Risk factors for late cytomegalovirus infection after allogeneic ... - Nature

4 downloads 306 Views 140KB Size Report
May 10, 2004 - Kyungpook National University Hospital, Daegu, 700-721, Korea. Summary: An increased incidence of late cytomegalovirus (CMV) infection ...
Bone Marrow Transplantation (2004) 34, 21–27 & 2004 Nature Publishing Group All rights reserved 0268-3369/04 $30.00

www.nature.com/bmt

Risk factors for late cytomegalovirus infection after allogeneic stem cell transplantation using HLA-matched sibling donor: donor lymphocyte infusion and previous history of early CMV infection DH Kim1,3, JG Kim1,3, NY Lee2, WJ Sung1, SK Sohn1,3, JS Suh2,3, KS Lee3 and KB Lee1,3 1

Department of Hematology/Oncology, Kyungpook National University Hospital, Daegu, 700-721, Korea; 2Department of Laboratory Medicine, Kyungpook National University Hospital, Daegu, 700-721, Korea; and 3Stem Cell Transplantation Center, Kyungpook National University Hospital, Daegu, 700-721, Korea

Summary: An increased incidence of late cytomegalovirus (CMV) infection has been reported during the last decade since the introduction of ganciclovir (GCV) prophylaxis or GCV pre-emptive therapy. Given that a donor lymphocyte infusion (DLI) can induce more severe GVHD, this may predispose a patient to late CMV infection. In all, 64 patients (median age 36, M/F 38/26) underwent allogeneic stem cell transplantation (SCT) using a matched sibling donor with bone marrow (n ¼ 9) or peripheral blood stem cells (n ¼ 55). The overall incidence of CMV infection, early and late CMV infection was 46.9 (30/64), 42.2 (27/ 64), and 16.4% (9/55), respectively. Early CMV infection was treated with GCV pre-emptive therapy that produced a 92.6% success rate. Among the 20 patients who received 35 DLIs, late CMV infection developed in eight (42.1%) of 19 evaluable cases with a median onset at 127 days post transplant. Risk factors for late CMV infection in a logistic regression analysis included DLIs (P ¼ 0.001) and a previous history of CMV infection (P ¼ 0.006). In conclusion, late CMV infection was strongly associated with DLIs and a previous history of early CMV infection. Accordingly, extended surveillance of CMV antigenemia is recommended for patients receiving DLIs or who have a previous history of CMV infection. Bone Marrow Transplantation (2004) 34, 21–27. doi:10.1038/sj.bmt.1704528 Published online 10 May 2004 Keywords: CMV infection; donor lymphocyte infusion; allogeneic PBSCT

An increased incidence of late cytomegalovirus (CMV) infection has been reported during the last decade since the introduction of ganciclovir (GCV) prophylaxis or GCV

Correspondence: Dr SK Sohn, Department of Hematology/Oncology, Kyungpook National University Hospital, 50 Samduk 2-ga, Jung-gu, Daegu, 700-721, Korea; E-mail: [email protected] Received 30 October 2003; accepted 16 January 2004 Published online 10 May 2004

pre-emptive therapy. Recently, the incidence of late CMV infection was reported in approximately 50% of patients after allogeneic stem cell transplantation (SCT).1 Boeckh et al1 noted that the most important risk factor for late CMV infection was the early detection of the CMV pp65 antigen within 3 months after SCT. In addition, other risk factors for late CMV infection include chronic graft-versushost disease (GVHD), delayed immune deficiency, or prolonged use of GCV.1,2 The pathogenesis of late CMV infection can be explained by the delayed recovery of CMV-specific T-cell immunity, which leaves patients at risk of late CMV infection when antiviral treatment is discontinued. Donor lymphocyte infusions (DLIs) are widely used to enhance the graft-versus-leukemia effect after allogeneic transplantation.3–5 However, a DLI can induce severe GVHD predisposing to an opportunistic infection. Thus, it is postulated that a DLI can predispose a patient to late CMV infection in the case of allogeneic recipients. Therefore, the current study defined the incidence and risk factors of early and late CMV infection after allogeneic SCT using matched sibling donors.

Patients and materials Study design The primary focus of the current study was to estimate the incidence of early and late CMV infection and the incidence of CMV infection following a DLI in allogeneic recipients. The clinical relevance of DLI and other risk factors with late CMV infection was investigated.

