Abstract
In the era of cytomegalovirus(CMV)-preemptive therapy, it is unclear whether CMV serostatus of donor or recipient affects outcome of allogeneic hematopoietic stem cell transplantation (HSCT) among children with leukemia. To investigate, consecutive patients age 0–8 who underwent primary HSCT for acute leukemia in 1997–007 (HLA-matched sibling or unrelated donor, myeloablative conditioning, unmanipulated bone marrow or peripheral blood, preemptive therapy, no CMV prophylaxis) were followed retrospectively through January 2008. Treatment failure (relapse or death) was analyzed using survival-based proportional hazards regression. Competing risks (relapse and non-relapse mortality, NRM) were analyzed using generalized linear models of cumulative incidence-based proportional hazards. Excluding 4 (2.8%) patients lacking serostatus of donor or recipient, there were 140 subjects, of whom 50 relapsed and 24 died in remission. Pretransplant CMV seroprevalence was 55.7% in recipients, 57.1% in donors. Thirty-five (25.0%) grafts were from seronegative donor to seronegative recipient (D−/R−). On univariate analysis, D−/R− grafts were associated with shorter relapse-free survival (RFS) than other grafts (median 1.06 versus 3.15 years, p<0.05). Adjusted for donor type, diagnosis, disease stage, recipient and donor age, female-to-male graft, graft source, and year, D−/R− graft was associated with relapse (hazards ratio 3.15, 95% confidence interval 1.46–6.76) and treatment failure (2.45, 1.46–4.12) but not significantly with NRM (2.00, 0.44–9.09). In the current era, children who undergo allogeneic HSCT for acute leukemia have reduced risk of relapse and superior RFS when recipient and/or donor is CMV-seropositive before transplantation. However, no net improvement in RFS would be gained from substituting seropositive unrelated for seronegative sibling donors.
INTRODUCTION
In the current era of effective prophylactic and preemptive therapy, cytomegalovirus (CMV), once a leading infectious cause of death after hematopoietic stem cell transplantation (HSCT), is now an infrequent cause of early mortality. Yet donor or recipient CMV seropositivity may still confer a survival disadvantage, particularly when the graft is T-cell depleted.[1] Mechanisms proposed for an indirect adverse effect of CMV include virally-mediated immunosuppression (resulting in increased risk of bacterial and fungal infections)[2,3] and increased risk of acute graft-versus-host disease (GVHD)[4,5].
In order to prevent CMV transmission from seropositive donor to seronegative recipient, it has been recommended that CMV-seronegative patients receive grafts from seronegative donors whenever possible.[6] For a seropositive recipient, on the other hand, the choice of donor is currently controversial.[1,6,7] Some studies have reported a beneficial effect of seropositive donor, either reduction in relapse[8,9] or reduction in non-relapse mortality (NRM)[10,11], while other studies have found no benefit from seropositive donor.[1,2,12,13]
Three pediatric studies have investigated the effect of donor and recipient CMV serostatus on HSCT outcomes. In the first 2 studies (one of Philadelphia chromosome-positive chronic myeloid leukemia[14], the other of acute or chronic leukemia[15]), preemptive antiviral chemotherapy was not routinely used until the final years of study enrollment. No association was detected between relapse or NRM and donor or recipient serostatus[14] or seronegative donor-recipient pair[15]. In the third study, preemptive antiviral therapy was routine.[16] However, CMV prophylaxis was also standard; in addition, the sample was small and combined nonmalignant with malignant cases. In that study, the primary endpoint was CMV disease within 12 months after HSCT, but a possible association was noted (p=0.05) between recipient CMV seropositivity and increased NRM.
Thus for pediatric leukemia patients in the current era of preemptive therapy, it remains unclear whether CMV serostatus of donor and/or recipient affects the outcome of allogeneic HSCT. To investigate this question, we undertook a retrospective study among children with acute leukemia who underwent primary allogeneic HSCT with routine use of CMV-preemptive therapy.
