Long-Lasting Production of New T and B Cells and T-Cell Repertoire Diversity in Patients with Primary Immunodeficiency Who Had Undergone Stem Cell Transplantation: A Single-Centre Experience

Table 4

Multivariable logistic models investigating factors affecting the probability of having an unrestricted repertoire in HSCT patients.

Outcome and contributing factors	OR	SE
Probability of having an unrestricted repertoire
Model 1
TRECs under the cut-off	1	—	—
TRECs over the cut-off	3.2391	0.8649	0.0000
GvHD × time^a	1.0403	0.0119	0.0010
GvHD × time × time^b	0.9998	0.0001	0.0030
Model 2
TRECs under the cut-off	1	—	—
TRECs over the cut-off	2.8469	0.7616	0.0000
GvHD × time^a	1.1050	0.0354	0.0020
GvHD × time × time^b	0.9994	0.0354	0.0050
Ablative conditioning × GvHD × time^c	0.9305	0.0329	0.0420
Ablative conditioning × GvHD × time × time^d	1.0005	0.0002	0.0260

OR: odds ratio; SE: standard error.
Two models were fitted; reported here are only the significant factors and interactions; predictions of model 2 are depicted in Supplementary Figure  4.
GvHD: graft-versus-host disease.
TRECs: T-cell receptor excision circles.
^aTime expressed in months since transplantation; the interaction represents the effect of GvHD on the slope of the regression line (see Supplementary Figure  4); ^bthis quadratic interaction represents the effect of GvHD on the rate of change of the slope, that is, the curvature of the regression line; ^c,dthese interactions represent the change of the effect of GvHD on the slope due to the presence or absence of ablative conditioning.