Transplantation procedure A review was conducted of the medical records of 64 patients who received allogeneic stem cell transplants based on an HLA-matched sibling donor from September 1998 to October 2002: nine patients (14.1%) received a bone marrow transplant (BMT) and 55 patients (85.9%) received a peripheral blood stem cell transplant (PBSCT). All cases were CMV IgG D þ /R þ , except for one (D/R þ ). Myeloablative conditioning was used in 52 cases (81.3%) and nonmyeloablative conditioning in 12 cases (18.7%).

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

22

The conditioning regimens were as follows: Busulfan (3.2 mg/kg/day i.v. in divided doses daily for 4 days) plus Cyclophosphamide (60 mg/kg once daily i.v. for 2 days, total dose 120 mg/kg) in 45 (70.3%), Fludarabine-based regimen in nine (14.1%), and other nonmyeloablative protocol in one (1.5%) with low-dose total body irradiation (200 cGy). Cyclophosphamide (50 mg/kg once daily i.v. for 4 consecutive days, total dose 200 mg/kg) plus equine antithymocyte globulin (ATG, Pharmacia & Upjohn, Kalamazoo, MI, USA. 30 mg/kg/day i.v. for 3 days) was used in nine patients (14.1%) with aplastic anemia. The PBSCs were harvested with a target of more than 4  106/ kg of CD34 after mobilization treatment with G-CSF alone or a G/GM-CSF combination or sequential injections from a matched sibling donor as reported previously.6 For the advanced disease group, additional PBSCs were collected and cryopreserved under 1701C for future DLIs in the case of a relapse or prophylactic use as previously reported.6,7 Acute GVHD prophylaxis consisted of cyclosporin A (Cipol-Ns, ChongKunDang Pharm., Seoul, Korea) beginning on day-1 and standard methotrexate therapy on D 1, 3, 6, and 11. For CMV prophylaxis, all patients received irradiated blood products that were depleted of leukocytes by filters, and infused intravenous immunoglobulin (500 mg/kg) every 2 weeks till day 100, and every month till 6 months. The infection prophylaxis consisted of ciprofloxacin (250 mg twice daily p.o.)/metronidazole (500 mg thrice daily p.o.)/fluconazole (100 mg once daily p.o.), beginning with the initiation of conditioning and acyclovir (600 mg twice daily p.o.) from day 1 until day þ 180. Co-trimoxazole was started after engraftment. Ursodeoxycholinic acid was used for the veno-occlusive disease (VOD) prophylaxis beginning with the initiation of conditioning. Chimerism studies with variable numbers of tandem repeats using a PCR with PB or BM mononuclear cells were performed according to protocol guidelines. Engraftment was confirmed by peripheral blood counts (myeloid: peripheral absolute neutrophil count of more than 0.5  109/l, megakaryocyte: peripheral platelet count of more than 20  109/l for 3 consecutive days).

Donor lymphocyte infusion Prophylactic DLIs were performed while continuing immunosuppression in six patients, whereas 14 relapsed patients were treated with chemotherapy followed by therapeutic DLIs as previously reported7,8 after stopping the immunosuppressants at least 2 weeks before DLI. The CD3 þ cell dose for each DLI was planned according to the disease status and purpose of the DLI. For a prophylactic DLI, the initial CD3 þ cell dose was planned within a range of 1–3  107/kg, with the next DLI at a fixed dose of 1  108/kg. For a therapeutic DLI in patients with relapsed CML, the initial CD3 þ cell dose was planned within a range of 1–3  107/kg, while for patients with relapsed AML or NHL, a fixed dose of 1  108/kg of CD3 þ cell was planned. No GVHD prophylaxis was administered in the postDLI period. Mild GVHD was not treated, yet moderate-toBone Marrow Transplantation

severe GVHD was treated with corticosteroids (methylprednisolone 1.2–5 mg/kg/day) or other immunosuppressive agents, including mycophenolate mofetil or tacrolimus. The pre- and post-DLI GVHD were classified according to the criteria proposed by Shulman et al.9

CMV antigenemia monitoring CMV antigenemia assay was performed every week till day 100, every 2 weeks till 6 months, and every 2 or 4 weeks till 12 months since engraftment. CMV antigenemia surveillance was extended when the late CMV infection or recurrent episode of CMV infection was documented or clinical sign of GVHD was observed. CMV antigenemia assay was performed as described elsewhere10 and determined as positive when any polymorphonuclear leukocytes (PMNLs) were stained in 200 000 PMNLs per slide.

