Background: In a randomized multicenter trial of 222 renal tx recipients treated with steroid minimization and a tacrolimus, MMF based regimen, a novel gene expression assay, kSORT was evaluated for its accuracy in diagnosing and predicting biopsy confirmed AR.

Methods: Blood samples were drawn at day 0, 10, months 3, 6, 12 and at graft dysfunction, to perform kSORT assay, a customized 17 gene assay, that provides a high risk (HR) or low risk (LR) immune score for acute rejection (AR). Biopsies (bx) were done on all study patients by protocol at engraftment and 12 months post-transplantation and central histology was read by Banff scores. The kSORT assay was run on 633 blood samples obtained from the first 111 enrolled patients who cp,pelted 1 yr followup. 214 blood samples were matched with protocol or indicated biopsies. 31 patients had clinically suspected AR, of which 18 were BPAR and an additional 6 were borderline (BL-AR). RNA was extracted and QPCR for all 17 genes was normalized to 18S; data was profiled using a customized algorithm kSAS.

Results: Of the 25 biopsy confirmed AR episodes, 21 had definite kSORT scores and 4 were indeterminate; 18/21 AR had high-risk KSORT scores. 11 AR episodes had prior blood samples collected per protocol in the previous 4 months; 8/11 of the pre-AR samples had high kSORT scores. Of the 163 biopsy matched blood samples without histological AR, 139 had definite kSORT scores, and 24 had intedeterminate calls. 132/139 blood samples matched with biopsies without AR, had low-risk kSORT scores. A diagnostic odds ratio was calculated to examine the odds of a (+)kSORT compared to the odds of a (-)kSORT in the confirmed AR group (dOR=39.3, p=4e-15). To evaluate prediction accuracy, there were 107 patients with samples either before on day 0 of the transplant. kSAS called 8/17 with confirmed AR high-risk. Of the 63 stable transplants, 61 were predicted to be low-risk.

Conclusion: Interim results of the diagnostic accuracy of the kSORT assay in a randomized prospective multicenter trial in renal transplantation, confirms that the assay has 85.7% sensitivity, 95.0% specificity, and 97.8% NPV for the non-invasive diagnosis of AR. 73% of AR could have been diagnosed by the kSORT assay days-months prior their current time-line for diagnosis based on the serum creatinine alone, supporting the use of this assay for serial monitoring of rejection risk and proactive immunosuppression customization to alloimmune risk.

CITATION INFORMATION: Lindner P., Shroeder A., Ekberg J., Hsieh S., Towfighi P., Damm I., Sigdel T., Sarwal M. Interim Results of the kSORT in the SAILOR Randomized Multicenter Trial Am J Transplant. 2017;17 (suppl 3).

« Back to 2018 American Transplant Congress