Charlson-Deyo-Comorbidity-Index Predicts Long-Term Survival in Lung Transplantation

Purpose The Charlson-Deyo-Comorbidity-Index is a popular age independent score to predict long term survival. It incorporates 17 weighted comorbidity conditions and is established in a variety of clinical fields. However, the feasibility for lung transplantation is not yet tested. We therefore validate this score for primary graft dysfunction (PGD), short- to long-term survival and onset of chronic allograft dysfunction (CLAD). Methods The Charleston-Deyo-Comorbidity-Index was assessed in all 466 lung transplant recipients at our center between 1992-2015. Discriminative ability was calculated by the area under the ROC curve (AUC) for 30-day to 10-year survival, severe PGD (grade III) and onset of CLAD. An AUC of 0.5 was considered as non-predictive and 1.0 as maximally predictive. Multiple Cox-regression was used to detect independent risk factors for mortality in the score. All calculations were adjusted for unilateral transplantation, retransplantation, recipient age, idiomatic pulmonary fibrosis, preoperative intensive care unit stay and marginality of the donor (Oto-Donor-Score). Results Median Charslon-Deyo-Comorbidity-Index was 1 (range 1-5). PGD had an AUC of 0.65 (95%CI 0.59-0.72). The predictability increased from 30-day AUC0.58 (95%CI 0.48-0.69) over 1-year AUC0.67 (95%CI 0.60-0.73) to 10-year survival AUC0.72 (95%CI 0.67-0.78). A trend of decreased predictability was observed for onset of CLAD at 1-year AUC0.64 (95%CI 0.55-0.76) to 10-years AUC0.60 (95%CI 0.55-0.66). Multiple Cox-regression revealed the following comorbidities of the index as independent risk factors for mortality: congestive heart failure HR1.40 (95%CI 1.01-1.94), coronary disease needing intervention HR1.72 (95%CI 1.10-2.68), end-stage diabetes mellitus HR5.93 (2.12-19.58), moderate liver disease HR2.53 (95%CI 1.71-3.74) and peripheral vascular disease HR2.43 (95%CI 1.18-5.00). Conclusion The Charlson-Deyo-Comorbidity-Index has a fair prediction for PGD, mid-term survival and onset of CLAD and a good prediction for long term survival after lung transplantation. This score might be used for further clinical decision making in lung allocation. A different weighing and sub-categorization may even improve its prediction.