Abstract
Background: We previously reported that graft failure due to nonadherence (GFNA) was a major cause of graft loss in kidney transplantation. Here, among 150 prospectively-followed kidney transplant recipients at 18 years post-transplant, we provide: updated (longer-term) estimates of cause-specific graft loss probabilities, risk factors for developing GFNA, and detailed characterizations of patients' overt nonadherent (NA) behavior, including timing, extent, and clinical consequences. Methods: Determination of the patient becoming NA in taking his/her immunosuppressive medications, and the underlying cause of graft loss, were determined prospectively by the attending physicians. For never-functioning-graft, GFNA, GF due to causes other than NA (Other GF), and death with a functioning graft (DWFG), cumulative incidence functions were used to estimate the cumulative probabilities of cause-specific graft loss. Cox stepwise regression was used to determine significant multivariable predictors for the hazard rate of developing GFNA. Results: GFNA was a major cause of graft loss (22/150 patients), particularly among African-American and Hispanic recipients <50 years of age-at-transplant (20/56 experienced GFNA), with estimated percentages of such patients ever developing GFNA ranging between 36.9 and 41.5%. These patients were also at a higher risk of developing Other GF. For the remaining patients (2/94 experienced GFNA), estimated percentages of ever-developing GFNA were much lower (range: 0.0-6.7%). The major cause of graft loss among recipients >= 50 years of age was DWFG; GFNA rarely occurred among older recipients. In 21/22 GFNA patients, NA behavior lasted continuously from the time of developing NA until GFNA. In total, 28/150 patients became NA, and 67.9% (19/28) occurred beyond 36 months post-transplant. A total of 25 of 28 NA patients (89.3%) developed biopsy-proven acute rejection and/or chronic rejection that was directly attributed to the NA behavior. Lastly, 25/28 admitted to NA behavior, with financial and psychological components documented in 71.4% (20/28) and 96.4% (27/28) of NA cases, respectively. Conclusions: These results highlight the importance of performing serial monitoring of patients for overt NA behavior throughout their post-transplant follow-up. Financial and psychological components to NA behavior need to be simultaneously addressed with the goal of achieving complete avoidance/elimination of NA behavior among higher risk patients.
