External quality assessment of p210 BCR-ABL1 transcript quantification by RT- qpcr: Findings and recommendations

Transcription

1 Received: 20 February 2018 Revised: 3 July 2018 Accepted: 17 July 2018 DOI: /ijlh ORIGINAL ARTICLE External quality assessment of p210 BCR-ABL1 transcript quantification by RT- qpcr: Findings and recommendations Yu Fu 1,2,3,4 Rui Zhang 1,4 Qisheng Wu 1,2,4 Jiawei Zhang 1,2,4 Lihua Bao 3 Jinming Li 1,2,4 1 National Center for Clinical Laboratories, National Center of Gerontology, Beijing Hospital, Beijing, China 2 Graduate School, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China 3 Department of Nuclear Medicine, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China 4 Beijing Engineering Research Center of Laboratory Medicine, Beijing Hospital, Beijing, China Correspondence Jinming Li, National Center for Clinical Laboratories, Beijing, China. jmli@nccl.org.cn Funding information National Natural Science Foundation of China, Grant/Award Number: ; the Special Fund for Health Scientific Research in the Public Interest from National Population and Family Planning Commission of the People s Republic of China, Grant/ Award Number: ; Beijing Natural Science Foundation, Grant/Award Number: Abstract Introduction: External quality assessment (EQA) is an essential tool for quality assurance of analytical testing processes of p210 BCR-ABL1 transcripts by RT- qpcr. As an EQA provider, the National Center for Clinical Laboratories organized an EQA scheme of p210 BCR-ABL1 testing in China for the first time to identify existing problems and ensure the reliability of p210 BCR-ABL1 testing. Methods: Using armored RNA technology, we first constructed pacyc- MS2- p210 and CG recombinant plasmids and expressed p210 and CG armored RNAs, with packaging segments of p210 BCR-ABL1 fusion gene (FG) and four common control gene (CG) transcripts. Using these armored RNAs, we prepared lyophilized p210 quality control (QC) sample panels and evaluated detection performance of participating laboratories in China. Results: Of the 66 participating laboratories, great variation was found with coefficient of variation (CV%) of raw p210 BCR-ABL1 results basically ranging from 60.0% to 100.0%. In 24 International Scale (IS) laboratories, the CV% of results decreased from 82.4% to 61.6%, and the percentage of laboratories within 2-, 3-, and 5- fold of the median values increased from 78.2%, 87.0%, and 92.1% to 80.1%, 89.4%, and 97.2%, respectively, after conversion with a laboratory- specific conversion factor (CF); however, poorly converted results were also observed in laboratories resulting from changed components of RT- qpcr procedures. False- negative and false- positive results were also found in the EQA. Conclusions: Various problems were found for p210 BCR-ABL1 detection in the EQA. By solving the existing problems, the performance of p210 BCR-ABL1 detection can be improved, ensuring robust laboratory diagnostic capacities in China. KEYWORDS BCR-ABL1, external quality assessment, leukemia, quality control, reverse transcriptionquantitative PCR 1 BACKGROUND The Philadelphia chromosome (Ph) and the resultant BCR-ABL1 fusion gene (FG), resulting from reciprocal translocation between the ABL-1 oncogene of chromosome 9 and a breakpoint cluster region (BCR) on chromosome 22, are common cytogenetic abnormalities in leukemia. 1,2 BCR-ABL1 FG can be divided into a number of subtypes according to the breakpoints in the ABL1 and BCR genes, including e13a2, Fu and Zhang contributed equally to this study John Wiley & Sons Ltd wileyonlinelibrary.com/journal/ijlh Int J Lab Hem. 2019;41:46 54.

