数字PCR检测慢性髓细胞性白血病患者国际标准化BCR::ABL1（P210）mRNA的室间比对

目的：

通过室间样本派发比对评估数字 PCR（dPCR）检测慢性髓细胞性白血病患者国际标准化 BCR::ABL1（P210）mRNA（BCR::ABL1IS）的能力。

方法：

本研究为多中心比对实验研究。采用待废弃的临床样本统一制备共 37 种不同 BCR::ABL1^IS 水平的比对样本，其中 22 种用于实时定量 PCR（RQ-PCR）检测以进行转换系数（CF）的再验证，其余 15 种加上用于 RQ-PCR 的 6 种用于 dPCR 比对。北京大学人民医院通过检测 WHO 国际标准品判定自身 dPCR 及 RQ-PCR 的 CF 的有效性。通过 Bland-Altman 一致性分析判断各实验室 CF 的有效性，若无合格 CF，则重新计算并评估有效性。

结果：

准参加 dPCR 比对的 14 家实验室中 13 家所有比对样本的 BCR::ABL1 均能检出。参加 RQ-PCR 比对的 13 家实验室中 5 家 CF 合格，其余 8 家重新计算的 CF 均合格。dPCR比对中，11 家采用同一商品化试剂盒，10 家合格；1 家采用自配试剂并利用国家临床检验中心标准品获得 CF 亦合格；无 CF 的 2 家重新计算的 CF 均合格。同时进行 dPCR 和 RQ-PCR 比对的样本结果比较显示，各家采用各自合格 CF 得出的 BCR::ABL1^IS 结果差异均无统计学意义（均 P≥0.1）；但 RQ-PCR 原有 CF 不合格的 6 家实验室中，2 家的 dPCR 检测结果与由原 CF 得出的 BCR::ABL1^IS 结果之间差异有统计学意义（均 P＜0.05），3 家趋于差异有统计学意义（0.05＜P＜0.1）。

结论：

dPCR 检测 BCR::ABL1^IS具有高敏感性、高准确性以及与 RQ-PCR 结果的高度一致性。

Objective: 

To evaluate the capability of digital PCR (dPCR) in detecting the internationally standardized BCR::ABL1 (P210) mRNA (BCR::ABL1^IS) of patients with chronic myeloid leukemia through inter-laboratory sample distribution and comparison.

Methods: 

This was a multi-center compatative trial study. A total of 37 samples with different BCR::ABL1^IS levels were uniformly prepared from patient bone marrow or peripheral blood to be discarded for comparison. Among them, 22 samples were used for real-time reverse transcription quantitative PCR (RQ-PCR) detection to revalidate the conversion factor (CF), whereas another 15 samples, along with 6 samples used in RQ-PCR, were used for dPCR comparison. Peking University People’s Hospital determined the validity of its own dPCR and RQ-PCR CFs by detecting World Health Organization international standards. Bland-Altman analysis was conducted to validate the CFs of participating laboratories. If no validated CF was available, the new CF was calculated using the current comparison samples, and its validity was assessed.

Results: 

Among the 14 laboratories participating in the dPCR comparison, 13 detected BCR::ABL1 in all samples. Among the 13 laboratories participating in the RQ-PCR comparison, 5 had validated CFs, whereas the other 8 had validated CFs after recalculating. In the dPCR comparison, 11 laboratories used the same commercial kit, and 10 of them were qualified. Further, 1 laboratory used a laboratory-developed testing reagent and obtained a CF using the standard sample from the National Clinical Laboratory Center, which was also qualified, whereas the 2 laboratories with no CFs both had qualified CFs after calculating. Comparison of results from samples simultaneously performing dPCR and RQ-PCR demonstrated that the BCR::ABL1^IS results derived from their respective qualified CFs showed no significant difference for each laboratory (all P≥0.1). However, among the 6 laboratories with unqualified original CFs in RQ-PCR, 2 showed significant differences between dPCR results and RQ-PCR results derived from the original CFs (P＜0.05), and 3 demonstrated tendencies toward significant differences (P＞0.05 to ＜0.1).

Conclusion: 

dPCR detectionof BCR::ABL1^IS shows high sensitivity and accuracy and has high consistency with RQ-PCR results.