METHODS
Sample
Consecutive patients age 0–18 who underwent primary, myeloablative, allogeneic HSCT were studied retrospectively with the approval of the medical center’s institutional review board. Eligible diagnoses were acute lymphocytic leukemia (ALL), acute myeloid or promyelocytic leukemia (AML), and myelodysplastic syndrome (MDS). Eligible donors were HLA-matched sibling (nonsyngeneic) or unrelated individuals. Eligible grafts were unmanipulated bone marrow or peripheral blood. Transplantations were performed between inception of the pediatric HSCT program in March 1997 and October 2007, and subjects were followed to relapse, NRM, or last contact through January 2008.
Surveillance for early CMV infection
Early CMV infection refers to viremia or disease with onset by day 100. Peripheral blood samples for CMV culture or polymerase chain reaction (PCR) were obtained twice weekly from day 21 through day 100. Specimens other than blood were obtained when clinically indicated. CMV viremia was defined as 1 positive culture, 2 consecutive positive PCR tests, or 1 quantitative PCR with viral load ≥5000/ml (or ≥1000/ml during high-dose corticosteroid therapy for acute GVHD). Per published guidelines[17], CMV disease was defined as clinical symptoms together with detection of CMV in fluid, lavage, or biopsy specimen from the affected organ, except CMV retinitis, which was diagnosed on retinal examination by an experienced ophthalmologist.
CMV-preemptive therapy
Preemptive therapy consisted of a week of induction therapy using either ganciclovir (5 mg/kg IV twice daily) or valganciclovir (450 mg/m2 orally twice daily), followed by 5 weeks of maintenance therapy with these same drugs given once daily, 5 days per week; foscarnet was used instead in 1 patient. CMV prophylaxis was not used. Standard HSCT procedures included acyclovir (pediatric dose 250 mg/m2 IV every 12 hours) from day −1 to day +25 as prophylaxis against varicella zoster and herpes simplex viruses.
Definitions
Disease stage was defined as early (AML and ALL in first complete remission and MDS subtype refractory anemia), intermediate (AML or ALL in second or subsequent complete remission or in first relapse), or advanced (AML or ALL in second or higher relapse or primary induction failure, MDS subtype refractory anemia with excess blasts or in transformation, or MDS, not otherwise classified). Acute GVHD specified cases that were grade 2–4 per Keystone Consensus Criteria[18].
Statistical analysis
Patients who did not achieve remission after transplantation were considered to have relapsed on day 1. No relapse or NRM was observed after 4 years from HSCT, by which milestone fewer than 20% of subjects remained in the cohort. Therefore, 4 years was the follow-up period chosen for study. Relapse-free survival was estimated using the Kaplan-Meier method[19]. Treatment failure (relapse or NRM) was analyzed using Cox proportional hazards regression[20].
Relapse and NRM constitute competing types of treatment failure: the occurrence of one type precludes the occurrence of the other. Cumulative incidence of competing risks was calculated and compared between groups as described by Gray[21]. In the presence of competing risks, the common practice in cancer studies of censoring one type of failure in order to model the other has been criticized as logically flawed.[22] Therefore, relapse and NRM were modeled using a methodology appropriate for competing risks: generalized linear models (complementary log-log link function, PROC GENMOD in SAS Version 9.1, SAS Institute Inc., Cary, North Carolina) of cumulative incidence-based proportional hazards were constructed using the pseudo-values approach of Klein and Andersen[23,24]. As in [24], a grid of 5 timepoints was used when calculating pseudo-values. Specifically, day 50, 90, 160, 240, and 540 demarcated approximately equal numbers of treatment failures per time period. The proportionality of hazards over time was verified.[24] If a hazard was time-dependent, a cutpoint was chosen among the 5 grid timepoints. A similar approach was used to model risk of acute GVHD through day 100, for which the competing risk was treatment failure.
Covariates
CMV serostatus of donor and of recipient were initially investigated using the 4 possible combinations of D/R serostatus. However, when 3 of the categories (D+/R+, D+/R−, and D−/R+) were associated with similar hazard ratios, these categories were combined to enhance statistical power and ease of interpretation. Thus, CMV serostatus was defined as D−/R− versus other graft. In addition to CMV serostatus of donor and recipient, standard covariates included in all models were diagnosis, disease stage at transplantation, donor type, recipient and donor age, female-to-male graft, graft source, and year of transplantation. Interaction between covariates was investigated. Selection of regimens for conditioning and GVHD prophylaxis was at physician discretion and varied according to donor type, disease stage, diagnosis, and year of transplantation. Therefore, treatment regimens were not considered as independent covariates.