CMV antigenemia-guided ganciclovir pre-emptive therapy and CMV treatment When positive CMV pp65 antigenemia was detected, preemptive ganciclovir therapy (5 mg/kg i.v. twice daily) was administered daily for at least 2 weeks until antigenemia resolved, followed by 5 mg/kg/day i.v. every other day for another week. Dose adjustment was made if renal function was impaired. Acyclovir was stopped during pre-emptive therapy. If leukopenia occurred (WBC less than 1.5  109/ l), ganciclovir was temporarily stopped and CSF was given for at least two consecutive days. CMV disease was treated with ganciclovir 5 mg/kg i.v. twice daily for at least 3 weeks followed by 5 mg/kg i.v. for 5 days a week until clinical signs of CMV disease improved and CMV antigenemia was undetectable. If CMV antigenemia persisted, the treatment was continued with a weekly antigenemia assay.

Definition and statistics CMV infection was defined as any positive result of CMV pp65 antigenemia without clinical signs and symptoms. Early CMV infection was defined as a CMV pp65 antigenemia occurring before day 100 after SCT, late CMV infection thereafter. CMV disease was defined as a symptomatic CMV infection, including CMV pneumonia, CMV gastroenteritis, or CMV retinitis. CMV pneumonia was diagnosed based on respiratory signs and symptoms with interstitial infiltrates on the chest X-ray in conjunction with a positive CMV culture from bronchoalveolar lavage or a proven histologic demonstration in a tissue specimen. The diagnosis of gastrointestinal CMV disease was based on endoscopic signs of bowel inflammation and histologic demonstration of CMV inclusions, or positive in situ hybridization assay. The clinical characteristics of the patients with or without early or late CMV infection were compared using Mann–Whitney’s U-test or Fisher’s exact test as appropriate. Logistic regression analysis was used to determine the risk factors for the development of early or late CMV infection. Those patients who did not survive more than

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

23

100 days after SCT were excluded from the analysis of late CMV infection in the current study. The parameters used to analyze the risk factor of early and late CMV infection were as follows; age and sex of recipient, MNC and CD34 cell dose, diagnosis, advanced disease status (more than CR2, relapse or primary refractoriness at transplant) vs standard risk disease status, conditioning regimen using ATG vs non-ATG regimen, nonmyeloablative regimen vs myeloablative regimen, PBSCs as the stem cell source vs BM, acute GVHD, chronic GVHD, and the GVHD treatment with mycophenolate mofetil or high-dose corticosteroids. The incidences of early and late CMV infection were estimated using the Kaplan–Meier method. A logrank test was used for a survival analysis of the incidence of late CMV infection. A cutoff value of 0.05 was adopted for all the statistical analyses. The statistical data were obtained using the SPSS software package (SPSS 10.0 Inc., Chicago, IL, USA).

Results With a median follow-up of 13 months (range 1–54 months), a total of 36 episodes of CMV infection were noted in 30 (46.9%) of 64 patients and the overall incidence of a first episode of CMV infection 1-year post transplant was 51.5% (Table 1).

Early CMV infection and risk factor The incidence of early CMV infection was 42.2% (27/64). In all, 27 patients received GCV pre-emptive therapy for early CMV infection. Among them, two patients (7.4%) who progressed to CMV diseases were enrolled on the CMV treatment protocol. Among the 25 patients who received a 2-week course of GCV pre-emptive therapy, CMV infection was completely resolved in 22 patients (88.0%), while the other three patients received an extended GCV therapy schedule beyond 2 weeks until a negative conversion of CMV antigenemia. Three patients (11.1%) developed leukopenia during GCV pre-emptive therapy. The patients’ characteristics between the groups that developed early CMV infection or not, were summarized in Table 1. Acute GVHD was found to be an independent risk factor for the development of early CMV infection (P ¼ 0.0042, relative risk [RR] 11.926).

