NEW YORK (GenomeWeb) – In a new study published late last week, CombiMatrix has reported its largest dataset to date showing that microarray analysis of more than 8,000 miscarriage samples yielded successful results more than 90 percent of the time, a significantly better performance than standard karyotyping approaches, which are known to fail up to 40 percent of the time.
The results, published late last week in Genetics in Medicine, provide an updated and much more comprehensive picture than previous studies of the efficacy and power of array-based testing to diagnose the cause of pregnancy loss.
In an earlier analysis of only about 4,000 clinically tested samples, Combimatrix reported that it had about an 85 percent success rate in obtaining a result.
Trilochan Sahoo, the company's vice president of clinical affairs and director of cytogenetics, told GenomeWeb this week that while the superior diagnostic power of chromosomal microarray analysis compared to karyotyping is well established in pediatric and prenatal testing spheres, there is still a relative lack of comprehensive evidence for the technology in the specific setting of miscarriage, or products of conception testing.
In amassing greater and more persuasive data for the added value of array-based miscarriage testing, the company stands to aid its own efforts in winning final reimbursement battles, which it recently stepped up.
"We believe the results of this study will further educate the … physician community on the clinical importance of miscarriage analysis testing," Combimatrix President and CEO Mark McDonough said in a statement.
"So far in 2016, 15 insurance providers, including Cigna Health Services and multiple plans in the Blue Cross Blue Shield network, have revised their medical policies to include microarray testing as medically necessary for the evaluation of recurrent pregnancy loss," he added.
The new study — authored by Sahoo, other CombiMatrix colleagues, and collaborators from several medical centers — did not directly compare karyotyping versus array-based testing in matched miscarriage samples. Rather, it looked only at CombiMatrix's overall success rate in obtaining a result in cases that the company accessioned clinically over about four years, as well as the frequencies of different alterations found in these cases.
Overall, amongst 8,118 cases tested, more than 91 percent were successfully processed with reportable results. Among those, 53.7 percent had an abnormal result of some kind, 44.3 percent had a normal result, and 2 percent yielded variants of unknown significance.
In terms of actual diagnostic results, Sahoo and his coauthors concede that very successful karyotype or extensive FISH testing would also be expected to detect the vast majority of the same genomic alterations that CombiMatrix identified in its analysis across these 8,000 samples.
However, the issue is that karyotyping is not always "very successful," failing about 20 to 40 percent of the time due to the necessity of culturing live cells, and the risk of erroneous results due to maternal cell contamination.
Considering those issues, the 90 percent success rate CombiMatrix saw in its 44 months of array-based testing demonstrates a clear improvement over traditional karyotyping.
Smaller studies (and CombiMatrix's own earlier data) have begun to accumulate similar evidence in recent years, but according to Sahoo, the company believes that accumulating and publishing this much larger dataset will provide much stronger evidence to the field of the superiority and clinical utility of an array-based approach.
Over the study period, CombiMatrix's miscarriage testing evolved over three different microarray platforms — first BAC-aCGH, then oligo-aCGH, and most recently a SNP array, Illumina's CytoSNP 850K, which is what the company still uses, and which made up more than 80 percent of the tests performed in the 8,000 cases reported.
Importantly, Sahoo said, the study confirmed what earlier Combimatrix data has demonstrated and what smaller studies have also suggested, which is that array-based miscarriage testing works effectively on preserved formalin-fixed paraffin-embedded tissue samples, not just the fresh samples required by traditional karyotyping.
Overall, the company's various array platforms were successful in determining a result for more than 91 percent of samples. This success was more than 92 percent when looking only at the fresh samples, but remained more than 86 percent even in the approximately 1,800 FFPE samples.
Abnormal results were present in close to 49 percent of the FFPE samples, compared to 54 percent of the overall cohort of successfully tested samples.
As expected, Sahoo and his colleagues wrote, the majority of abnormalities identified were trisomies for either a single or multiple chromosomes, which were present in 67 percent of the cases.
Next were monosomies, which the team observed in 11 percent of the cohort, and further polyploidy in another 12 percent. Another 181 cases, 4.5 percent of the cohort, had segemental imbalances.
About 95 percent of the cases with abnormalities in CombiMatrix's cohort would most likely also have been picked up by a "very successful" karyotype or extensive FISH analysis, Sahoo and his coauthors wrote.
However, a small number of cases with segmental aneuploidies involving segments shorter than 10 Mb — 58 cases among the overall 8,000-case cohort — would probably not have been detectable by standard karyotyping.
Sahoo declined to discuss how CombiMatrix is thinking about future platform or technological shifts in the miscarriage testing space, for example the possibility of incorporating next-gen sequencing in addition to or in lieu of array-based analysis.
ComibMatrix recently adopted NGS for the first time for its preimplantation genetic testing segment, and President and CEO Mark McDonough said at the time that the company was considering other areas of its business where NGS might be at least functionally equivalent while offering throughput advantages or other clinical benefits.