2 47 e14a2, e1a2, e19a2, e6a2, e8a2, e13a3, and e14a Among these, e13a2 and e14a2 are the major BCR-ABL1 fusion subtypes, occurring in approximately 95% of chronic myeloid leukemia (CML) patients. 5 Because the BCR-ABL1 FG plays a central role in patients with CML by encoding a deregulated tyrosine kinase (the p210 oncoprotein), it has become a target for CML therapy. 6 An international study showed that 70% of patients treated with imatinib achieved complete cytogenetic remission. 7 In recent years, the application of second- generation tyrosine kinase inhibitors (TKIs), such as nilotinib, dasatinib, and bosutinib, has enabled earlier cytogenetic and deeper molecular responses than imatinib Deep molecular responses in CML provide a rational approach to treatment discontinuation and may predict survival Therefore, quantification of p210 BCR- ABL1 transcript levels may be increasingly important in the future management of leukemia patients. The BCR-ABL1 transcript can be sensitively detected by the widely available and sensitive method of reverse transcriptionquantitative polymerase chain reaction (RT- qpcr). However, due to differences in protocols, RT- qpcr is characterized by significant variation and lack of reproducibility between different laboratories. 8 Although efforts to harmonize various laboratory procedures and reporting methods have been made in order to standardize optimal treatment response criteria and facilitate comparisons across laboratories and patients, wide variability in the detection of p210 BCR-ABL1 transcripts remains, 19,20 hindering clinicians from making appropriate decisions on disease diagnosis and treatment. External quality assessment (EQA) is an effective tool to compare results between laboratories and improve laboratory diagnostic capacities. 21 The United Kingdom National External Quality Assessment Service for Leucocyte Immunophenotyping (UKNEQAS LI) has conducted an international EQA scheme for p210 BCR-ABL1 RT-qPCR analysis for years, covering countries around the world. 19 However, no EQA schemes have been organized in China, and the current performance of p210 BCR-ABL1 testing is unknown. As an EQA provider, the National Center for Clinical Laboratories organized an EQA scheme for p210 BCR-ABL1 testing for the first time, to identify existing problems and ensure the reliability of p210 BCR-ABL1 testing in China. 2 MATERIALS AND METHODS the primers listed in Figure S1. Then, two restriction enzyme sites (FseI/PacI) and two 19- nucleotide sequence- specific MS2 cistrons (pac sites) were added directly to the final targeted sequence of p210 BCR-ABL1 and CG using the primers listed in Table S2. Finally, the targeted segments of the p210 BCR-ABL1 and CG transcripts were cloned into the pacyc- MS2 recombinant plasmid for expression of p210 and CG armored RNAs according to previously described protocol. 22 The p210 and CG armored RNAs were validated by RT- qpcr in different days using a commercial kit (YUAN QI BIO, Shanghai, China) on an ABI 7500 platform (Applied Biosystems, Foster City, CA) and a laboratory- developed test (LDT) using the Europe Against Cancer (EAC) primer/probe sets 23,24 synthesized by Invitrogen Trading (Shanghai, Co., Ltd.) on an ABI 7500 platform (Applied Biosystems, Foster City, CA), respectively. Before the batch production, the p210 and CG armored RNAs must be characteristic of homogeneity between vials and of stability for transportation or storage at different temperatures after they were harvested. Homogeneity and stability tests of the corresponding p210 and CG armored RNAs were conducted following the process presented in Data S1, Section Preparation of the p210 QC sample panel To improve the similarity to clinical samples, we conducted a survey to acknowledge the practical conditions of p210 BCR-ABL1 detection before preparing the QC samples, such as the distribution of p210 BCR-ABL1 and CG transcripts and %BCR-ABL1 ratios (data not shown). Then, variable amounts of p210 armored RNA solutions were diluted with a constant solution of CG armored RNAs to produce a series of p210 QC samples with expected %BCR-ABL1 ratios covering > 10.0% (Level 1(L1)), 1.0%- 10.0% (L2), 0.1%- 1.0% (L3), 0.01%- 0.1% (L4), and <0.01% (L5) intervals according to standardized definitions of molecular responses in CML. 25 Note that we set two pairs of duplicated samples at low and high %BCR- ABL1/CG values to investigate the precision of the detection platform, and one negative sample (phosphate- buffered saline) to investigate laboratory specificity in p210 BCR-ABL1 transcript quantification. Eventually, the p210 BCR- ABL1 QC panels, containing nine positive samples (labeled as 1601, ) and one negative sample (1602) in each panel, were prepared and lyophilized (Data S1, Section 2) for long storage. 