Late CMV infection Among 55 patients in whom CMV status was monitored beyond 100 days after transplant, nine late CMV infections (16.4%) were observed with six recurrent and three first episodes. The patients’ characteristics between the groups that developed late CMV infection or not, were summarized in Table 1. In a multivariate analysis using a logistic regression model, the group that received DLIs (P ¼ 0.001, RR 109.926, 95% C.I. [6.072, 1990.13]), and a group with previous history of early CMV infection (P ¼ 0.006, RR 25.486, 95% CI [2.548, 254.863]) had a higher risk of late CMV infection, respectively (Table 1). The advanced disease group was also identified as a potential risk factor

for late CMV infection in univariate analysis, but not statistically significant in multivariate analysis. The cumulative incidence of late CMV infection between groups that did and did not receive DLI were shown in Figure 1. In the group receiving DLI, the incidence of late CMV infection at 2-year post transplant was 58.4%, while in the group not receiving DLI 4.2%. One case developed late CMV infection without prior DLI was successfully treated with GCV pre-emptive therapy. Among the eight patients who developed late CMV infections after DLIs, three patients were successfully treated with GCV pre-emptive therapy, and one patient died of CMV pneumonia refractory to GCV. Four patients with CMV antigenemia of less than 5/200 000 PMNLs were treated with G-CSF instead of GCV because they were severely myelosuppressed when their CMV antigen appeared and then came to exhibit undetectable CMV antigen levels at the follow-up of 1-week interval.

DLIs and late CMV infection In total, 20 patients (31.3%) received 35 DLIs for prophylactic (n ¼ 6) or therapeutic purposes (n ¼ 14). The median total CD3 þ cell dose was 1.31  108/kg (range 0.06–2.90) with a median of two DLIs (range 1–3). The median day for the DLIs and the diagnosis of late CMV infection after SCT was 127 (range 28–660) and 186 (range 114–564) days, respectively. The characteristics of the patients are summarized in Table 2. The clinical descriptions for the six cases that received prophylactic DLIs were as follows: four cases of acute leukemia in a refractoriness or relapse status at transplant, one case of non-Hodgkin’s lymphoma that relapsed after autologous SCT, and one case of refractory colon cancer that recurred with lung metastasis and was treated with a pDLI. In 14 cases that received therapeutic DLIs, 11 patients with AML, one with NHL, one with CML, and one with MDS were included. The grades of chronic GVHD before and after the DLIs are listed in Table 2. Among the 20 patients who received a DLI, seven patients (35%) exhibited limited chronic GVHD before the DLI. After the DLIs, limited and extensive chronic GVHD were documented in eight (42%) and eight patients (42%), respectively.

CMV diseases Among the 30 patients who developed CMV antigenemia, four episodes of CMV disease, including two cases of CMV pneumonia, one case of CMV enteritis, and one case of CMV esophagitis, were noted in the current study. As regards the two cases of CMV pneumonia, one patient who received a prophylactic DLI on day 66 after SCT became positive to CMV antigenemia on day 243. This patient finally died of CMV pneumonia with limited chronic GVHD, yet had no previous history of early CMV infection. The second patient showed an early CMV infection on day 54 that was resolved with GCV preemptive therapy, yet 52 days after the complete resolution of CMV antigenemia (day 142 post transplant), CMV pneumonia developed with a flare up of GVHD and the patient died 196 days post transplant. The patient with Bone Marrow Transplantation

24

Bone Marrow Transplantation

Table 1

Patient characteristics for cases with early cytomegalovirus infection and late cytomegalovirus infection (total 37 episodes) Early CMV infection (n ¼ 27/64)

Category

Sex Age Diagnosis

Conditioning Cell source Cell dose

GVHD

GVHD treatment

(M/F, % of male) (median, range) Aplastic anemia Acute leukemia Chronic leukemia Others Advanced disease status NST Use of ATG BM (n ¼ 9) PBSC (n ¼ 55) MNC (  108/kg) CD34+cell (  106/kg) Acute GVHD Acute GVHD, Gr 2–4 Chronic GVHD Extensive GVHD MMF

