Risk factors for late CMV disease and death 3 months after transplantation
| Covariate . | Late CMV disease . | Death . | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariate, RR . | 95% CI . | Multivariable, RR§ . | 95% CI . | Univariate, RR . | 95% CI . | Multivariable, RR∥ . | 95% CI . | |
| Absolute lymphocyte count after d 40, 100/mm3 or less | 2.5 | 1.1, 5.73-150 | — | — | 1.8 | 1.1, 3.03-150 | — | — |
| CD4 count at 3 mo, 50/mm3or less | 2.8 | 1.2, 6.33-150 | — | — | 1.9 | 1.1, 3.13-150 | — | — |
| CD8 count at 3 mo, 50/mm3 or less | — | — | — | — | 2.5 | 1.5, 4.23-151 | — | — |
| Undetectable CMV-specific CD4+ TH responses at 3 mo | — | — | — | — | 2.5 | 1.4, 4.53-150 | 2.1 | 1.1, 4.03-150 |
| CMV antigenemia before d 95 | ||||||||
| Positive, any level | 2.4 | 1.0, 5.73-150 | 3.4 | 1.2, 9.93-150 | — | — | — | — |
| More than 5 | 3.4 | 1.5, 7.53-151 | 3.5 | 1.5, 8.33-151 | 1.8 | 1.0, 3.13-150 | — | — |
| More than 10 | 2.6 | 1.1, 6.13-150 | 2.5 | 1.0, 6.153-150 | — | — | — | — |
| More than 100 | 3.8 | 1.1, 12.63-150 | — | — | — | — | — | — |
| Covariate . | Late CMV disease . | Death . | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariate, RR . | 95% CI . | Multivariable, RR§ . | 95% CI . | Univariate, RR . | 95% CI . | Multivariable, RR∥ . | 95% CI . | |
| Absolute lymphocyte count after d 40, 100/mm3 or less | 2.5 | 1.1, 5.73-150 | — | — | 1.8 | 1.1, 3.03-150 | — | — |
| CD4 count at 3 mo, 50/mm3or less | 2.8 | 1.2, 6.33-150 | — | — | 1.9 | 1.1, 3.13-150 | — | — |
| CD8 count at 3 mo, 50/mm3 or less | — | — | — | — | 2.5 | 1.5, 4.23-151 | — | — |
| Undetectable CMV-specific CD4+ TH responses at 3 mo | — | — | — | — | 2.5 | 1.4, 4.53-150 | 2.1 | 1.1, 4.03-150 |
| CMV antigenemia before d 95 | ||||||||
| Positive, any level | 2.4 | 1.0, 5.73-150 | 3.4 | 1.2, 9.93-150 | — | — | — | — |
| More than 5 | 3.4 | 1.5, 7.53-151 | 3.5 | 1.5, 8.33-151 | 1.8 | 1.0, 3.13-150 | — | — |
| More than 10 | 2.6 | 1.1, 6.13-150 | 2.5 | 1.0, 6.153-150 | — | — | — | — |
| More than 100 | 3.8 | 1.1, 12.63-150 | — | — | — | — | — | — |
Additional covariate factors considered in the univariate baseline model included patient age and sex, CMV donor serostatus before transplantation, HSV recipient serostatus before transplantation, underlying disease risk status, HLA donor matching (matched-related vs unrelated or mismatched donor), conditioning regimen, GVHD prophylaxis regimen, presence of neutropenia between engraftment and study entry, GVHD (acute grades 2-4 and 3-4; chronic clinical-extensive), use of ganciclovir (preemptive therapy vs ganciclovir prophylaxis), invasive aspergillosis, year of transplantation, and CMV viremia by culture.
Covariates for late CMV disease considered for inclusion in the final multivariable model were aspergillosis, GVHD prophylaxis, CD4 cell count lower than 50/mm3, GVHD (grades 2-4 or chronic clinical-extensive), lack of CMV-specific T-cell responses, and postengraftment lymphocytopenia (<100/mm3). For pp65 antigenemia before day 95, each cut-off variable was entered singly in a model adjusted for aspergillosis, GVHD prophylaxis, and CD4 cell count lower than 50/mm3.
Covariates for death considered for inclusion in the final multivariable model were CMV donor serostatus, aspergillosis, GVHD prophylaxis, underlying disease, donor type, CD4 cell count lower than 50/mm3, GVHD (grades 2-4 or chronic clinical-extensive), lack of CMV-specific T-cell responses, and postengraftment lymphocytopenia (<100/mm3). For pp65 antigenemia, each cut-off variable was entered singly in a model adjusted for donor CMV serostatus, GVHD prophylaxis, CMV-specific T-cell negativity, and CD4 cell count lower than 50/mm3.
— indicates results not significant.
P ≤ .05 (because of the number of statistical tests being carried out, values between .05 and .01 should be considered suggestive and those less than .01 considered significant).
P ≤ .01.