2.1 Construction of pacyc- MS2- p210 and CG recombinant plasmids and expression and characteristic evaluation of the corresponding armored RNAs The target segments of p210 BCR-ABL1 (e14a2) and four common CGs (BCR, GUSB, B2M, and ABL1) (Table S1) were amplified from the cdna of a patient expressing the BCR-ABL1 e14a2 transcript. The target segment of p210 BCR-ABL1 was amplified using the forward primer 5 - TCTCTGCACCAAGCTCAAGA- 3 and the reverse primer 5 - CTGCACCAGGTTAGGGTGTT- 3. The four segments of CG were amplified and linked by overlapping PCR at a 1:1 ratio with 2.3 Distribution of p210 QC sample panels to participating laboratories The p210 QC sample panels were shipped to participating laboratories at ambient temperature. The participating laboratories were asked to measure the transcript concentration of QC samples using their routine operating procedures. The following information was requested: date of measurement, resuspension volume, volume of the cdna synthesis, final volume of the cdna reaction, the CG, the standards used and their origin, the RNA extraction method, the assay types, instrument models, the raw BCR-ABL1 and CG transcript values, the raw %BCR-ABL1/CG ratio, and the converted

3 48 %BCR-ABL1/CG ratio. To monitor the effect of transportation on the samples, a blind- mailed sample was sent to a deliberately nonexistent destination and analyzed in our laboratory when it returned using a commercial kit (YUAN BIO, Shanghai, China). 2.4 Data analysis Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, IL, USA) or GraphPad Prism 5.0 (GraphPad Prism Software, San Diego, CA, USA). The Gaussian distribution of p210%bcr-abl1/ CG was tested for normality using the Kolmogorov- Smirnov test. Differences between the experimental groups were subjected to paired t tests and one- way ANOVA. Pearson s correlation analysis was applied to examine the relationship between high and low p210%bcr-abl1/cg ratios. P- values < 0.05 were considered statistically significant. 3 RESULTS 3.1 Verification of pacyc- MS2- p210 and CG recombinant plasmids and corresponding armored RNAs The sequencing results indicated that the targeted segments of p210 BCR-ABL1 and CG were inserted into multiple cloning site 2 of the recombinant plasmid and segments of the four CGs were linked at a 1:1 ratio (Figure 1A). Restriction enzyme digestion indicated the correct sizes for the target segments of p210 BCR-ABL1 (1399 bp) and CG (2151 bp; Figure 1B). After RNA extraction and cdna synthesis, the p210 and CG armored RNAs could be detected by RT- qpcr using both a commercial kit (YUAN BIO, Shanghai, China) and the LDT (Figure 1C). 3.2 Characteristics of p210 and CG armored RNAs Compared with samples stored at 80 C, no statistical differences were found among each of the four p210 and CG armored RNAs upon their storage at 20 C, 4 C, 25 C, or 37 C for 3, 7, 15, and 30 days; no differences were also found among the six p210 and CG armored RNAs upon their storage at 20 C for 1, 2, 3, 4, 5, and 6 months by paired t tests (P > 0.05, Figure 1D). An interunit homogeneity test showed that no outlying results were found among eight randomly chosen samples among 200 vials of the p210 and CG armored RNAs by one- way ANOVA (P > 0.05, Figure 1E). 3.3 Responses from participating laboratories A total of 80 laboratories participated in the trial, among which 71 (88.8%) returned results by the end of the deadline, including 66 clinical laboratories in hospitals, 3 commercial laboratories, and 2 FIGURE 1 Verification and characteristics of p210 BCR-ABL1 FG and CG recombinant plasmids and the corresponding armored RNAs. A, The targeted segments of p210 BCR-ABL1 FG and CG transcripts were successfully detected by sequencing. B, Restriction enzyme digestion indicated that the targeted segments of p210 BCR-ABL1 FG (1399 bp) and CG (2151 bp) had the correct sizes. C, p210 and CG armored RNA could be detected using a commercial kit (left) and LDT (right) targeted to p210 BCR-ABL1 FG and CG (ABL1) segments. D, Stability test of p210 and CG armored RNAs at 37, 25, 4, and E, Homogeneity test of p210 and CG armored RNAs [Colour figure can be viewed at wileyonlinelibrary.com]

