- Email: [email protected]
Herpes zoster infection after bone marrow transplantation in children
Herpes zoster infection after bone marrow transplantation in children Hirohide Kawasaki, MD, Jun T a k a y a m a , MD, a n d Mutsuro Ohira, MD From the Department of Pediatrics, Kansai Medical University,Osaka, Japan, and the Department of Pediatrics, National Cancer Center Hospital, Tokyo, Japan
Objective: To determine the frequency of, risk factors for, and clinical course of herpes zoster (HZ) after bone marrow transplantation (BMT) in children. Study design: A total of 107 children with hematologic malignancy or solid tumor who underwent allogeneic or autologous BMT were studied retrospectively. Results:.Of the 107 patients, HZ developed in 35 (33%) after BMT; 31 (89%) of these 35 patients had localized HZ. The median onset of infection was day 96 after BMT, and 89% of cases of HZ occurred before day 365 after BMT. HZ developed in 26 (58%) of 45 patients (13/21 [62%] allogeneic and 13/24 [54%] autologous patients) with hematologic malignancy; most of these patients had undergone total body irradiation. Of 33 patients with solid tumor, HZ developed in 9 (27%). All patients with HZ were treated with acyclovir, and no patients died of complications directly resulting from HZ. Conclusion: Herpes zoster occurred earlier after BMT than in adults, and it occurred frequently in children who had hematologic malignancy and/or had undergone total body irradiation. Prompt antiviral therapy reduced the mortality rate and significant morbidity associated with HZ. (J PEDIATR1996;128:353-6) Allogeneic and autologous bone marrow transplantation is associated with variable periods of severe neutropenia, and morbidity and death in patients after BMT are largely due to infectious complications. Herpes zoster is a common disease caused by reactivation of the varicella-zoster virus in immunosuppressed patients. 1 Some reports have described the occurrence of HZ infection in adults after BMT, and have demonstrated several risk factors for it, including age, 2 underlying disease,3 conditioning regimen including total body irradiation,4 and the presence of acute or chronic graft-versus-host disease. 5 However, there are few reports 6 concerning the occun'ence of HZ after BMT in children, and HZ after BMT in children is poorly documented. We reviewed the overall incidence and timing of HZ after BMT and examined the risk factors for HZ in children undergoing BMT.
Submitted for publication July 24, 1995; accepted Oct. 20, 1995. Reprint requests: Hirohide Kawaski, MD, Fumizonocho 10-15, Moriguchi, Osaka 570, Japan. Copyright © 1996 by Mosby-Year Book, Inc. 0022-3476/96/$5.00 + 0 9/20/70011
METHODS Patients. This study included all 107 children (60 boys) undergoing allogeneic BMT or ABMT at the National Cancer Center Hospital Department of Pediatrics, Tokyo, Japan, between September 1982 and June 1993. Case identification was retrospective, and information was collected by chart review. The mean age was 9.6 years (range, 1.0 to 18.0 ABMT BMT GVHD HZ TBI VZV
Autologousbone marrow transplantation Bone marrow transplantation Graft-versus-hostdisease Herpes zoster Total body irradiation Varicella-zostervirus
years). Serologic tests for VZV were performed with the enzyme-linked immunosorbent assay before BMT for 65 patients. A total of 59 patients had a hematologic malignancy: 23 had acute lymphoblastic leukemia, 22 acute myeloid leukemia, 8 chronic myelogenous leukemia, 4 non-Hodgkin lymphoma, and 2 other malignancies. Of these 59 children, 28 underwent allogeneic BMT and 31 underwent ABMT. Forty-eight patients had solid tumors: 11 had rhabdomyo-
353
354
Kawasaki, Takayama, and Ohira
The JournalofPediatrics March 1996
Table. Underlying diseases AIIo-BMT
Hematologicmalignancy Acute lymphoblasticleukemia Acute myeloid leukemia Chronic myelogonousleukemia Non-Hodgkin lymphoma Other hematologic malignancies Solid tumor Neuroblastoma Rhabdomyosarcoma Osteosarcoma Brain tumor Ewing sarcoma Retinoblastoma Others solid tumors
28
TOTAL
30
Auto-BMT
31 8 9 7 2 2
2
Total
59 15 13 1 2 0
46 2 0 0 0 0 0 0
48 7 11 7 5 4 4 8
77
107
Allo, Allogeneic; Auto, autologous.
sarcoma, 9 neuroblastoma, 7 osteosarcoma, 5 brain tumor, 4 Ewing sarcoma, 4 retinoblastoma, and 8 other tumors. Of these 48 children, 46 underwent ABMT; two children with neuroblastoma underwent allogeneic BMT. Of the 59 patients with hematologic malignancy, 43 underwent conditioning regimens, including TBI; six patients underwent non-TBI regimens. Of the 48 patients with solid tumors, 44 underwent non-TBI regimens and four had TBI (Table). Definitions of VZV infection. Varicella was defined as vesicular skin lesions with generalized onset and without dermatomal distribution, irrespective of previous VZV infection. Herpes zoster was defined as the presence of vesicular lesions beginning in a dermatomal distribution in a patient with a history of varicella, a positive resuk on a VZV enzyme-linked immunosorbent assay, or both. Statistical analysis. The incidence of HZ was determined by the Kaplan-Meier method with 95 % confidence intervals. The curves were compared by the log-rank test. Potential risk factors for HZ included age, gender, underlying disease, conditioning regimen, duration of post-BMT aplasia, presence of acute or chronic GVHD, and antibody indexes as determined by enzyme-linked immunosorbent assay. Comparisons between selected patient groups with infection were made with the univariate chi-square test and the two-sided Student t test.
