Statistical analysis

The data were expressed as the mean (standard deviation, SD), median (interquartile range, IQR), and percentage. SPSS21 (IBM Corporation) and GraphPad9 (GraphPad Software) were used for statistical calculations. The normality of each variable was evaluated using the Kolmogorov–Smirnov test. For normally distributed data, the comparison of two variables was performed using unpaired or paired two-tailed Student’s t-tests. A one-way ANOVA test followed by Tukey’s multiple comparisons test was applied for comparing two more independent samples. When the data were not normally distributed, the comparison of variables was performed with a Wilcoxon matched-pairs signed-rank test for paired or a Mann–Whitney U test for unpaired data, respectively. In the case of comparing two more independent samples, the Kruskal-Walli’s test followed by Dunn’s multiple comparisons test was applied. Correlation coefficients were calculated for nonparametric distributions using Spearman’s correlation test. The Chi-square test was used to compare categorical variables.

When comparing the incidence of poor immune reconstitution in different groups of CD70, the continuous variable of CD70 expression on CD4⁺ T cells was converted into a categorical variable according to the optimal cutoff value (with the highest Youden index) which was obtained by the Receiver operating characteristic (ROC) curve. The ROC analysis was also applied to detect the predicted potential. For recovery possibilities, the Kaplan-Meier survival plot with log-rank test P value was performed. The endpoint of Kaplan-Meier analysis was CD4⁺ T cell recovery, which was defined by the data of the first values of CD4⁺ T cell count ≥ 500 cells/μl. P < 0.05 was considered statistically significant.