University of Michigan Dexter, Michigan, United States
Background/Case Studies: CD34 is a critical analyte in stem cell processing and is subject to proficiency testing (PT). Limited commercial CD34-PT are available from domestic (PT-D) and international (PT-InL) vendors. We describe our experience with two CD34-PT vendors over 11 years. To our knowledge, an institutional comparison of different commercial CD34-PT has not been previously reported.
Study
Design/Methods: CD34-PT were purchased from a domestic (PT-D [USA], CAP) and international vendor (PT-InL [Europe], UK-NEQA). CD34-PT were analyzed by flow cytometry using a dual platform, modified ISHAGE protocol on day of receipt. Elements for comparison included %CD34, CD34/mcL, total CD34 events, and whether PT challenge was graded. Our results were compared to the study mean by paired t-test, linear regression, Z-score and chi-square. For PT-D, results were also compared to the certificate of analysis (CoA). PT vendors were also compared relative to standard deviation (SD), % coefficient variation (%CV) and ISHAGE CD34 minimum count success rate (>200 CD34+ per sample).
Results/Findings: We participated in 64 CD34-PT over 11 years. We initially participated in PT-D which provides stabilized cell samples. In 2019, we switched to PT-InL which supplies fixed, nonviable cell samples. Linear regression showed a good correlation (R > 0.95) between our results and mean PT results for %CD34 and CD34/mcL regardless of vendor. The %CV for PT-D and PT-InL was similar for %CD34 (19% vs 16%) whereas %CV for CD34/mcL was 3x higher for PT-D (37% vs 13%, P=0.002). Our %CD34 average Z-score, a measure of our result variance from the mean, showed most %CD34 results falling within 1SD (>75%) and 2SD (>90%) for both vendors (Table 1). In contrast, CD34/mcL varied significantly: The PT-D mean Z-score was 2.5x higher, with only 17% results falling within 1SD (vs 77% PT-InL, P=0.0001) and 46% > 2SD (P=0.0013, Z=3.7). In the same PT-D challenges, %CD3 and CD3/mL showed a similar pattern, with %CD3 within 1SD and CD3/mL > 2-3 SD suggesting sample degradation and cell loss. For PT-D, the mean CD34/mcL was 22% lower (P=0.16, paired-t; range 2-53%) and our result 48% lower (range 32%-92%; p=0.003) than the CD34/mcL listed on CoA. PT-InL analysis reached the minimum ISHAGE CD34 count (> 200 events) in 92% challenges vs only 71% in PT-D (P=0.02). The majority (97%) of PT-InL challenges were graded with > 350 participants. PT-D participation was significantly lower ( < 75 centers) and were ungraded until 2016. Conclusions: Based on our experience, CD34-PT using fixed cells (PT-InL) was superior to stabilized cell samples (PT-D), which showed evidence of degradation resulting in low CD34/mcL and total CD34 counts ( < 200). Issues encountered with PT-InL are occasional shipment delays due to US customs clearance.
Importance of research: CD34 proficiency testing is required by regulatory bodies overseeing cell processing laboratories although there are very limited options on the market. The 'real-life' performance of commercial CD34 PT vendors based on PT sample preparation is of immediate interest to cell therapy professionals.
