Development and validation of multivariable prediction models of serological response to SARS-CoV-2 vaccination in kidney transplant recipients

Abstract

Repeated vaccination against SARS-CoV-2 increases serological response in kidney transplant recipients (KTR) with high interindividual variability. No decision support tool exists to predict SARS-CoV-2 vaccination response to third or fourth vaccination in KTR. We developed, internally and externally validated five different multivariable prediction models of serological response after the third and fourth vaccine dose against SARS-CoV-2 in previously seronegative, COVID-19-naïve KTR. Using 20 candidate predictor variables, we applied statistical and machine learning approaches including logistic regression (LR), least absolute shrinkage and selection operator (LASSO)-regularized LR, random forest, and gradient boosted regression trees. For development and internal validation, data from 590 vaccinations were used. External validation was performed in four independent, international validation cohorts comprising 191, 184, 254, and 323 vaccinations, respectively. LASSO-regularized LR performed on the whole development dataset yielded a 20- and 10-variable model, respectively. External validation showed AUC-ROC of 0.840, 0.741, 0.816, and 0.783 for the sparser 10-variable model, yielding an overall performance 0.812. A 10-variable LASSO-regularized LR model predicts vaccination response in KTR with good overall accuracy. Implemented as an online tool, it can guide decisions whether to modulate immunosuppressive therapy before additional active vaccination, or to perform passive immunization to improve protection against COVID-19 in previously seronegative, COVID-19-naïve KTR.

A machine learning model for predicting serum neutralizing activity against Omicron SARS‐CoV‐2 BA.2 and BA.4/5 sublineages in the general population

“…The model used a small number of easily collectable variables, including age, vaccination schedule, and history of prior SARS‐CoV‐2 infection. Studies resembling ours have reported on ML models designed to predict SARS‐CoV‐2 seroconversion or suboptimal (arbitrarily defined) anti‐RBD antibody levels following vaccination in solid organ or allogeneic hematopoietic stem cell transplant recipients, which displayed variable precision 12–15 . For instance, via ML analysis, Papadopoulos et al 12 showed that age, body mass index and the presence of autoimmune diseases influenced the level of NtAbs against SARS‐CoV‐2 after vaccination.…”

“…For instance, via ML analysis, Papadopoulos et al 12 showed that age, body mass index and the presence of autoimmune diseases influenced the level of NtAbs against SARS‐CoV‐2 after vaccination. Osmanodja et al 13 applied ML algorithms to predict the positive serological response after third and fourth doses in kidney transplant recipients, but it was built and applied only in seronegative COVID‐naïve patients. Alejo et al 14 also built an ML model for the prediction of antibody levels after vaccination in solid organ transplant recipients.…”

“…[12][13][14][15] For instance, via ML analysis, Papadopoulos et al 12 showed that age, body mass index and the presence of autoimmune diseases influenced the level of NtAbs against SARS-CoV-2 after vaccination. Osmanodja et al 13 applied ML algorithms to predict the positive serological response after third and…”

“…This may be explained in part by the fact that the model was built using a cohort that was tested a median of 166 days Studies resembling ours have reported on ML models designed to predict SARS-CoV-2 seroconversion or suboptimal (arbitrarily defined) anti-RBD antibody levels following vaccination in solid organ or allogeneic hematopoietic stem cell transplant recipients, which displayed variable precision. [12][13][14][15] For instance, via ML analysis, Papadopoulos et al 12 showed that age, body mass index and the presence of autoimmune diseases influenced the level of NtAbs against SARS-CoV-2 after vaccination. Osmanodja et al 13 applied ML algorithms to predict the positive serological response after third and…”

“…4,[9][10][11] Supervised ML methods have been successfully employed to predict either qualitative or quantitative antibody responses elicited by COVID-19 vaccines in a variety of clinical settings. [12][13][14][15] Here, we explored the reliability of a ML approach (Classification and Regression Tree, CART) 16 to predict anti-SARS-CoV-2-RBD total antibody levels, which are strongly associated with presence of detectable NtAb responses against Omicron BA.2 and BA.4/5 sublineages in the general population.…”