4 49 reagent manufacturers. Excluding five participants with < reported CG copies, assay configurations in 66 of the laboratories are listed in Table 1. All the participants used plasmid materials to generate standard curves; 65 (98.5%) participants used ABL1 as a CG, while only one laboratory used GUSB; 55 (83.3%) participants used TRIzol reagent to extract RNA, and 11 (16.7%) of laboratories used a Spin column method. Note that 54 (81.8%) participants employed a common commercial kit from YUAN QI BIO (Shanghai, China). The kit used ABL1 as CG and gene- specific primer for reverse transcription and 1- step RT- qpcr method (cdna synthesis and qpcr amplification in the same run) for %BCR-ABL1 detection. In addition, 24 (36.4%) participants could convert their results to the International Scale (IS) using laboratory- specific conversion factors (CFs). 3.4 Performance of participating laboratories The CV% was calculated with the original detection results of samples. Overall, the difference between the maximum and the minimum CV% was large, ranging from 60.0% to 100.0%, and the percentage of laboratories (%Lab) within 2-, 3-, and 5- fold of the median QC value were 45.5%- 80.3%, 68.2%- 92.4%, and 72.7%- 95.5%, respectively. We also found that the CV% tended to be larger and %Lab within 2-, 3-, and 5- fold of QC tended to be lower when the %BCR-ABL1/CG ratio was<1.0% (Table 2). For the two pairs of duplicated samples (low level: 1601 & 1606; high level: 1608 & 1609), no differences were observed between results from the participating laboratories (P > 0.05), and %BCR-ABL1/ CG value of p210h (the mean value of sample 1608 & 1609) was strongly related to that of p210l (the mean value of sample 1601 & 1606) (r 2 = , P < 0.001; Figure S2A), indicating that the measurement systems between laboratories were relatively stable. The distribution of %BCR-ABL1 values in each QC samples was skewed using Kolmogorov- Smirnov test (P < 0.05). Moreover, the detection results from blind- mailed samples were located at the main distribution region of %BCR-ABL1 values in each QC samples in the trial (Figure S2B), indicating that after freeze- drying, the p210 QC sample panels were stable under conditions of transport for normal use in participated laboratories. In addition, 2 (3.0%) laboratories returned false- positive results and 1 (1.5%) laboratory reported CG > copies in the negative sample (1602), suggesting cross- contamination in their daily work. For positive samples with %BCR-ABL1/CG ratios < 1.0%, 15 falsenegative results were reported from 11(16.7%) laboratories. 3.5 Performance of IS laboratories Among the 24 laboratories that could convert their results to IS using CFs, the CV% of the unconverted results was 59.7% % with an average value of 82.4%; while the CV% of the converted results was 48.4%- 80.6% with an average value of 61.6% (Table 2). The unconverted %Lab within 2-, 3-,and 5- fold were 78.2% (50.0%- 87.5%), 87.0% (70.8%- 91.7%), and 92.1% (75.0% %), TABLE 1 Assay configurations of the 66 p210 BCR-ABL1 tests in the EQA scheme Parameter Control gene Number of labs ABL BCR GUSB B2M Other Assay type Commercial kit (YUAN QI BIO) Commercial kit (other) LDTs Reporting on International Scale Yes No IS calibration method Sample exchange Not on IS RNA Extraction method TRIzol Spin columns step vs 2-step RT-qPCR 1-step step Reverse transcription primers Random primer Gene- specific primer Instrument and model ABI ABI 7500Fast ABI ABI StepOnePlus ABI QSDx Roche LightCycler480 II Roche Cobas z Bio-Rad CFX Agilent StrataGene Mx3000P Agilent AriaMx G8830A Qiagen Rotor- Gene In- run standard Plasmid (YUAN QI BIO) Plasmid (other) Plasmid (LDTs) Percentage LDTs, laboratory- developed tests; IS, International Scale; RT- qpcr, reverse transcription- quantitative PCR.