Sensitivity analysis
To assess the robustness of the association between D−/R− graft and relapse, further models of relapse were constructed among subjects who were alive and relapse-free at day 90 (permitting history of early CMV infection and acute GVHD to be included as covariates), among subjects with similar disease (AML or ALL in first or second complete remission), and among subjects who received similar conditioning and GVHD prophylaxis.
RESULTS
Subjects
Four (2.8% of 144) eligible patients, relapse-free at last contact, were excluded for lack of information on CMV serostatus of donor (D) or recipient (R). In the study sample (n=140, median age 11 (range 0–18) years), the age distribution was bimodal, with peaks in infancy (age 1) and adolescence (age 12–14). Pretransplant CMV seroprevalence was similar among donors and recipients (57.1% and 55.7%, respectively). Distribution of CMV serostatus in D/R pairs was D+/R+ 37.9%, D+/R− 19.3%, D−/R+ 17.9%, and D−/R− 25.0%. Concordant serostatus (D+/R+ or D−/R−) was more common among sibling than unrelated pairs (70.5% vs 53.2%, chi-square test, p<0.05). Table 1 presents baseline characteristics for the sample overall and for subjects with and without the risk factor of current interest (D−/R− graft). No significant differences were present between subjects with D−/R− versus other grafts.
Table 1.
Baseline Characteristics of Pediatric HSCT Recipients (n=140), Overall and By CMV Serostatus of Donor/Recipient.
| Overall N=140 N (%) |
D−/R− Graft N=35 N (%) |
Other Graft N=105 N (%) |
|
|---|---|---|---|
| Male | 78 (55.7) | 20 (57.1) | 58 (55.2) |
| Diagnosis | |||
| ALL | 81 (57.9) | 18 (51.4) | 63 (60.0) |
| AML | 54 (38.6) | 16 (45.7) | 38 (36.2) |
| MDS | 5 (3.6) | 1 (2.9) | 4 (3.8) |
| Disease Stage | |||
| Early | 49 (35.0) | 14 (40.0) | 35 (33.3) |
| Intermediate | 81 (57.9) | 19 (54.3) | 62 (59.1) |
| Advanced | 10 (7.1) | 2 (5.7) | 8 (7.6) |
| Ethnicity | |||
| Hispanic White | 65 (46.4) | 11 (31.4) | 54 (51.4) |
| Non-Hispanic White | 56 (40.0) | 20 (57.1) | 36 (34.3) |
| Asian | 13 (9.3) | 2 (5.7) | 11 (10.5) |
| African-American | 6 (4.3) | 2 (5.7) | 4 (3.8) |
| Donor Type | |||
| Sibling | 78 (55.7) | 23 (65.7) | 55 (52.4) |
| Unrelated | 62 (44.3) | 12 (34.3) | 50 (47.6) |
| Donor Age | |||
| 0–19 Years | 74 (52.9) | 23 (65.7) | 51 (48.6) |
| 20–39 Years | 43 (30.7) | 8 (22.9) | 35 (33.3) |
| 40–55 Years | 23 (16.4) | 4 (11.4) | 19 (18.1) |
| Graft Source | |||
| Bone Marrow | 77 (55.0) | 18 (51.4) | 59 (56.2) |
| Peripheral Blood | 63 (45.0) | 17 (48.6) | 46 (43.8) |
| Year of Transplantation | |||
| 1997–2002 | 69 (49.3) | 18 (51.4) | 51 (48.6) |
| 2003–2007 | 71 (50.7) | 17 (48.6) | 54 (51.4) |
| Female-to-Male Graft | |||
| Yes | 32 (22.9) | 8 (22.9) | 24 (22.9) |
| No | 108 (77.1) | 27 (77.1) | 81 (77.1) |
| Conditioning Regimen | |||
| *Total Body Irradiation + Etoposide | 74 (52.9) | 17 (48.6) | 57 (54.3) |
| *Total Body Irradiation + †CTX(±other) | 45 (32.1) | 12 (34.3) | 33 (31.4) |
| Busulfan + †CTX | 21 (15.0) | 6 (17.1) | 15 (14.3) |
| GVHD Prophylaxis | |||
| Cyclosporin + Methotrexate(±other) | 81 (57.9) | 24 (68.6) | 57 (54.3) |
| Tacrolimus(±other) | 51 (36.4) | 11 (31.4) | 40 (38.1) |
| Cyclosporin±Mycophenolate Mofetil | 8 (5.7) | 0 | 8 (7.6) |
Irradiation was delivered in fractionated doses in 116/119 cases.