High-dose steroids Relapse Death Previous history of CMV infection Outcome

DLI

EI () n ¼ 37, 57.8%

16/11(59.3%) 22/15(59.5%) 37(16–56) 36(17–53) 4(14.8%) 5(13.5%) 17(63.0%) 20(54.1%) 4(14.8%) 4(10.8%) 2(7.4%) 8(21.6%) 9(33.3%) 18(48.6%) 5(18.5%) 7(25.9%) 6(22.2%) 21(77.8%) 7.6570.95 7.1971.19

P-value

Relative risk, [95% CI]

NS NS NS

— —

NS



NS NS NS

— — —

LI (+) n ¼ 9, 16.4%

LI () n ¼ 46, 83.6%

P-value

Relative risk, [95% CI]

6/3(66.7%) 39(29–56) 1(11.1%) 6(66.7%) 1(11.1%) 1(11.1%) 7(77.8%)

28/18(60.9%) 36(16–54) 8(17.5%) 25(54.3%) 7(15.2%) 6(13.0%) 15(32.6%)

NS NS NS



0.022

7.24 [1.33,38.46]

3(33.3%) 1(11.1%) 1(11.1%) 8(88.9%) 8.3071.38 5.7973.89

8(17.4%) 10(21.7%) 5(10.9%) 41(89.1%) 8.1370.63 9.1772.15

NS NS NS

— — —

8/9(88.9%)

33/46(71.7%)

NS



7/9(77.8%)

27/46(58.7%)

NS



NS NS NS

— — — —

7(18.9%) 5(13.5%) 3(8.1%) 34(91.9%) 7.4170.63 10.9572.70

NS 0.027

24/26(88.9%) 22/36(61.1%)

0.014a

22/26(84.6%) 18/36(50.0%)

0.005

11.926 [1.094,130.015] —

13/24(54.2%) 17/32(53.1%) 7/24(29.2%) 10/32(31.3%) 14/37(37.8%) 3/27(11.1%)

NS NS 0.022

— — —

6/9(66.7%) 3/9(33.3%) 14/46(30.4%)

24/45(53.3%) 14/45(31.1%) 2/9(22.2%)

NS 0.031 NS

— — —

5/46(10.9%) 5(55.6%) 4(44.4%) 6(66.7%)

1/9(11.1%) 18(39.1%) 18(39.1%) 17 (37.0%)

NS NS NS 0.098b

8(88.9%)

11(23.9%)

0.001c

4/37(10.8%) 6(22.2%) 10(37.0%)

6/27(22.2%) 18(48.6%) 20(54.1%)

EI ¼ early CMV infection; LI ¼ late CMV infection; NST ¼ nonmyeloablative stem cell transplantation; MMF ¼ mycophenolate mofetil; NS ¼ not significant. a P ¼ 0.042 in logistic regression analysis for the risk factor of early CMV infection. b P ¼ 0.006 in logistic regression analysis for the risk factor of late CMV infection. c P ¼ 0.001 in logistic regression analysis for the risk factor of late CMV infection.



NS NS

— 25.486 [2.548,254.863] 109.926 [6.072,1990.13]

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

EI (+) n ¼ 27, 42.2%

Late CMV infection (n ¼ 9/55)

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

25

CMV enteritis tested positive to CMV antigenemia on day 65 and died of gastrointestinal and CNS bleeding on day 71. The patient with CMV esophagitis tested positive to CMV antigenemia on day 22, yet this was completely resolved with GCV pre-emptive therapy. However, the patient died of exacerbated refractory acute GVHD with the reappearance of CMV antigenemia on day 83 post transplant.

Incidence of late CMV infection

1.0

0.8 DLI (+)

0.6

0.4

0.2 DLI (−) 0.0 60 120 180 240 300 360 420 480 540 600 660 720 Days after transplantation

Figure 1 Cumulative incidence of late cytomegalovirus infection in the groups that did (n ¼ 19) or did not receive donor lymphocyte infusions (n ¼ 36). The cumulative incidence of late CMV infection of two groups at 2-year post transplant was 58.4 and 4.2%, respectively.