5 50 TABLE 2 p210 BCR-ABL1 detection results of the 66 participants in the EQA scheme L1 L2 L3 L4 L a 1609 a a 1606 a All participant laboratories (n = 66) Mean SD CV% Median Min Max Range %Lab within 2- fold c %Lab within 3- fold c %Lab within 5- fold c IS participant laboratories (n = 24) Mean SD CV% Median Min Max Range %Lab within 2- fold c %Lab within 3- fold c %Lab within 5- fold c IS participant laboratories b (n = 24) Mean SD CV% Median Min Max Range %Lab within 2- fold c %Lab within 3- fold c %Lab within 5- fold c a Duplicated samples (1601&1606; 1608&1609); b Converted results from 24 IS participant laboratories; c Rate of laboratories with BCR-ABL1 values locating in the ±2- fold, 3- fold, or 5- fold of the median values of samples. respectively; by contrast, the converted %Lab within 2-, 3-,and 5- fold were 80.1% (62.5%- 91.7%), 89.4% (70.8%- 95.8%), and 97.2% (83.3% %), respectively (Table 2). Importantly, as displayed in Figure 2, we found that the results from majority of laboratories, such as Lab #41 and Lab #55 (marked in yellow), were well converted by CFs, while results from only a few laboratories, such as Lab #43 and Lab #44 (marked in red), were poorly converted, with results deviating from the median value compared to the unconverted result. 3.6 Follow- up of CF changes in IS laboratories To survey whether changes in the detection system affected the conversion in participating IS laboratories, particularly for Labs #43 and #44, we performed a follow- up on the condition of CF revalidation after the EQA scheme. As shown in Table 3, 16 laboratories did not need to revalidate their using CFs because their CFs were within the effective period. For the 8 laboratories that revalidated their results, 5 passed the revalidation, while 3 did not pass, and new

6 51 FIGURE 2 Performance evaluation of the 24 IS laboratories. Diagram of unconverted and converted BCR-ABL1 values with specific CFs. Yellow points: two representative laboratories (Lab #41 and Lab #55) obtained good conversion effects with CFs; red points: two representative laboratories (Lab #43 and Lab #44) obtained poor conversion effects [Colour figure can be viewed at wileyonlinelibrary.com] CFs were calculated in accordance with the procedure of reference laboratory in China. Labs #43 and #44 were included in these 3 laboratories with changed components of their procedures, and the newly recalculated CFs produced well- converted results. In addition, although no changes occurred in the RT- qpcr platform, Lab #18 did not pass the revalidation, and a new CF was calculated. When using the newly recalculated CFs, the CV% of the converted results can further decrease from 61.6% to 50.6%. 4 DISCUSSION Armored RNA technology is a mature technology that has been used to generate versatile calibrators and control materials for multiple viral assays. 22,26-28 In recent years, armored RNA has also been assessed for use as a candidate reference material for BCR-ABL1 FGs and was successfully manufactured into a secondary reference panel calibrated with the first World Health Organization (WHO) international genetic reference panel. 29,30 In this study, we successfully developed an armored RNA- based p210 QC sample panel and organized the first ever EQA scheme for p210 BCR-ABL1 detection in China. Compared to cell- based samples, an excellent characteristic of armored RNA- based control materials is their ease of mass production. In this study, to further improve the stability of the armored RNA, we conducted a lyophilization process after the sample panel was prepared, and similar to previous studies, 22,27,28 this study has also demonstrated that armored RNA is stable and homogeneous. Moreover, considering that different CGs might be used in different laboratories, we selected four common used CGs (ABL1, BCR, GUSB, and B2M) and linked them together by overlapping PCR at a 1:1 ratio. The returned results demonstrated that all participants could successfully detect segments of the p210 BCR-ABL1 and CG transcripts, regardless of the RNA extraction methods and assay types employed, indicating that the armored RNA based- simulated samples had good practicability as QC materials in the detection of p210 BCR-ABL1 transcripts. Harmonization of the methodology for BCR-ABL1 transcript quantification contributes to decreased variability between laboratories. 17 In this EQA scheme for BCR-ABL1 testing, although 81.8% and 98.5% of the laboratories used a commercial kit (YUANQI BIO, Shanghai, China) and a common CG (ABL1), respectively, the CV% of p210 BCR-ABL1 detection results ranged from 60.0% to 100.0%, slightly exceeding a previous result reported