Cyclophosphamide
Loss to follow-up during the 4-year study window was 5.0% (7/140 subjects). Another 34 (24.3%) subjects were relapse-free within the study window when their follow-up was censored at the time of study closure. Thus follow-up of relapse-free survivors was median 3.22 years (minimum 90 days).
Engraftment failed to occur in 2 (1.4%) recipients (CMV D−/R− and D−/R+, respectively). Both patients had received bone marrow grafts from unrelated donors over age 40 during the early years of the study. Among recipients who did engraft, cumulative incidence of acute GVHD (prior to any relapse) was 39.1(±4.2)% at 100 days. Among such cases (n=54), time to onset was median 20 days (maximum 50 days, except in an outlier case which began on day 91). In multivariate analysis that adjusted for all covariates, acute GVHD was not associated with D−/R− graft (0.53, 95% CI 0.21–1.34, p=0.18) or with serostatus of donor or recipient.
At 100 days, cumulative incidence of CMV viremia or disease was 12.9(±3.0)%. Time to onset of early CMV infection (n=18) was median 42 (range 15 to 77) days. All but 1 CMV infection occurred in recipients seropositive before transplant; the remaining case occurred in a seronegative recipient with a seronegative donor (D−/R−). All cases occurred in relapse-free patients receiving high-dose corticosteroids (prednisone at least 15 mg bid, tacrolimus at least 1 mg per day, and/or cyclosporin >50 mg bid) as GVHD treatment or prophylaxis.
One third of early CMV infections developed into CMV disease (n=5 pneumonia, n=1 retinitis). Infection usually resolved with treatment but infrequently was a primary cause of death (n=2, CMV pneumonia) or persisted intermittently (n=1, asymptomatic infection) until death from vancomycin-resistant enterococcal pneumonia and sepsis. The latter 3 deaths occurred within 90 days after transplantations performed during the first half of the study.
Relapse and non-relapse mortality
During follow-up, relapse occurred in 50 subjects (2 of whom had persistent leukemia). NRM occurred in another 24 subjects, 14 of whom had infection as a primary or contributing case of death. On univariate analysis, median relapse-free survival differed between recipients of D−/R− (1.06 (95% CI 0.40–3.02) years) and other grafts (3.15 (95% CI 1.48-upper limit not reached) years, difference p<0.05) (Fig.1). When cumulative incidences of relapse and NRM were plotted (Fig.2), D−/R− graft was associated with cumulative incidence of relapse (Gray test[21], p=0.012) and not of NRM.
Figure 1. Relapse-free Survival, By Donor/Recipient CMV Serostatus (n=140).
Number of subjects at baseline, year 1, year 2, and year 3 were, for D−/R− graft, 35, 16, 8, and 6, and for other grafts, 105, 59, 44, and 33, respectively.
Figure 2. Cumulative Incidence of Relapse and Non-Relapse Mortality (NRM), By onor/Recipient CMV Serostatus (n=140).
Cumulative incidence is indicated using solid lines for relapse, dashed lines for NRM. Gray olor corresponds to D−/R− Graft, black to Other Grafts.
After adjustment for standard covariates, D−/R− graft remained a significant predictor of relapse (Table 2). The association with D−/R− graft did not vary with time. No interaction was present between D−/R− graft and other covariates, indicating that the effect of D−/R− graft on relapse was not limited to a particular subgroup.
Table 2.
