Work

Mitigation of errors on an FDA-approved platform for cytomegalovirus viral load assay

Year: 2024

Type: article

Abstract: KEYWORDS automated platforms, CMV viral load monitoring, device recall O n 14 September 2023, Hologic Inc. issued an urgent medical device recall notifica tion for the Aptima CMV Quant Assay (Aptima) for ML2 errors on the Panther Instrument.Approved by the FDA, the performance characteristics of the assay had been compared in a large study previously (1).We verified the assay in our laboratory using COBAS CMV assay as the comparator and started clinically using this assay in February of 2023.Within a month of adopting the assay, we encountered "ML2 error flags" that are due to clogging of the magnetic wash station aspirators after initial aspiration of the sample according to the recall notification.An ML2 error refers to a liquid level sensing error where the Magnetic Wash Station reports a higher than acceptable liquid level after wash fluid is dispensed and it leads to an invalid specimen result.The vendor also indicated in this notification document that the errors have been reported with a frequency of 0.54% and an investigation conducted by Hologic indicated that an abnormally high concentration of globulin proteins in a sample may result in clogging of the analyzer during assay processing.In our experience, the errors occurred randomly and once the error occurred, the sample causing the error and all samples that are queued after the sample that caused the error were flagged as "invalid" by the instrument.These invalid results were random, and affected results for CMV associated with the Aptima assay; and the EBV viral load assay that was also being performed as a Fusion Open Access Taqman assay on the same instrument.Since samples could not be re-tested due to lack of residual volume, these errors led to fresh blood draws from those patients whose specimens were affected.We performed an investigation into the ML2 errors and identified nine samples from seven patients on the nine different occasions and over a period of 63 days triggered the error and resulted in invalid results for 95 samples for quantitative CMV viral load assay.We performed a total of 1,039 CMV assays during this time; thus, the frequency of ML2 errors in our experience was 0.87%.In addition, invalidation of additional samples being tested for CMV accounted for a much greater failure rate of about 10% during this time.Based on the investigation from Hologic indicating high globulin levels, we checked the immunoglobulin and total protein levels of the plasma samples that were associated with the ML2 errors.We also randomly selected 27 samples from an additional 26 patients that had documented results for CMV/EBV quantitation or had a non-ML2 error flag (these included a failure of control or a volume pipetting error) on the instrument.Only two samples of these 27 randomly selected samples that triggered errors other than ML2 had high immunoglobulin (IgG) levels of 2,098 and 1,910 mg/dL (normal range: 556-1,631 mg/dL); in contrast, six of the nine patients with samples causing the ML2 error had either elevated immunoglobulin IgG or IgM (n = 5; range: 1,772-6,453 mg/dL) level or a high γ-globulin in serum electrophoresis test (n = 1; level 2.9; normal range: 0.8-1.8g/dL) or high total protein (n = 1; 10.2; normal range: 6.4-8.3 g/dL) levels

Source: Journal of Clinical Microbiology

DOI: 10.1128/jcm.00416-24