Discussion CMV infection is one of the most troublesome complications after allogeneic SCT. CMV infection is commonly seen in the early phases post transplant. A possible explanation for this is the reactivation of latent CMV infection due to the severely suppressed immune status in the recipient and uncontrolled CMV proliferation in the absence of CMV-specific immunoprotective CD8 þ cells after transplantation.11 Since the introduction of GCV prophylaxis or pre-emptive therapy in allogeneic transplantation, late CMV infection has become an important cause of late mortality in recipients. Exposure to GCV and treatment efforts for GVHD can delay the recovery of CMV-specific T-cell immunity after allogeneic SCT, which can persist after the discontinuation of GCV therapy, predisposing a patient to late CMV infection.12 The incidence of late CMV infection has been reported in up to approximately 50% of patients after allogeneic SCT in the Boeckh et al1 study, which was higher than that of 16.4% in the current study. However, they were not all matched sibling donors (a mixture of sibling and unrelated donors), while the current study focused on matched sibling transplantation and 85.9% of patients received peripheral blood as the stem cell source, which may improve immune reconstitution, thus preventing CMV infection. Among the nine patients who developed late CMV infection, only one patient eventually progressed to a CMV disease. For the eight remaining patients, there was no evidence of clinical CMV disease whether they were treated with GCV therapy or not. This result would seem to imply that late CMV infection has a lesser potential of developing into a clinical

Table 2 Summary of donor lymphocyte infusions, cytomegalovirus infection, graft-versus-host disease status before and after donor lymphocyte infusions in patients who received donor lymphocyte infusions (n ¼ 20) UPN

Donor lymphocyte infusion Day Purpose Times Diagnosis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

34 38 44 45 54 98 135 147 209 278 304 660 28 66 99 119 143 172 185 265

pDLI pDLI tDLI tDLI pDLI tDLI tDLI tDLI tDLI tDLI tDLI tDLI pDLI pDLI tDLI pDLI tDLI tDLI tDLI tDLI

2 2 1 3 2 1 2 2 1 2 1 2 2 2 2 1 3 1 1 2

AML, M2, refractory ALL, L2, refractory AML, M7, relapse NHL, DLBL, refractory NHL, MCL, relapse AML, M2, relapse CML, AP AML, mixed, complex AML, M2, 5q AML, M2, complex MDS, CMML AML, M6 Colon cancer, relapse AML, M0, refractory AML, M2, refractory AML, M2, relapse AML, mixed, ph+ AML, M2, refractory AML, M2, del (7q) AML, M2

CMV infection

GVHD

Current status

CD3+ (  108/kg) Early infection Late infection Pre-DLI Post-DLI 2.90 0.95 1.12 1.88 1.33 0.30 0.39 0.06 1.57 0.67 0.97 1.32 1.50 1.67 1.70 0.35 1.43 1.29 2.05 1.12

            +(77)  +(18)  +(29) +(41) +(86) 

  NE          +(126) +(243) +(114) +(203) +(174) +(564) +(198) +(309)

No No No No No Limited No No Limited No No No Limited No No Limited Limited Limited No Limited

Extensive No NE Limited Extensive Extensive Limited Extensive Limited Limited Limited Limited Limited Limited No Extensive Extensive Extensive No Extensive

159 113 56 260 1510 546 276 238 271 340 829+ 1113 138 316 262 1547+ 218+ 311+ 274 351

NE ¼ not evaluable; UPN ¼ unique patient number; pDLI ¼ prophylactic DLI; tDLI ¼ therapeutic DLI; AML ¼ acute myeloid leukemia; NHL ¼ nonHodgkin’s lymphoma; DLBL ¼ diffuse large B cell lymphoma; MCL ¼ mantle cell lymphoma; MDS ¼ myelodysplastic syndrome; CMML ¼ chronic myelomonocytic leukemia; AP ¼ accelerated-phase; complex, complex cytogenetic abnormalities. The column of CMV infection notes the onset day of CMV infection inside parenthesis. The current status column showed status of the patients ( alive, + dead) with the last follow-up days after SCT.

