By Diagnostics World News Staff
September 29, 2026 | Since launching AlloSure Plus in July 2025, precision medicine company CareDx has been on a fast-moving expansion of AI-driven diagnostics aimed at one of medicine's highest-stakes patient populations: organ and bone marrow transplant recipients.
The company's flagship tool, AlloSure Plus, earned an Innovative Practices Award earlier this year at the Bio-IT World Conference & Expo — recognition for a kidney transplant rejection risk predictor that CareDx says is already cutting unnecessary biopsies and enabling earlier treatment. Speaking at the event, Chief Data and AI Officer Jing Huang framed the company's pace of development in stark terms: "We move fast because we know those patients need us."
That urgency has produced two additional platforms. AlloHeme tracks relapse risk in blood cancer patients following bone marrow transplant, while VANTx serves as an analytics layer for cohort analysis and clinical trial design across solid organ transplant programs.
The Science Behind the Award
AlloSure Plus pairs familiar clinical markers — eGFR, prior rejection history, proteinuria, graft instability — with donor-derived cell-free DNA, then applies AI to generate a single risk score delivered straight into physicians' Epic EHR workflow. The tool has been validated in more than 2,700 kidney transplant biopsies across pediatric and adult patients, U.S. and European centers, and both first-time and repeat transplant recipients. That breadth, Huang said, makes the results "relevant, applicable, and generalizable."
The numbers back up the award: cell-free DNA alone predicts rejection with an ROC value just over 0.7, but the AI-combined score reached nearly 0.82 — what Huang called a first in transplant rejection prediction. Fresh data presented at the 2026 American Transplant Congress showed an AUROC of 0.79 for distinguishing confirmed rejection from non-rejection.
Beyond Kidneys
AlloHeme, unveiled at the 2026 Tandem Meetings, applies similar next-generation sequencing and AI methodology to bone marrow transplant patients, functioning like a "GPS" that recalibrates relapse risk based on time since transplant. In the ACROBAT study, it posted a 0.89 ROC value and detected relapse signals a median of 41 days before clinical relapse occurred — what Huang described as a "golden window" for physician intervention. Tellingly, most of the tool's false negatives traced back to patients who missed scheduled testing, not a shortcoming in the algorithm itself.
Now used at more than 100 kidney and 30-plus heart transplant centers, VANTx rounds out the ecosystem by surfacing center-level insights that might otherwise go unnoticed. At one transplant center, comparing patient cohorts before and after 2024 showed that expanding treatment to intermediate-risk — not just high-risk — patients drove high-risk populations down to near zero within months, a finding that will be detailed in a forthcoming paper.
Across all three platforms, CareDx is pushing toward the same underlying goal: shifting immunosuppression dosing away from rigid, demographics-based formulas and toward real-time biological response.
For the full report, including CareDx's traffic-light risk model and additional findings from the KOAR registry study, visit Bio-ITWorld.com.