Hazards Ratios (HR) of Relapse, NRM, and Treatment Failure Among Pediatric HSCT Recipients (N=140).
| ----------Relapse----------- | ------------NRM-------------- | ------Treatment Failure----- | ||||
|---|---|---|---|---|---|---|
| *HR (95% CI) | p value | *HR (95% CI) | p value | †HR (95% CI) | p value | |
| CMV Serostatus | ||||||
| D−/R− | 3.15 (1.46–6.76) | 0.003 | 2.00 (0.44–9.09) | 0.37 | 2.45 (1.46–4.12) | <0.001 |
| Other | 1.00 | 1.00 | 1.00 | |||
| Disease Stage | ||||||
| Early | 0.20 (0.09–0.46) | <0.001 | 0.54 (0.09–3.04) | 0.48 | 0.50 (0.28–0.88) | 0.02 |
| Other | 1.00 | 1.00 | 1.00 | |||
| Recipient Age | ||||||
| 0–7 Years | 4.28 (1.90–9.64) | <0.001 | 0.07 (0.01–0.68) | 0.02 | 1.58 (0.94–2.67) | 0.08 |
| 8–18 Years | 1.00 | 1.00 | 1.00 | |||
| Donor Type, Diagnosis | ||||||
| Unrelated, ALL | 0.96 (0.39–2.40) | 0.93 | 38.07 (4.82–300.8) | <0.001 | 2.51 (1.39–4.52) | 0.002 |
| Unrelated, AML/MDS | 1.48 (0.58–3.77) | 0.41 | 0.44 (0.02–8.02) | 0.58 | 1.22 (0.56–2.65) | 0.62 |
| Sibling, Any Diagnosis | 1.00 | 1.00 | 1.00 | |||
| Female-to-Male Graft | ||||||
| Yes | 0.37 (0.12–1.14) | 0.08 | 18.12 (4.39–74.84) | <0.0001 | ‡3.85 (1.20–12.34) | 0.02 |
| No | 1.00 | 1.00 | 1.00 | |||
| Donor Age | ||||||
| 40–55 Years | 0.75 (0.25–2.27) | 0.61 | 2.06 (0.38–11.28) | 0.40 | 1.87 (0.99–3.53) | 0.05 |
| 0–39 Years | 1.00 | 1.00 | 1.00 | |||
| Graft Source | ||||||
| Bone Marrow | 2.14 (0.93–4.93) | 0.07 | 0.29 (0.05–1.64) | 0.16 | 0.79 (0.48–1.29) | 0.35 |
| Peripheral Blood | 1.00 | 1.00 | 1.00 | |||
| Year of Transplantation | ||||||
| 1997–2002 | 1.27 (0.58–2.78) | 0.55 | 1.46 (0.41–5.14) | 0.56 | 1.43 (0.87–2.38) | 0.16 |
| 2003–2007 | 1.00 | 1.00 | 1.00 | |||
Unlike relapse, NRM was not significantly associated with D−/R− graft in multivariate analysis (Table 2). Of note, young recipient age had opposite effects on the 2 competing outcomes, increasing the risk of relapse while decreasing the risk of NRM.
When relapse and NRM were combined into a single outcome (treatment failure), D−/R− graft remained a significant predictor in multivariate analysis (Table 2). Female-to-male graft, whose effect did not vary over time in the model of NRM, was nevertheless a time-dependent covariate in the model of treatment failure, being statistically significant through the first 50 days only. During that early period, treatment failures were predominantly (7/12, 58.3%) NRM; thereafter, NRM constituted a minority (17/62, 27.4%) of treatment failures.
Sensitivity analysis
To determine whether the association between D−/R− graft and relapse was influenced by early CMV infection or acute GVHD, an additional model of relapse was constructed using relapse-free survivors at day 90 (n=115, including 36 who subsequently relapsed and 15 who died in remission). Among these day 90 survivors were 12 (10.4%) who had developed early CMV infection and 45 (39.1%) who had developed acute GVHD. Adjusted for standard covariates, relapse after day 90 remained associated with D−/R− graft (hazards ratio 2.65, 95% CI 1.18–5.98, p=0.02) and was not associated with early CMV infection or acute GVHD; neither variable altered the association between relapse and D−/R− graft.