7 52 Lab # Old CF New CF Process of validation Changes in RT- qpcr procedure TABLE 3 Follow- up of CF changes in the 24 IS laboratories No validation No change No validation No change No validation No change Pass validation No change No validation No change No validation No change No validation No change Recalculation No change Pass validation No change No validation No change No validation No change Pass validation No change No validation No change No validation No change No validation No change Pass validation No change No validation No change Recalculation Laboratory equipment was maintained Recalculation Enzyme mixture and primer changed Pass validation No change No validation No change No validation No change No validation No change No validation No change CF, conversion factor; IS, International Scale; RT- qpcr, reverse transcription- quantitative PCR. by Branford et al 16 and the ranges between the maximum and minimum values of positive samples were very large, indicating that large differences existed between laboratories. We also found that the lower the BCR-ABL1 values, the larger the variability became in the EQA, which is in accordance with previous studies. 19,30 Because deeper molecular responses after treatment can reflect the treatment effect in patients with leukemia, 31,32 a large variability in the detection of low BCR-ABL1 values can hinder clinicians from making appropriate decisions with regard to the diagnosis and treatment of the disease. According to Branford et al 16 performing replicate analysis (during RT, qpcr, or both) and averaging the results can optimize measurement reliability. Thus, routine duplicate analysis of samples might improve the precision of detection results. No differences were found in the duplicated samples at low or high %BCR-ABL1values, and the high %BCR-ABL1/CG values were strongly related to the low values, indicating that the measurement systems between laboratories were relatively stable. However, inherent internal variability should not be neglected in routine work. According to a survey conducted by Branford et al 33 the variability at a high BCR-ABL1 level was 18.0%, while the low BCR-ABL1 level displayed greater variability, with a CV of 32.0%. In a 6- year analysis, Ruiz et al 34 also showed that the CV% of low BCR-ABL1 levels was > 30%, while the CV% of high BCR-ABL1 levels was <30%. Inherent internal variability can be minimized by using internal QC samples within each run and by defining criteria for the acceptance or rejection of a run. 16 It is very important for clinicians to appropriately assess treatment response of leukemia patients by using %BCR-ABL1 values reported on IS. 8 For the 24 IS laboratories, the CV% of the converted result was lower and the %Lab within 2-, 3-, and 5- fold were higher compared to those of the unconverted results, indicating that CFs can harmonize p210 BCR-ABL1 transcript quantification between laboratories, consistent with previous findings. 19,34,35 The improved comparability of p210 BCR-ABL1 transcript values between laboratories is mainly attributed to effective work toward the standardization of p210 BCR-ABL1 transcript quantification, such as a defined standardized baseline, 7,36 the proposal of IS, 17 establishment of CFs, 37 and development of primary and secondary reference reagents. 29,38,39 However, because of limitations in establishing

8 53 specific CFs by exchanging samples with a reference laboratory, IS laboratories accounted for <30% of all the laboratories in China. To facilitate regional standardization of p210 BCR-ABL1 monitoring, we have developed an armored RNA- based secondary reference material in accordance with the WHO primary standards, and assay stability between laboratories provides a prerequisite for standardization of p210 BCR-ABL1 quantification in China. Importantly, we found two laboratories with poor conversion effects in the EQA scheme and we suspected that they were using CFs that were no longer suitable for IS conversion. Follow- up results showed that certain components had changed in these laboratories. According to Branford et al 37 minor alterations to analytical systems may have significant impacts on the measurements, requiring CF recalculation. However, in the EQA scheme, we also found a laboratory with no component changes in the detection system that did not pass validation. Indeed, imperceptible changes are always occurring in laboratories, such as instrument aging, reagent lot changes, operator differences, which might affect the stability of measurement. Therefore, the use of CF in IS laboratories should be revalidated frequently, particularly for laboratories that have changed their RT- qpcr components, which can further improve the comparability between laboratories. However, there has been no consensus on the frequency of CF revalidation. According to the recommendation of Müller et al 35 repetition of patient sample evaluation to confirm the CF should be performed every 2 years, or when parts of the methods need to be changed. In our daily work, we can monitor