphoblastic leukaemia (ALL) enrolled on clinical trials in both Europe and Australia.1 Most laboratories in Europe utilize the generic Taqman probes published in Leukemia for the Junctional (J) region of the immunoglobulin heavy-chain gene (IGH) and the immunoglobulin--deleting element (IGK-Kde) to provide effective real-time quantitative (RQ) PCR-based detection of these rearrangements.2, 3 Two MRD markers with high sensitivity (10-4) are generally required in MRD intervention clinical trials to reduce false-negative results arising from clonal evolution,1, 4 and although IGK-Kde markers have been recommended for use as second markers,2 they have been under-utilized due to the lower sensitivity of some assays.1 We have trialled new MRD probes for both IGK and IGH with the aim of improving marker sensitivity and the efficiency of identifying suitable MRD markers (Table 1). These probes were designed to target the consensus Variable (V) region family sequences instead of the J or Kde gene segments and in most instances were based on minor groove binding or locked nucleic acid chemistry, which can provide better binding efficiency than the conventional Taqman probes. A comparison of the MRD probes was carried out in conjunction with the routine identification and optimization of MRD markers performed for a cohort of 400 patients enrolled on the Australian and New Zealand Children's Haematology Oncology Group Study 8. We compared the performance of the new probes with the published Taqman probes specific for IGH J1245, J3 and J6 sequences,3 which were used for some patients with IGH(V–J) rearrangements as well as all patients with IGH(VH5-J) or immature IGH(D–J) rearrangements. For IGK rearrangements, new probes, of which each bind a consensus sequence for a family of Vk sequences (VK1, VK2 and VK3) and an IGK(intron) probe were compared with the single generic IGK(Kde) probe used by most laboratories for all IGK-Kde rearrangements.2 Each RQ-PCR assay was based on a pair of patient-specific primers designed to amplify a clonal rearrangement identified in the patient's diagnostic sample by routine PCR and sequencing.1, 4, 5 The 50-l reactions, which included 500 ng of DNA, 250 ng of each primer, 2.5 pmol of probe and the manufacturer's mastermix, were run on an ABI PRISM 7700 or a Biorad IQ5 platform at 95 °C for 10 or 3 min followed by 50 cycles of 15 s at 95 °C and 45 s at the optimal annealing temperature (57–67 °C). Established ESG-MRD-ALL guidelines were used to assess the optimized RQ-PCR standard curves and to estimate the quantitative range as a measure of assay sensitivity.6 Our analysis of the sensitivity of assays achieved with the different probes targeting different gene segments is shown in Figure 1a for IGH and Figure 1b for IGK. A higher proportion of the IGH MRD tests based on V region probes were sensitive (83 versus 59% of mature IGH rearrangements tested with IGH(J) probes; P<0.0001, n=418, Fisher's exact test)). Similarly, the IGK(V) probes were also sensitive in more patients than the generic IGK(Kde) probe (67 versus 50% of IGK-Kde rearrangements; P<0.05, n=130, Fisher's exact test). This enhancement probably arises from a combination of both improving the probe chemistry, which has been reported previously for minor groove binding, versus non minor groove binding probes for IGH,7 and the greater selectivity of probes targeting the V gene segments. It can be argued that gene segments that are less common in rearrangements make better probe targets from the point of view of sensitivity, although they will be useful for fewer patients. In support of this, a further analysis comparing the J6 probe with the more commonly used J1245 probe showed that it was more likely to provide sensitive assays (71 versus 57%; P<0.05, n=216, Fisher's exact test). It is also clear from Figure 1a that the more generic J1245 and Kde probes are functional but result in a higher incidence of less-sensitive markers, which are less useful but still have some value for identifying high-risk patients. Neither of the probes used with IGK intron-Kde rearrangements performed well (Figure 1b). This is consistent with the lower sensitivity of IGK(intron-Kde) markers described in a much larger study by Flohr et al.,1 who proposed an order of preference for MRD markers of IGH > IGK(Vk-Kde) > TCR- (TCRD) > TCR- (TCRG) and IGK(intron-Kde) markers on the basis of availability and sensitivity. In view of its poorer performance and utility for fewer rearrangements, the new Taqman IGK(intron) probe is not recommended to other laboratories. Identification of low- or standard-risk patients in clinical trials usually requires two sensitive MRD markers. A more practical and empirical measure of the value of the new MRD probes is therefore the proportion of RQ-PCR assays based on different probe types that were selected as one of the best two tests for patient MRD stratification. It should be noted that in addition of being tested for IGH and IGK rearrangements, these patients were also screened for rearrangements of the TCR genes and (TCRG, TCRD), and in some cases, TCR-- (TCRD-A) and TCR- (TCRB) markers. On average, four Ig/TCR markers were tested in RQ-PCR assays for each patient, and the two best markers were selected mainly on the basis of sensitivity for the MRD studies in the clinical trial. For both IGH and IGK markers, the probe group had a significant effect on whether an MRD marker was selected for usage in the clinical trial. For the mature IGH markers, 71% of MRD tests based on IGH(V) probes were used compared to only 41% using IGH(J) probes (P<0.0001). For IGK Vk-Kde rearrangements, 70% of tests based on IGK(V) probes were selected and 40% of those with the IGK(Kde) probe (P=0.0018). This selective advantage was also borne out in the direct comparisons for those individual markers in which both probes were trialled, reinforcing our conclusion that the new probes usually provided better sensitivity. From a practical point of view, there are two potential disadvantages to be considered in association with introducing these new probes. First, all of these fluorescent probes are expensive and the original IGH(J) probes and IGK(intron) probes would still be needed for MRD for immature IGH and IGK(intron-Kde) markers and to detect rearrangements involving rare V regions (VH5, VK4, VK5, VK6 and VK7). This additional outlay may not be justifiable for laboratories with small numbers of patients. Second, MRD markers based on IGH(V) probes may be subject to false negativity due to IGH V-region substitution during clonal evolution, reinforcing the established need to use a second marker for MRD detection and monitoring.4 Our previous evaluation of V versus J probes for TCRG found no benefit compared with J probes,8 and Uchiyama et al.7 have demonstrated the value of minor groove binding compared to Taqman probes for MRD; so we conclude that in seeking to improve probes for other markers, it may be more important to adopt probes with higher binding efficiency (at little extra cost) than to change the probe target. In summary, the inclusion of a variety of probes in our routine RQ-PCR optimizations for MRD testing of patients enrolled on Australian and New Zealand Children's Haematology Oncology Group Study 8 has enabled us to compare the effectiveness of different probes for both IGK and IGH and to establish the value of probes with higher binding efficiency targeting the V regions of the rearrangements. Further research is required to establish better options for IGK(intron-Kde) rearrangements. The approach outlined in this paper should also be valuable for assessing MRD probes for other rearrangements such as TCRG and TCRD, thus improving the overall efficiency of MRD monitoring. We acknowledge the NH&MRC, the Cancer Council of NSW, private donors and the Leukaemia Foundation for their financial support of MRD studies and also the Australian and New Zealand Children's Haematology Oncology Group Study 8 committee and participating clinicians and the I-BFM-SG. Standardization and quality control for MRD testing is supported by the ESG-MRD-ALL. Children's Cancer Institute Australia for Medical Research is affiliated with both the University of New South Wales and Sydney Children's Hospital.