Further models of relapse ascertained that the association with D−/R− graft did not depend on the study’s inclusion criteria. The association remained significant when the sample was restricted to patients with similar disease (AML or ALL in first or second complete remission, n=105), among whom the hazards ratio for D−/R− graft was 12.47, 95% CI 3.67–42.37 (p<0.0001). Likewise, the association remained significant when the sample was restricted to patients who received similar treatment (irradiation-based conditioning and GVHD prophylaxis with cyclosporin plus methotrexate or with a tacrolimus-based regimen, n=111), among whom the hazards ratio was 2.66, 95% CI 1.13–6.29 (p=0.026).
DISCUSSION
According to the current study, children who undergo allogeneic HSCT for acute leukemia in the era of CMV-preemptive therapy have reduced risk of relapse and superior relapse-free survival, with no increase in risk of NRM or acute GVHD, when the recipient and/or donor is seropositive for CMV prior to transplant. This finding suggests that the general recommendation to select seronegative donors for seronegative recipients[6] may not be appropriate for pediatric patients receiving non-depleted grafts in the current era.
For a seronegative recipient with seronegative sibling donor, the current analysis does not support replacing a seronegative sibling with a seropositive unrelated donor, even if one could be located without delay. According to the multivariate model of treatment failure (Table 2), such a substitution would achieve no net improvement in relapse-free survival, because the reduction in risk achieved by acquiring a seropositive donor, thus eliminating the D−/R− graft and its 2.45-fold hazard, would be negated by the increase in risk gained from acquiring an unrelated donor, most clearly in patients with ALL, where unrelated donor conferred a 2.51-fold hazard of treatment failure.
Neither early CMV infection nor acute GVHD explained or altered the current association between relapse and D−/R− graft, nor did the association depend on study inclusion criteria. The current incidence of early CMV infection (12.9±3.0%) was similar to a report of 11% DNAemia among pediatric recipients of unmanipulated grafts.[25] Consistent with reported predictors of early CMV infection [25–27], all but 1 current case occurred in seropositive recipients, and at onset of infection, all current cases were receiving high-dose corticosteroids as GVHD treatment or prophylaxis.
Both an adequate sample of pediatric leukemia patients and routine use of CMV-preemptive therapy without CMV prophylaxis may be essential for a study to be able to detect benefit from CMV seropositivity in recipient and/or donor. The requirement for a pediatric sample may arise from greater thymic regenerative capacity[28,29] and greater potential to mount a CMV-specific immune response[30] reported in children relative to adults. The requirement for preemptive therapy arises from the necessity to effectively control CMV and its complications before any potential benefit of CMV immunity or subclinical reactivation can be detected. Such benefit could be blocked by CMV prophylaxis.[16] Therefore, studies undertaken to verify the current association between D−/R− serostatus and relapse will need to exclude subjects receiving CMV prophylaxis in addition to preemptive therapy or else employ an analysis that controls for possible interaction between D−/R− serostatus and receipt of CMV prophylaxis.
The associations currently observed with D−/R− graft await confirmation from separate pediatric samples undergoing HSCT for acute leukemia with T-cell replete grafts and CMV-preemptive therapy but not CMV prophylaxis. Limitations of the current study sample include its relatively small size (insufficient for subgroup analyses, ie by HLA-A2 antigen status[8,13]), its origin at a single institution, and lack of data on cytogenetics and other factors potentially associated with relapse. Also, because the study was retrospective, it was not possible to investigate the mechanism underlying the current association between D−/R− graft and relapse, which likely involves poor graft-versus-leukemia response.