shifts in data by including internal QC samples in every run to determine whether the CF needs to be validated or not, according to the recommendations of Branford et al. 37 Moreover, false- positive and false- negative results were also reported in the EQA scheme. False- positive results can be avoided by reducing cross- contamination following the strict measures suggested by Hughes et al. 17 False negatives can be improved by increasing the sensitivity of the RT- qpcr platform. According to the recommendations of Alikian et al 40 adding more cdna to the reaction or increasing the reaction volume could improve sensitivity using RT- digital PCR (an adaptation of RT- qpcr).therefore, increasing the amount of template in the reaction system, particularly for laboratories using LDTs, decreases false- negative results in BCR- ABL1 testing by RT- qpcr. Avoiding false- positive and false- negative results would help increase comparability between laboratories. 5 CONCLUSION In the EQA scheme for p210 BCR-ABL1 detection in China, an unacceptable level of variability of detection results was found from the participating laboratories, even from the IS laboratories. So we call for an urgent reevaluation of the used methods and the wider standardization for p210 BCR- ABL1 detection via the secondary reference material, which has been developed in China. We believe that this EQA scheme can improve the reproducibility and accuracy of p210 BCR-ABL1 detection to ensure robust laboratory diagnostic capacities in China. ACKNOWLEDGEMENTS We thank all the laboratories that participated in this program. We also thank the commercial kit manufacturer (YUAN QI BIO) for providing the detection reagent in the study. This work was supported by grants from National Natural Science Foundation of China (Grant ), Beijing Natural Science Foundation (Grant ), and the Special Fund for Health Scientific Research in the Public Interest from National Population and Family Planning Commission of the People s Republic of China (No ). COMPETING INTERESTS The authors declare that they have no competing interests. AUTHOR CONTRIBUTIONS R. Zhang and J. Lin designed the research study. Y. Fu performed the research, analyzed the data, and wrote the manuscript. Q. Wu, J. Zhang, and L. Bao reviewed manuscript and done critical revision. All authors read and approved the final manuscript. ETHICS APPROVAL AND CONSENT TO PARTICIPATE The study design was approved by the Ethics Committee of National Center for Clinical Laboratories, Beijing Hospital (reference number: NCCLLLKYPJ ). ORCID Yu Fu REFERENCES 1. Rowley JD. Identification of a translocation with quinacrine fluorescence in a patient with acute leukemia. Ann Genet. 1973;16: Shtivelman E, Lifshitz B, Gale RP, Canaani E. Fused transcript of abl and bcr genes in chronic myelogenous leukaemia. Nature. 1985;315: Hochhaus A, Reiter A, Skladny H, et al. A novel BCR- ABL fusion gene (e6a2) in a patient with Philadelphia chromosome- negative chronic myelogenous leukemia. Blood. 1996;88: Melo JV. The diversity of BCR- ABL fusion proteins and their relationship to leukemia phenotype. Blood. 1996;88: Vinhas R, Cordeiro M, Pedrosa P, Fernandes AR, Baptista PV. Current trends in molecular diagnostics of chronic myeloid leukemia. Leuk Lymphoma. 2016;58: Quintás-Cardama A, Cortes J. Molecular biology of bcr- abl1- positive chronic myeloid leukemia. Blood. 2009;113: O Brien SG, Guilhot F, Larson RA, et al. Imatinib compared with interferon and low- dose cytarabine for newly diagnosed chronic- phase chronic myeloid leukemia. N Engl J Med. 2003;348: Baccarani M, Deininger MW, Rosti G, et al. European LeukemiaNet recommendations for the management of chronic myeloid leukemia: Blood. 2013;122:

9 54 9. Jabbour E, Kantarjian HM, Saglio G, et al. Early response with dasatinib or imatinib in chronic myeloid leukemia: 3- year follow- up from a randomized phase 3 trial (DASISION). Blood. 2014;123: Saglio G, Kim DW, Issaragrisil S, et al. Nilotinib versus imatinib for newly diagnosed chronic myeloid leukemia. N Engl J Med. 2010;362: Khoury HJ, Cortes JE, Kantarjian HM, et al. Bosutinib is active in chronic phase chronic myeloid leukemia after imatinib and dasatinib and/or nilotinib therapy failure. Blood. 2012;119: Hochhaus A, Rosti G, Cross NC, et al. Frontline nilotinib in patients with chronic myeloid leukemia in chronic phase: results from the European ENEST1st study. Leukemia. 2016;30: Mahon FX, Réa D, Guilhot J, et al. Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre Stop Imatinib (STIM) trial. Lancet Oncol. 2010;11: Ross DM, Branford S, Seymour JF, et al. Safety and efficacy of imatinib cessation for CML patients with stable undetectable minimal residual disease: results from the TWISTER study. Blood. 