Bone Marrow Transplantation

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

26

CMV disease. The role of CMV-specific donor T-cell activity may have a more important role than that of host immunity in the development of late CMV infection after DLIs, which was also observed by Krause et al,11 who reported that the patients lacking a proliferative response of CMV-specific T cells after day 120 post transplant have a risk of at least 31% of late CMV disease. Hakki et al13 also reported that subclinical CMV reactivation may stimulate CMV-specific immunity, thereby preventing the development of overt CMV diseases, which may explain the relatively low incidence of late CMV disease in the current study. The present study also found that most cases of late CMV infection had a previous history of CMV infection, which stimulated CMV-specific cytotoxic T-cell immunity. Since mortality related to CMV infection is high as 50% in an allogeneic setting, monitoring the CMV status regularly is very important even in the late phase of transplant so as to enable treatment of CMV viremia as early as possible. For the early detection of late CMV infection, knowledge of the high risks related to late CMV infection is needed. In the current study, it was found that the use of DLIs and a previous history of early CMV infection were independent risk factors for predicting late CMV infection after allogeneic SCT. Boeckh et al1 also reported that the detection of the CMV pp65 antigen within 3 months after SCT was a high-risk factor predicting late CMV infection. Meanwhile, Peggs et al14 reported that the risk of late CMV infection was minimal if no CMV infection was documented before day 100 in CMV-seropositive patients. As briefly mentioned above, a possible explanation for this outcome is that GCV delays the recovery of CMV-specific T-cell immunity, thereby leaving patients at risk of late CMV infection when the antiviral treatment is discontinued. Patients receiving GCV therapy for early CMV infection exhibit a decreased proliferative response to CMV because of the myeloid toxicity associated with GCV.11 In the current study, DLIs were strongly related with the development of late CMV infection. One of the concerns related to a DLI in terms of CMV infection is that the additional infusion of donor-derived lymphocytes can increase the CMV viral load in the recipients. Boeckh et al15 reported that the risk of transmitting CMV via the marrow or stem cell products from a seropositive donor is approximately 15–20%, suggesting that the CMV viral load is related with the incidence of late CMV disease. Gor et al16 also reported that the viral load is the major factor associated with CMV disease independent of acute GVHD and the recipient’s CMV serostatus. Even though most cases have a D þ /R þ CMV serostatus, the DLI procedure would be expected to increase the viral load, thereby increasing the incidence of late CMV infection. In the current study, among the nine patients who developed CMV infection after DLIs, three patients (33.3%) manifested late CMV infection as the first episode. The development of CMV infection after DLIs in these four patients may have been related to an increased viral load through peripheral blood lymphocytes independent of a previous history of CMV infection. The treatment effort for GVHD induced or aggravated by DLIs may also have placed an important role in the development of late CMV Bone Marrow Transplantation

infection in our four patients who manifested moderate-tosevere GVHD after DLIs. The advanced disease group was associated with increasing late CMV infection (P ¼ 0.022) in univariate analysis for the cumulative incidence of late CMV infection, but not statistically significant in multivariate analysis. The patients with advanced disease were already exposed to multiple chemotherapeutic agents and had a history of more transfusions that could increase the risk of CMV transmission. Moreover, prophylactic or therapeutic DLIs were performed much more frequently in patients with advanced disease at high risk of relapse. Therefore, based on the results of the current study, it is strongly recommended that CMV surveillance should be extended beyond day 100 after SCT, especially in the case of patients receiving DLIs or with a previous history of CMV infection. Patients with both these risk factors have a particularly high risk of late CMV infection, thus more stringent monitoring of CMV surveillance is encouraged. Recently, several investigators emphasized extended CMV surveillance beyond day 100 after SCT.14,17–19 According to Junghanss et al,18 weekly CMV surveillance and preemptive therapy with GCV should be continued for at least 1 year in patients who had GVHD and/or CMV reactivation before day 100. Peggs et al14 recommended extended surveillance in cases that tested positive in a CMV PCR before day 100, and proposed selective surveillance in cases with a history of early CMV infection. Machado et al17 also recommended extended surveillance until day 365 for the patients who developed chronic GVHD. Accordingly, the current study found that late CMV infection was strongly associated with DLIs and a previous history of early CMV infection. As such, extended surveillance of CMV antigenemia for patients receiving DLIs, with a previous history of CMV infection history is recommended.

Acknowledgements The valuable work of our nursing staff is gratefully acknowledged.