In conclusion, among pediatric patients undergoing HSCT for acute leukemia in the era of CMV-preemptive therapy, we observed a reduction in relapse with no increase in NRM when the recipient and/or donor was CMV-seropositive before transplantation. This finding runs counter to experience from the previous era.[1] However, with the current era has come the possibility that the effect of CMV serostatus on the outcome of HSCT may have changed, a possibility supported by evidence from the current pediatric study. If confirmed in a separate sample of children undergoing HSCT for acute leukemia, the finding of clinical benefit from CMV seropositivity in donor or recipient will be of use in the care and study of this patient population, informing donor selection, identification of patients at high risk of relapse, and design of clinical trials.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Boeckh M, Nichols WG. The impact of cytomegalovirus serostatus of donor and recipient before hematopoietic stem cell transplantation in the era of antiviral prophylaxis and preemptive therapy. Blood. 2004;103:2003–2008. doi: 10.1182/blood-2003-10-3616. [DOI] [PubMed] [Google Scholar]
- 2.Nichols WG, Corey L, Gooley T, Davis C, Boeckh M. High risk of death due to bacterial and fungal infection among cytomegalovirus (CMV)-seronegative recipients of stem cell transplants from seropositive donors: evidence for indirect effects of primary CMV infection. J Infect Dis. 2002;185:273–282. doi: 10.1086/338624. [DOI] [PubMed] [Google Scholar]
- 3.Boeckh M, Nichols WG. Immunosuppressive effects of beta-herpes viruses. Herpes. 2003;10:12–16. [PubMed] [Google Scholar]
- 4.Broers AE, van Der Holt R, van Esser JW, et al. Increased transplant-related morbidity and mortality in CMV-seropositive patients despite highly effective prevention of CMV disease after allogeneic T-cell depleted stem cell transplantation. Blood. 2000;95:2240–2245. [PubMed] [Google Scholar]
- 5.Cornelissen JJ, Carston M, Kollman C, et al. Unrelated marrow transplantation for adult patients with poor-risk acute lymphoblastic leukemia: strong graft-versus-leukemia effect and risk factors determining outcome. Blood. 2001;97:1572–1577. doi: 10.1182/blood.v97.6.1572. [DOI] [PubMed] [Google Scholar]
- 6.Ljungman P, Reusser P, de la Camara R, et al. Management of CMV infections: recommendations from the infectious diseases working party of the EBMT. Bone Marrow Transplant. 2004;33:1075–1081. doi: 10.1038/sj.bmt.1704505. [DOI] [PubMed] [Google Scholar]
- 7.Ljungman P. Risk assessment in haematopoietic stem cell transplantation: viral status. Best Pract Res Clin Haematol. 2007;20:209–217. doi: 10.1016/j.beha.2006.09.003. [DOI] [PubMed] [Google Scholar]
- 8.Nachbaur D, Clausen J, Kircher B. Donor cytomegalovirus seropositivity and the risk of leukemic relapse after reduced-intensity transplants. Eur J Haematol. 2006;76:414–419. doi: 10.1111/j.1600-0609.2005.00625.x. [DOI] [PubMed] [Google Scholar]
- 9.Jacobsen N, Badsberg JH, Lönnqvist B, et al. The Nordic Bone Marrow Transplantation Group. Graft-versus-leukaemia activity associated with CMV-seropositive donor, post-transplant CMV infection, young donor age and chronic graft-versus-host disease in bone marrow allograft recipients. Bone Marrow Transplant. 1990;5:413–418. [PubMed] [Google Scholar]
- 10.Ljungman P, Brand R, Einsele H, Frassoni F, Niederwieser D, Cordonnier C. Donor CMV serologic status and outcome of CMV-seropositive recipients after unrelated donor stem cell transplantation: an EBMT megafile analysis. Blood. 2003;102:4255–4260. doi: 10.1182/blood-2002-10-3263. [DOI] [PubMed] [Google Scholar]
- 11.Ringdén O, Schaffer M, Le Blanc K, et al. Which donor should be chosen for hematopoietic stem cell transplantation among unrelated HLA-A, -B, and -DRB1 genomically identical volunteers? Biol Blood Marrow Transplant. 2004;10:128–134. doi: 10.1016/j.bbmt.2003.10.001. [DOI] [PubMed] [Google Scholar]
