Clinical validation of an artificial neural network trained to identify acute allograft rejection in liver transplant recipients.

Artificial neural networks (ANNs) are techniques of nonlinear data modeling that have been studied in a wide variety of medical applications. An ANN was developed to assist in the diagnosis of acute rejection in liver transplant recipients. We investigated the diagnostic accuracy of this ANN on a new data set of patients from the same hospital. In addition, we compared the diagnostic accuracy of the ANN with that of the individual input variables (alanine aminotransferase [ALT] and bilirubin levels and day posttransplantation). Clinical and biochemical data were collected retrospectively for 124 consecutive liver transplantations (117 patients) over the first 3 months after transplantation. Diagnostic accuracy was calculated using receiver operating characteristic (ROC) curve analysis. The ANN differentiated rejection from rejection-free episodes in the new data set over the first 3 months posttransplantation with an area under the ROC curve of 0.902 and sensitivity and specificity of 80.0% and 90.1% at the optimum decision threshold, respectively. The ANN was significantly more specific than ALT or bilirubin level or day posttransplantation at their corresponding optimum decision thresholds (P <.0001). Peak ANN output occurred 1 day earlier than peak values for either ALT or bilirubin (P <.005). The diagnostic accuracy of the ANN was greater than that of any of the individual variables that had been used as inputs. It would be a useful adjunct to conventional liver function tests for monitoring liver transplant recipients in the early postoperative period.

Neuro-computing versus linear statistical techniques applied to liver transplant monitoring: a comparative study.

This paper explores the potential for the application of neurocomputing in on-line monitoring in the liver transplantation domain. It extends our previously documented work to provide both an assessment of the performance gains achievable by incorporating temporal and dynamical information about the measurements made on a patient as well as presenting a novel computerized clinical decision aid for this domain. A comparison of the performance of linear and nonlinear classification system is made and used to motivate the final selection of the diagnostic inputs.

Blood levels of TGFbeta1 in liver transplant recipients receiving either tacrolimus or micro-emulsified cyclosporine.

BACKGROUND: Transforming growth factor beta-1 (TGFbeta1) is pro-fibrotic in addition to being a potent immunosuppressive cytokine. Cyclosporine (cyclosporin A[CsA]) has been found to increase circulating TGFbeta1 levels in patients (1, 2). To determine whether tacrolimus (FK506) similarly increases TGFbeta1 we have measured TGFbeta levels in blood samples from liver graft recipients who were of known TGFbeta1-responder status. METHODS: Sequential serum and plasma samples were obtained from liver transplant recipients in the UK trial of tacrolimus versus microemulsified CsA, with a follow up period of between 50 and 265 days. Twelve patients received CsA and 13 received tacrolimus. Active and total TGFbeta1 protein were measured and plasma beta thromboglobulin (betaTG) levels were used as an indirect indication of platelet-derived TGFbeta contamination of samples. RESULTS: We found no correlation between trough drug levels and active TGFbeta1 levels in serum of either set of patients. Plasma beta thromboglobulin was detected in platelet-depleted plasma samples, indicative of platelet damage before plasma separation. CONCLUSION: Neither CsA nor tacrolimus induced active TGFbeta1 blood levels in liver transplant recipients during a follow up period of < or = 265 days.

Monitoring eosinophil activation and liver function after liver transplantation.

BACKGROUND: Portal tract eosinophil infiltration and an increase in the blood eosinophil count (EOS) have been shown to be specific markers of liver allograft rejection. The graft eosinophil infiltration is associated with the local release of eosinophil cationic protein. Therefore, serum eosinophil cationic protein concentration (ECP) is a potential marker for acute allograft rejection. We investigated the chronological relationship among and diagnostic value of serial changes in EOS, ECP, and liver function tests (LFTs) following liver transplantation. METHODS: EOS, ECP, serum alpha-glutathione S-transferase concentration, and conventional LFTs were measured in serial samples collected over the first 3 postoperative months following 58 liver transplants. The diagnostic potential of each test, alone or in combination, was reviewed over the entire follow-up period. RESULTS: EOS and ECP increased at a median period of 3.5 and 4 days, respectively, before the diagnosis of acute rejection, and this increase was significantly earlier than the corresponding changes in LFTs (P<0.05). There was a significant correlation between the day of the first increase in EOS and alpha-glutathione S-transferase (rs=0.535; P=0.009) and EOS and alanine transaminase (rs=0.629; P=0.004). The optimum combination of tests for the diagnosis of acute rejection was an increase in both EOS and GST with a predictive efficiency of 84%. CONCLUSIONS: Increases in EOS and ECP are early indicators of acute liver allograft rejection and precede evidence of hepatocellular damage. However, an increase in ECP was also frequently associated with infection. Therefore, we recommend the regular monitoring of EOS in conjunction with routine LFTs after liver transplantation as an aid to the early diagnosis of acute rejection.

Randomized trial to evaluate the clinical benefits of serum alpha-glutathione S-transferase concentration monitoring after liver transplantation.

BACKGROUND: An increase in serum alpha-glutathione S-transferase concentration (GST) has been shown to be a more sensitive and specific marker of hepatocellular damage than equivalent increases in transaminase activities. A randomized clinical trial of 60 liver transplants in 49 patients was carried out to assess the clinical benefits of GST monitoring as a supplementary test to routine liver function tests during the first 3 postoperative months after liver transplantation. METHODS: Mortality and morbidity were compared in graft recipients who had their GST reported daily to the ward (reporting group) and graft recipients who did not. RESULTS: The 3-month survival rate was significantly greater in the reporting group (P=0.033) and the risk of graft loss was halved (relative hazard ratio=0.50; P=0.29). The reporting group also had significantly more patients who spent less than 3 weeks in the hospital throughout the follow-up period (P=0.036). In addition, the reporting group experienced a lower frequency of biopsies per graft (P=0.038), less severe rejection (P=0.015), and a lower incidence of infection episodes per graft (P=0.03). GST increased by >50% above the upper limit of the reference range at a median of 1 day before the equivalent change in alanine transaminase in association with allograft rejection in the combined groups (95% confidence interval=1 to 2 days) but was lower on the day of diagnosis of rejection in the reporting group (P=0.02). This is compatible with the earlier diagnosis of rejection in the reporting group. CONCLUSIONS: We conclude that the monitoring of GST may improve patient care, reducing both mortality and morbidity.