2013;122: Gambacorti-Passerini C, Antolini L, Mahon FX, et al. Multicenter independent assessment of outcomes in chronic myeloid leukemia patients treated with imatinib. J Natl Cancer Inst. 2011;103: Branford S, Cross NCP, Hochhaus A, et al. Rationale for the recommendations for harmonizing current methodology for detecting BCR- ABL transcripts in patients with chronic myeloid leukaemia. Leukemia. 2006;20: Hughes T, Deininger M, Hochhaus A, et al. Monitoring CML patients responding to treatment with tyrosine kinase inhibitors: review and recommendations for harmonizing current methodology for detecting BCR- ABL transcripts and kinase domain mutations and for expressing results. Blood. 2006;108: Foroni L, Wilson G, Gerrard G, et al. Guidelines for the measurement of BCR- ABL1 transcripts in chronic myeloid leukaemia. Br J Haematol. 2011;153: Scott S, Travis D, Whitby L, Bainbridge J, Cross NCP, Barnett D. Measurement of BCR- ABL1 by RT- qpcr in chronic myeloid leukaemia: findings from an International EQA Programme. Br J Haematol. 2017;177: White H, Deprez L, Corbisier P, et al. A certified plasmid reference material for the standardisation of BCR- ABL1 mrna quantification by real- time quantitative PCR. Leukemia. 2015;29: Miller WG, Jones GR, Horowitz GL, Weykamp C. Proficiency testing/external quality assessment: current challenges and future directions. Clin Chem. 2011;57: Zhan S, Li J, Xu R, Wang L, Zhang K, Zhang R. Armored long RNA controls or standards for branched DNA assay for detection of human immunodeficiency virus type 1. J Clin Microbiol. 2009;47: Gabert J, Beillard E, van der Velden VH, et al. Standardization and quality control studies of real- time quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia - a Europe Against Cancer program. Leukemia. 2003;17: Beillard E, Pallisgaard N, van der Velden VH, et al. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using real- time quantitative reversetranscriptase polymerase chain reaction (RQ- PCR) - a Europe against cancer program. Leukemia. 2003;17: Cross NCP, White HE, Müller MC, Saglio G, Hochhaus A. Standardized definitions of molecular response in chronic myeloid leukemia. Leukemia. 2012;26: Fu Y, Li J. A novel delivery platform based on Bacteriophage MS2 virus- like particles. Virus Res. 2016;211: Zhang L, Hao M, Zhang K, et al. External quality assessment for the molecular detection of MERS- CoV in China. J Clin Virol Feb;75: Wang G, Sun Y, Zhang K, et al. External quality assessment of molecular detection of ebola virus in China. PLoS ONE. 2015;10:e White HE, Hedges J, Bendit I, et al. Establishment and validation of analytical reference panels for the standardization of quantitative BCR- ABL1 measurements on the international scale. Clin Chem. 2013;59: White H, Cross NC. Armored RNA as reference material for standardisation of BCR-ABL RQ-PCR methods: report of field trial evaluation Available at: Mahon FX, Etienne G. Deep molecular response in chronic myeloid leukemia: the new goal of therapy? Clin Cancer Res. 2014;20: Hehlmann R, Müller MC, Lauseker M, et al. Deep molecular response is reached by the majority of patients treated with imatinib, predicts survival, and is achieved more quickly by optimized high- dose imatinib: results from the randomized CML- study IV. J Clin Oncol. 2014;32: Branford S. Molecular monitoring in chronic myeloid leukemiahow low can you go. Hematology Am Soc Hematol Educ Program. 2016;2016: Ruiz MS, Medina M, Tapia I, et al. Standardization of molecular monitoring for chronic myeloid leukemia in Latin America using locally produced secondary cellular calibrators. Leukemia. 2016;30: Müller MC, Cross NC, Erben P, et al. Harmonization of molecular monitoring of CML therapy in Europe. Leukemia. 2009;23: Hughes TP, Kaeda J, Branford S, et al. Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia. N Engl J Med. 2003;349: Branford S, Fletcher L, Cross NCP, et al. Desirable performance characteristics for BCR- ABL measurement on an international reporting scale to allow consistent interpretation of individual patient response and comparison of response rates between clinical trials. Blood. 2008;112: White HE, Matejtschuk P, Rigsby P, et al. Establishment of the first World Health Organization International Genetic Reference Panel for quantitation of BCR- ABL mrna. Blood. 2010;116:e111 e Cross NCP, White HE, Ernst T, et al. Development and evaluation of a secondary reference panel for BCR- ABL1 quantification on the International Scale. Leukemia. 2016;30: Alikian M, Whale AS, Akiki S, et al. RT- qpcr and RT- Digital PCR: a comparison of different platforms for the evaluation of residual disease in chronic myeloid leukemia. Clin Chem. 2017;63: SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of the article. How to cite this article: Fu Y, Zhang R, Wu Q, Zhang J, Bao L, Li J. External quality assessment of p210 BCR-ABL1 transcript quantification by RT- qpcr: Findings and recommendations. Int J Lab Hem. 2019;41: ijlh.12919