References 1 Boeckh M, Leisenring W, Riddell SR et al. Late cytomegalovirus disease and mortality in recipients of allogeneic hematopoietic stem cell transplants: importance of viral load and T-cell immunity. Blood 2003; 101: 407–414. 2 Chakrabarti S, Mackinnon S, Chopra R et al. High incidence of cytomegalovirus infection after nonmyeloablative stem cell transplantation: potential role of Campath-1 H in delaying immune reconstitution. Blood 2002; 99: 4357–4363. 3 Kolb HJ, Holler E. Adoptive immunotherapy with donor lymphocyte transfusions. Curr Opin Oncol 1997; 9: 139–145. 4 Kolb HJ, Mittermuller J, Clemm C et al. Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 1990; 76: 2462–2465. 5 Kolb HJ, Schattenberg A, Goldman JM et al. Graft-versusleukemia effect of donor lymphocyte transfusions in marrow grafted patients. European Group for Blood and Marrow

Impact of DLI on late CMV infection after allogeneic SCT DH Kim et al

27

6

7

8

9

10

11

12

Transplantation Working Party Chronic Leukemia. Blood 1995; 86: 2041–2050. Sohn SK, Kim JG, Chae YS et al. Large-volume leukapheresis using femoral venous access for harvesting peripheral blood stem cells with the Fenwal CS 3000 Plus from normal healthy donors: predictors of CD34+ cell yield and collection efficiency. J Clin Apheresis 2003; 18: 10–15. Sohn SK, Jung JT, Kim DH et al. Prophylactic growth factorprimed donor lymphocyte infusion using cells reserved at the time of transplantation after allogeneic peripheral blood stem cell transplantation in patients with high-risk hematologic malignancies. Cancer 2002; 94: 18–24. Kim JG, Sohn SK, Kim DH et al. A pilot study of cytoreductive chemotherapy combined with infusion of additional peripheral blood stem cells reserved at time of harvest for transplantation in case of relapsed hematologic malignancies after allogeneic peripheral blood stem cell transplant. Bone Marrow Transplant 2003; in press. Shulman HM, Sullivan KM, Weiden PL et al. Chronic graftversus-host syndrome in man. A long-term clinicopathologic study of 20 Seattle patients. Am J Med 1980; 69: 204–217. Boeckh M, Bowden RA, Goodrich JM et al. Cytomegalovirus antigen detection in peripheral blood leukocytes after allogeneic marrow transplantation. Blood 1992; 80: 1358–1364. Krause H, Hebart H, Jahn G et al. Screening for CMV-specific T cell proliferation to identify patients at risk of developing late onset CMV disease. Bone Marrow Transplant 1997; 19: 1111–1116. Junghanss C, Storb R, Boeckh M. Risk of cytomegalovirus infection after transplantation with nonmyeloablative and reduced intensity conditioning regimens. Blood 2002; 100: 2675–2676.

13 Hakki M, Riddell SR, Storek J et al. Immune reconstitution to cytomegalovirus after allogeneic hematopoietic stem cell transplantation: impact of host factors, drug therapy, and subclinical reactivation. Blood 2003; 102: 3060–3067. 14 Peggs KS, Preiser W, Kottaridis PD et al. Extended routine polymerase chain reaction surveillance and pre-emptive antiviral therapy for cytomegalovirus after allogeneic transplantation. Br J Haematol 2000; 111: 782–790. 15 Boeckh M, Riddel SR, Woogerd P et al. Primary CMV infection via marrow: incidence, response to early treatment, CMV-specific immune response, and the risk of late CMV disease. 9th International Symposium on Infections in the Immunocompromised Host, Assisi, Italy, 1996. 16 Gor D, Sabin C, Prentice HG et al. Longitudinal fluctuations in cytomegalovirus load in bone marrow transplant patients: relationship between peak virus load, donor/recipient serostatus, acute GVHD and CMV disease. Bone Marrow Transplant 1998; 21: 597–605. 17 Machado CM, Menezes RX, Macedo MC et al. Extended antigenemia surveillance and late cytomegalovirus infection after allogeneic BMT. Bone Marrow Transplant 2001; 28: 1053–1059. 18 Junghanss C, Boeckh M, Carter RA et al. Incidence and outcome of cytomegalovirus infections following nonmyeloablative compared with myeloablative allogeneic stem cell transplantation, a matched control study. Blood 2002; 99: 1978–1985. 19 Meijer E, Dekker AW, Verdonck LF. Influence of antithymocyte globulin dose on outcome in cytomegalovirus-seropositive recipients of partially T cell-depleted stem cell grafts from matched-unrelated donors. Br J Haematol 2003; 121: 473–476.

Bone Marrow Transplantation