Correction to: Germinal Immunogenetics predict treatment outcome for PD-1/PD-L1 checkpoint inhibitors

Correction to: Invest New Drugs (2020) 38(1):160–171

https://doi.org/10.1007/s10637-019-00845-w

Corrections are needed to the original version of this article. The in-text citations of tables and figures in the section “results” are not in sequential order. Moreover in section “results” (Associations between treatment outcome and SNPs), the reader should read “Almost half the patients (43.5%, N = 41) were in the low-risk group” and not “Almost half the patients (43.5%, N = 53) were in the low-risk group”.

The authors apologize for corrections and further state that changes to the text and the sequence of figures and tables does not affect the overall outcome of the study.

• Table 3 should show the univariate analysis for the 163 SNPs according to treatment outcome. Instead in the original paper it is presented in Table 4. For rs13900 the alleles associated with recessive model are “C/C or C/T” and not “C/C or T/T” as written in the table.

• Table 4 should show the univariate analysis for the 163 SNPs according to toxicity, instead in the original paper it is presented in Table 3.

• Table 5 should respectively show the weight of each SNP for multivariate predictive model according to treatment outcome and toxicity. Moreover in “Treatment outcome” the modality for rs13900 is T/T and not C/T.

• Table 6 should show the classification of patients based on risk group and risk evaluation of each group. For a better reading of table, the results “Treatment Outcome” and “Toxicity” have been reversed.

• Figure 1a should show the performance of the treatment predictive model, instead in the original paper it is presented in the Fig. 1b

• Figure 1b should show the performance of the toxicity predictive model, instead in the original paper it is presented in the Fig. 1a

Table 3 Univariate analysis for the 163 SNPs according to treatment outcome

Table 4 Univariate analysis for the 163 SNPs according to toxicity.

Table 5 Weight of each variable for multivariate predictive model

Table 6 Classification of patients based on risk group and risk evaluation of each group

Fig. 1

ROC curves representation and AUC estimation for predictive model. a Treatment outcome, b Toxicity

Fig. 2

Patients distribution in the different risk groups according on the predictive model. a Treatment outcome, b Toxicity