Efficiency in Diagnostic Resolution: Imaging vs. Molecular Tissue of Origin

Transcript:

Tom Beer, MD, FACP: Dr O’Donnell, could you briefly discuss current approaches to diagnostic resolution following a positive MCED [multicancer early detection] result?

Elizabeth O’Donnell, MD: So this is a critical clinical question. Part of the attractiveness of MCED testing is the fact that it offers the potential to help the health care system, both for the patients and for the providers. And so one of the goals of MCED diagnostic tests is that the resolution needs to not add to the burden of the providers who are ordering these tests. The current methods that are being utilized in the 2 different products that we’ve talked about today employ different resolution strategies. One uses a tissue of origin localization, a molecular signal classifier, and the other uses upfront imaging using a head CT scan to image from neck to thigh. Both of these strategies will trigger additional workup, and it’s important to consider the efficiency of each potential diagnostic resolution approach.

Efficiency in this setting is really the total number of diagnostic procedures that those images or biopsies [in order] to achieve diagnostic resolution. Efficiency is influenced by the positive predictive value of the test itself. It’s also influenced by the accuracy of the tissue of origin prediction, in the case of the molecular tissue of origin approach, and also the clinical decision-making by the provider. And so when we think about the different approaches, the molecular tissue of origin approach was used in the PATHFINDER study. This was a study that allowed for the return of results and enabled the assessment of the time to the diagnostic resolution. And what it showed, I think you mentioned this before, Dr Beer, was that most of the tests displayed tissue of origin; molecular signal origin required additional diagnostic imaging. In fact, 92% of the positive tests were followed by imaging of any kind, and 89% were followed by advanced imaging, which is defined as either a CT scan, an MRI, or a PET CT.

The time to diagnostic resolution for all of the participants was 79 days. That was the median. But you can see that there was a very broad range. For the patients who had true positive results, the median was only 57 days, but for the false positive results, the median was about 162 days, which is [approximately] 5 months. The DETECT-A study completed an image-based diagnostic resolution pathway. So after a patient had a positive test, Dr Buchanan explained that there was kind of an MCED tissue tumor board, and then patients would then go on to a head CT scan. This study was not designed to assess the time to diagnostic resolution. What we do know is that the results from the longitudinal analysis of false positives demonstrated that the image-based diagnostic resolution pathway was highly effective and that there was low risk for cancer in patients who had a false positive test but negative imaging and clinical evaluation. I think you’re supposed to ask me what advantages or disadvantages might be associated with that.

Tom Beer, MD, FACP: You guys are doing better than I am keeping up with. Thank you for those comments. Dr O’Donnell, could you comment on what advantages and disadvantages may be associated with suggested approaches?

Elizabeth O’Donnell, MD: That’s a great question. This is still such an early science that we don’t have head-to-head comparisons. A recent study did try to model this, to understand quantitatively, to compare the 2 and examine the diagnostic burden. So this study went through the diagnostic scenarios for MCED test outcomes for the imaging-based approach. It was either a true positive or a false positive. For the real test, it was either true positive with accurate tissue of origin, true positive with inaccurate tissue of origin, or false positive. Then [they would] estimate the number of diagnostic procedures or diagnostic scenarios for that MCED outcome and develop a mathematic expression that incorporated the positive predictive value of the MCED test, the accuracy of the tissue of origin localization, and the number of procedures for each diagnostic outcome. And the 2 approaches were compared by estimating the absolute difference in the diagnostic burden across a range of MCED diagnostic performance levels.

And so the analysis predicted that an imaging diagnostic resolution was more efficient than a molecular tissue of origin strategy. Imaging-based resolution was predicted to have a 28% lower mean overall diagnostic burden as compared with molecular tissue of origin. So an average of 2.6 procedures of imaging or biopsy versus 3.6. And when they scale this or model this at a variety of different positive predictive values and tissue of origin accuracy combinations, there was a lower diagnostic [burden] than 95.5% of the different scenarios. The kind of crossover threshold was a 90% break-even threshold. It was 90% molecular tissue of origin accuracy and a 79% positive predictive value for a tissue of organ strategy to have the same diagnostic burden as an image-based diagnostic resolution strategy. So I think this would favor what we do have, which is only models. We don’t have head-to-head comparisons, which would seem to suggest less of a burden in terms of number of procedures required. But there may be a nuanced approach. As we get better at this, perhaps as the tissue of origin accuracy goes up, we can think of ways to best combine and utilize the 2 approaches.

Transcript is AI generated and edited for readability.