Development and validation of a nomogram for predicting the occurrence of renal dysfunction after treatment of immune checkpoint inhibitor: a retrospective case–control study

Abstract

Purpose The administration of immune checkpoint inhibitors (ICIs) may lead to renal adverse events, notably including renal dysfunction. To early predict the probability of renal dysfunction after ICIs therapy, a retrospective case–control study was conducted.Methods Clinical information on ICIs-treated patients was collected. Multivariable logistic regression was applied to identify risk factors for renal dysfunction after ICIs treatment. Moreover, a nomogram model was developed and validated internally.Results A total of 442 patients were included, among which 35 (7.9%) experienced renal dysfunction after ICIs treatment. Lower baseline estimated glomerular filtration rate (eGFR) (OR 0.941; 95% CI 0.917 to 0.966; p<0.001), concurrent exposure of platinum(OR 4.014; 95% CI 1.557 to 10.346; p=0.004), comorbidities of hypertension (OR 3.478; 95% CI 1.600 to 7.562; p=0.002) and infection (OR 5.402; 95% CI 1.544 to 18.904; p=0.008) were found to be independent associated with renal dysfunction after ICIs treatment. To develop a predictive nomogram for the occurrence of renal dysfunction after ICIs treatment, the included cases were divided into training and validation groups in a ratio of 7:3 randomly. The above four independent risk factors were included in the model. The area under the receiver operating characteristic curves of the predictiive model were 0.822 (0.723–0.922) and 0.815 (0.699–0.930) in the training and validation groups, respectively.Conclusions Lower baseline eGFR, platinum exposure, comorbidities of hypertension and infection were predictors of renal dysfunction in ICIs-treated patients with cancer. A nomogram was developed to predict the probability of renal dysfunction after ICIs treatment, which might be operable and valuable in clinical practice.