- 12.Kollman C, Howe CW, Anasetti C, et al. Donor characteristics as risk factors in recipients after transplantation of bone marrow from unrelated donors: the effect of donor age. Blood. 2001;98:2043–2051. doi: 10.1182/blood.v98.7.2043. [DOI] [PubMed] [Google Scholar]
- 13.Erard V, Guthrie KA, Riddell S, Boeckh M. Impact of HLA A2 and cytomegalovirus serostatus on outcomes in patients with leukemia following matched-sibling myeloablative allogeneic hematopoietic cell transplantation. Haematologica. 2006;91:1377–1383. [PubMed] [Google Scholar]
- 14.Cwynarski K, Roberts IA, Iacobelli S, et al. Stem cell transplantation for chronic myeloid leukemia in children. Blood. 2003;102:1224–1231. doi: 10.1182/blood-2002-12-3637. [DOI] [PubMed] [Google Scholar]
- 15.Gustafsson Jernberg A, Remberger M, Ringdén O, Winiarski J. Risk factors in pediatric stem cell transplantation for leukemia. Pediatr Transplant. 2004;8:464–474. doi: 10.1111/j.1399-3046.2004.00175.x. [DOI] [PubMed] [Google Scholar]
- 16.Bordon V, Bravo S, Van Renterghem L, et al. Surveillance of cytomegalovirus (CMV) DNAemia in pediatric allogeneic stem cell transplantation: incidence and outcome of CMV infection and disease. Transpl Infect Dis. 2008;10:19–23. doi: 10.1111/j.1399-3062.2007.00242.x. [DOI] [PubMed] [Google Scholar]
- 17.Ljungman P, Griffiths P, Paya C. Definitions of cytomegalovirus infection and disease in transplant recipients. Clin Infect Dis. 2002;34:1094–1097. doi: 10.1086/339329. [DOI] [PubMed] [Google Scholar]
- 18.Przepiorka D, Weisdorf D, Martin P, et al. Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1994;1995(15):825–828. [PubMed] [Google Scholar]
- 19.Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481. [Google Scholar]
- 20.Cox DR. Regression models and life-tables (with discussion) J Royal Stat Soc. 1972;B34:187–220. [Google Scholar]
- 21.Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat. 1988;16:1140–1154. [Google Scholar]
- 22.Klein JP. Modelling competing risks in cancer studies. Stat Med. 2006;25:1015–1034. doi: 10.1002/sim.2246. [DOI] [PubMed] [Google Scholar]
- 23.Klein JP, Andersen PK. Regression modeling of competing risks data based on pseudovalues of the cumulative incidence function. Biometrics. 2005;61:223–229. doi: 10.1111/j.0006-341X.2005.031209.x. [DOI] [PubMed] [Google Scholar]
- 24.Klein JP, Gerster M, Andersen PK, Tarima S, Perme MP. SAS and R functions to compute pseudo-values for censored data regression. Comput Methods Programs Biomed. 2008;89:289–300. doi: 10.1016/j.cmpb.2007.11.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Matthes-Martin S, Lion T, Aberle SW, et al. Pre-emptive treatment of CMV DNAemia in paediatric stem cell transplantation: the impact of recipient and donor CMV serostatus on the incidence of CMV disease and CMV-related mortality. Bone Marrow Transplant. 2003;31:803–808. doi: 10.1038/sj.bmt.1703927. [DOI] [PubMed] [Google Scholar]
- 26.Paulin T, Ringdén O, Lönnqvist B, Wahren B, Nilsson B. The importance of pre bone marrow transplantation serology in determining subsequent cytomegalovirus infection. An analysis of risk factors. Scand J Infect Dis. 1986;18:199–209. doi: 10.3109/00365548609032328. [DOI] [PubMed] [Google Scholar]
- 27.Meyers JD, Fluornoy N, Thomas ED. Risk factors for cytomegalovirus infection after human marrow transplantation. J Infect Dis. 1986;153:478–488. doi: 10.1093/infdis/153.3.478. [DOI] [PubMed] [Google Scholar]
- 28.Mackall CL, Fleisher TA, Brown MR, et al. Age, thymopoiesis, and CD4+ T-lymphocyte regeneration after intensive chemotherapy. N Engl J Med. 1995;332:143–149. doi: 10.1056/NEJM199501193320303. [DOI] [PubMed] [Google Scholar]
- 29.Mackall CL, Hakim FT, Gress RE. Restoration of T-cell homeostasis after T-cell depletion. Semin Immunol. 1997;9:339–346. doi: 10.1006/smim.1997.0091. [DOI] [PubMed] [Google Scholar]
- 30.Lilleri D, Fornara C, Chiesa A, Caldera D, Alessandrino EP, Gerna G. Human cytomegalovirus-specific CD4+ and CD8+ T-cell reconstitution in adult allogeneic hematopoietic stem cell transplant recipients and immune control of viral infection. Haematologica. 2008;93:248–256. doi: 10.3324/haematol.11912. [DOI] [PubMed] [Google Scholar]