Table 3.
Multivariate analysis demonstrating that the high risk of donor and/or recipient CMV seropositivity is compounded by PTCy regardless of donor source
| Variables | N | HR | 99% CI lower limit | 99% CI upper limit | P | Category P |
|---|---|---|---|---|---|---|
| Main effect variable | ||||||
| CMV serostatus | ||||||
| D−/R−, SibCNI* | 325 | 1.0 | <.0001 | |||
| R+, HaploCy | 551 | 49.8 | 13.7 | 181.2 | <.0001 | |
| R+, SibCy | 281 | 47.8 | 12.9 | 177.0 | <.0001 | |
| R+, SibCNI | 1077 | 24.0 | 7.0 | 82.2 | <.0001 | |
| D+/R−, HaploCy | 48 | 17.7 | 3.9 | 80.3 | <.0001 | |
| D+/R−, SibCy | 36 | 9.5 | 2.0 | 44.2 | .0002 | |
| D+/R−, SibCNI | 159 | 7.8 | 2.2 | 27.6 | <.0001 | |
| D−/R−, HaploCy | 129 | 1.5 | 0.2 | 8.9 | .59 | |
| D−/R−, SibCy | 78 | 1.8 | 0.2 | 12.9 | .46 | |
| Notable pairwise comparisons | ||||||
| R+, HaploCy vs R+, SibCy | 1.0 | 0.8 | 1.41 | .72 | ||
| R+, HaploCy vs R+, SibCNI | 2.1 | 1.4 | 3.11 | <.0001 | ||
| R+, SibCy vs R+, SibCNI | 2.0 | 1.3 | 3.12 | .0001 | ||
| D+/R−, HaploCy vs D+/R−, SibCy | 1.9 | 0.4 | 8.24 | .28 | ||
| D+/R−, HaploCy vs D+/R−, SibCNI | 2.3 | 0.9 | 5.52 | .02 | ||
| D+/R−, SibCy vs D+/R−, SibCNI | 1.2 | 0.4 | 4.19 | .69 | ||
| D−/R−, HaploCy vs D-/R−, SibCy | 0.8 | 0.3 | 2.24 | .63 | ||
| Other variables tested | ||||||
| Graft type | ||||||
| Bone marrow* | 612 | 1.0 | .79 | |||
| Peripheral blood | 2072 | 1.0 | 0.7 | 1.6 | .79 | |
| Donor-recipient sex match | ||||||
| Male-male* | 933 | 1.0 | .47 | |||
| Male-female | 608 | 1.2 | 0.9 | 1.5 | .12 | |
| Female-male | 659 | 1.1 | 0.8 | 1.4 | .46 | |
| Female-female | 484 | 1.1 | 0.8 | 1.4 | .54 | |
| Age at transplant, y | ||||||
| 0-20* | 230 | 1.0 | .04 | |||
| 21-40 | 475 | 1.8 | 1.0 | 3.1 | .009 | |
| 41-60 | 946 | 2.0 | 1.0 | 4.0 | .01 | |
| >60 | 1033 | 2.3 | 1.1 | 4.9 | .005 | |
| Karnofsky/Lansky performance at HCT | ||||||
| <80* | 1535 | 1.0 | .58 | |||
| 80-89 | 379 | 0.9 | 0.6 | 1.3 | .35 | |
| ≥90 | 770 | 1.0 | 0.7 | 1.4 | .90 | |
| Conditioning regimen intensity | ||||||
| Myeloablative* | 1422 | 1.0 | .03 | |||
| RIC/NMA | 1262 | 0.8 | 0.6 | 1.0 | .03 | |
| TBI | ||||||
| No* | 1519 | 1.0 | .08 | |||
| Yes | 1165 | 1.2 | 0.9 | 1.6 | .08 | |
| Time from diagnosis to transplant | ||||||
| <6 mo* | 1343 | 1.0 | .04 | |||
| 6 mo to 1 y | 643 | 1.1 | 0.9 | 1.4 | .13 | |
| ≥1 y | 698 | 1.2 | 1.0 | 1.5 | .02 | |
| Year of transplant | ||||||
| 2012-2014* | 1034 | 1.0 | .29 | |||
| 2015-2017 | 1650 | 0.9 | 0.7 | 1.2 | .29 | |
| Cytogenetics/IPSS-R | ||||||
| AML/ALL normal* | 146 | 1.0 | .5 | |||
| AML/ALL favorable | 78 | 0.7 | 0.3 | 1.4 | .18 | |
| AML/ALL intermediate | 888 | 1.1 | 0.7 | 1.7 | .68 | |
| AML/ALL poor | 682 | 1.1 | 0.8 | 1.7 | .39 | |
| MDS very low | 92 | 0.8 | 0.4 | 1.8 | .48 | |
| MDS low | 218 | 1.0 | 0.6 | 1.7 | .95 | |
| MDS intermediate | 230 | 1.3 | 0.8 | 2.2 | .21 | |
| MDS high | 109 | 1.2 | 0.7 | 2.2 | .43 | |
| MDS very high | 54 | 0.8 | 0.4 | 1.7 | .50 | |
| Other/not tested/missing | 187 | 0.8 | 0.4 | 1.5 | .35 |
We performed Cox proportional hazard models as a multivariate analysis to assess the combined impact of CMV serostatus, donor source, and PTCy on incidence of CMV infection (defined as CMV DNAemia ± organ disease). Because of the numerous variables and larger sample sizes, all analyzes are reported with 99% confidence intervals with a level of significance defined as P < .01. Recipient seropositivity (R+) conferred a high risk for CMV infection across all groups, which was doubled by PTCy use in both HaploCy and SibCy HCT. Donor-positive (D+)/R− serostatus also had a higher risk of CMV infection, which was increased by PTCy, more so in HaploCy HCT (see supplemental Data for pairwise comparisons between serostatus/graft type combinations). Other than serostatus, no other variables examined interacted with the outcome in question. IPSS-R, Revised International Prognostic Scoring System.
Reference cohort for each variable category analyzed.