Forty patients aged sixteen or older were chosen for this study. They were newly diagnosed with B-cell acute lymphoblastic leukemia (B-ALL) according to the WHO classification of tumors of hematopoietic and lymphoid tissues in 2016. Among the 40 patients, 20 tested positive for both Ph-chromosome and BCR::ABL1 fusion transcript, while the remaining 20 tested negative for both. All patients underwent comprehensive diagnostic evaluations, including complete blood count, morphological assessment, immunophenotyping, cytogenetic analysis, and molecular biology testing. All results were fully documented. The cytogenetic and molecular biology techniques followed standard protocols. The patients were enrolled at Cho Ray Hospital from March 2021 to September 2022. The study was approved by the Cho Ray Hospital Institutional Review Board (approval number 1170-BVCR-HĐĐĐ), and each patient provided signed informed consent.

Analysis We used the ClearLLab 10C system (Beckman Coulter, California, USA) for the diagnosis and classification of immunophenotype in patients, including ClearLLab 10C B cell tube (Cat no B96805) and M2 cell tube (Cat no B96808) according to the manufacturer's recommendations. Additionally, a single staining tube including CD66c (Cat no IM2039U), CD25 (Cat no IM0479U), CD22 (Cat no A60791), CD45 (Cat no B36294) was used, along with another tube including cyTdT (Cat no IM3524U), cyMPO (Cat no IM3455U), cyCD79a (Cat no A60793), cyCD3 (Cat no A66329), CD45 (Cat no B36294), to complement the diagnostic panel. Samples were acquired on the Navios EX cytometers using acquisition protocols provided by the manufacturer. The results of the immunophenotype analysis are collected using Kaluza software version 2.0 and analyzed by a professional hematologist.

The t(9;22) translocation or Ph-chromosome was detected in the bone marrow sample using the FISH technique with the Vysis BCR/ABL1/ASS1 Tri-color DF FISH Probe Kit (Cat no 05N5420, Abbott Molecular, Illinois, USA) according to the manufacturer's instructions. Briefly, the white blood cells were collected and incubated with a 0.075 M potassium chloride (KCl) solution and Carnoy's fixation solution. After that, the cells were dropped on the marked area of a charged slide and incubated with a 10 µl probe mixture at 75°C/3 mins and 37°C/16–20 hours. Finally, the slide was washed with SSC/NP-40 solution and stained with DAPI-II solution before being analyzed using the Bioview fluorescence microscope system (Abbott Molecular, Illinois, USA). Slides were automatically scanned on the system, and the results were analyzed by professional molecular biology and genetics technicians. Positive results are calculated as the percentage of abnormal cells over the total number of analyzed cells (at least 200 cells) based on signal images captured by the system.For the BCR::ABL1 fusion transcript detection, total RNA from the bone marrow sample was extracted using the QIAamp RNA Blood Mini Kit (Cat no 52304, Qiagen, Hilden, Germany). The BCR::ABL1 fusion transcript was confirmed by the RQ-PCR technique using the ipsogen BCR::ABL1 Mbcr IS MMR DX Kit (Cat no 670823, Qiagen, Hilden, Germany) for the e13a2 and e14a2 transcript (major variant) and the ipsogen BCR::ABL1 mbcr (Cat no 670023, Qiagen, Hilden, Germany) for the e1a2 transcript (minor variant) according to the manufacturer's instructions. PCR reactions were performed on the RotorGene Q 5Plex HRM platform and analyzed using RotorGene Q 2.0 software.

The Chi-square or Fisher’s exact tests were used to compare frequencies, while the Kruskal–Wallis rank test was used to compare the density expression of each antigen between groups. The Bayesian Model Averaging (BMA) statistic was applied to identify the markers associated with BCR::ABL1 rearrangement and the optimal diagnostic model. Logistic regression was employed to construct the receiver operating characteristic (ROC) curve and define the cut-off point, sensitivity, specificity, and the value under the ROC curve (area under the curve: AUC) of each antigen and the optimal diagnostic model for B-ALL with BCR::ABL1 fusion transcript. All data analyses were completed using R statistical software v.3.5.1 (R Foundation, Austria). A p < 0.05 was considered statistically significant.

Figure 1

Abbreviations: -ALL: B-cell Acute Lymphoblastic Leukemia; BCR::ABL1: Breakpoint Cluster Region - Abelson 1; WBC: White Blood Cell

In this study, we diagnosed B-ALL by complete blood count, morphology, and flow cytometry panel using 24 immunophenotype markers. The presence of BCR::ABL1 fusion transcript was detected by RQ-PCR testing. By analyzing many factors related to BCR::ABL1 fusion transcript in B-ALL, we found that four potential factors: WBC count (p = 0.004), CD25 (p = 0.002), CD38 (p = 0.003), and CD66c (p < 0.001), which were closely associated with BCR::ABL1 rearrangement in B-ALL.

In the univariable logistic regression, B-ALL with BCR::ABL1 fusion transcript was differentiated from B-ALL non-BCR::ABL1 with an accuracy of 76.5% (AUC = 0.765, cut-off: ≥ 38.6 × 10⁹/L) by WBC count, 72.2% by CD25 (AUC = 0.722, cut-off: ≥ 24% cells), 77.8% by CD38 (AUC = 0.778, cut-off: ≤ 84% cells), and 80.8% (AUC = 0.808; cut-off: ≥ 20% cells) by CD66c (Table 1).

Table 1

Factor	Cut-off	AUC (CI 95%)	Sensitivity, % (CI 95%)	Specificity, % (CI 95%)
WBC	≥ 38.6 x 109/L	0.765 (0.611-0.919)	75.0 (55.0-90.0)	80.0 (60.0-95.0)
CD25	≥ 24% cells	0.722 (0.599-0.846)	50.0 (30.0-70.0)	95.0 (85.0-100.0)
CD38	≤ 84% cells	0.778 (0.623-0.932)	70.0 (50.0-90.0)	80.0 (60.0-95.0)
CD66c	≥ 20% cells	0.808 (0.674-0.941)	75.0 (55.0-90.0)	90.0 (75.0-100.0)

Youden’s index identified the cut-off value, Area Under the Curve (AUC) value, sensitivity, and specificity of 4 potential factors in B-ALL with BCR::ABL1 rearrangement: WBC count, CD25, CD38, and CD66c

Figure 3

Abbreviations: BMA: Bayesian Model Averaging; WBC: White Blood Cell; AUC: Area Under the Curve; B-ALL: B-cell Acute Lymphoblastic Leukemia; BCR::ABL1: Breakpoint Cluster Region - Abelson 1