RESULTS Herpes zoster. Of the 107 patients, 78 (72.9%) had a history of varicella, a positive VZV test result, or both, before BMT. Herpes zoster occurred in 35 (32.8%) of the 107 patients; all 35 had a history of varicella (44.9% of such patients) (Fig. 1). The patients with HZ included 17 boys and 18 girls with a mean age of 10.4 years (range, 2.3 to 17.8 years). The median onset of HZ was 96 days after BMT (range, 0 to 1021 days). HZ occurred in a total of 20 patients
(57.1%) and 31 patients (88.6%), respectively, within the first 100 days and 365 days after BMT. Of the 35 patients with varicella, 26 (74%) had a hematologic malignancy and 9 had solid tumors. Thirteen (37%) underwent allogeneic BMT and 22 (63%) underwent ABMT. Thirty-one patients (89%) had localized HZ. The locations of cutaneous infection were as follows: cranial dermatome, 6 (17%); cervical, 3 (9%); thoracic, 17 (49%); lumbar, 1 (3%); and sacral, 4 (11%). Three patients had disseminated cutaneous HZ based on dermatomal presentations, one patient had varicelliform reactivated latent zoster (cutaneous dissemination), and no patient had visceral dissemination. All patients were treated with acyclovir (15 to 30 mg/kg per day). Postherpetic neuralgia occurred in four patients (11%). However, none of the complications were either permanent or disabling. No death occurred in relation to HZ. Risk factors for HZ. There were significant differences in the incidence of HZ, depending on the underlying disease. Herpes zoster occurred in 26 children (58%) with hematologic malignancy, in comparison with 9 children (27%) with a solid tumor (p <0.01) (Fig. 2). The use of TBI in the conditioning regimen was associated with a significantly higher rate of HZ (p <0.01). At our institution, however, the choice of a conditioning regimen was based on the underlying disease. Fifty-three children with hematologic malignancy (96%) underwent conditioning regimens including TBI, and 44 children with a solid tumor (92%) received non-TBI regimens. The results of multivariate analysis indicated that there was no independent association between TBI and HZ. There was no significant difference in the frequency of HZ between patients undergoing allogeneic BMT (59%) and those undergoing ABMT (39%). Even among patients with hematologic malignancy, there was no significant difference in the frequency of HZ between those who underwent allo-
The Journal of Pediatrics Volume 128, Number 3
Kawasaki, Takayama, and Ohira
355
% 100"
90" N 80' -r "6 70' ID
g 60, --
varicella (+)
50'
:>
40' O
,
total
30' 2010" j'""
)3
=
6 g
1'2
24
1'8
"
:30
36
42
4'8
i
54
60
months after BMT
Fig. 1. Cumulative incidence of HZ among all children undergoing BMT was 32.8% (solid line), whereas that among 78 children with a history of varicella was 44.9% (dotted line).
% 100" 90" N 80" '-r~6 Q) 70O t--
Hematologic malignancy
o~ 60-g_ O
.=-- 50._> _N 40-
,,
I-"
|
P
E
Solid tumor
= 30O 2010,,,
0
; 1'2
1'8
,,
44
,t2
4'8
d4
months after BMT Fig, 9. Cumulative incidence of HZ among children with hematologic malignancy was 57.8% (solid line), whereas that
among children with solid tumor was 27.3% (dotted line). HZ was significantlymore frequent in children with hematologic malignancy than in those with solid tumor (p <0.01).
geneic BMT (62%) and those who underwent ABMT (54%). Eight children had varicella after an initial diagnosis of malignancy or during treatment, and underwent BMT within a comparatively short period after varicella (<150 days). Their cases of HZ occurred at a significantly higher rate (100%) and occurred earlier (range, 0 to 80 days; median, 40 days)
than did those in the other children. There were no differences in median pretransplantation liters of IgG antibody to VZV between patients who did and those who did not have HZ after BMT. Shnilarly, age, gender, and the presence of acute or chronic GVHD were each unassociated with an increase in the incidence of HZ after BMT,
356
Kawasaki, Takayama, and Ohira
DISCUSSION This study showed that herpes zoster occurred with a high level of frequency after BMT in children. In our study, HV developed in one third of children undergoing BMT, in contrast to 16.5% of children with leukemia who underwent only chemotherapy. 7 For adults who have undergone BMT, the reported incidence of HZ has ranged from 30% to 62%2-5; the reported time to onset was 5 months. 8 For patients with Hodgkin disease, a history of primary varicella or HZ after the initial diagnosis is a significant risk factor for the development of HZ early after ABMT. 9 In our population, HZ occurred significantly earlier in those with a history of primary varicella after the initial diagnosis of malignant disease than in other patients. Thus the early development of HZ in children was the result of the comparatively short interval between primary varicella infection and BMT. Reported risk factors for I-LZinclude patient age, 2 underlying disease, 3 conditioning regimen including TBI, 4 and acute and chronic GVHD. 5 In this study the only risk factor found to be significantly associated with the occurrence of HZ after BMT was underlying disease. In our hospital, however, most children with hematologic malignancy undergo a pre-BMT regimen, including TBI, and most children with solid tumor undergo a non-TBI regimen. A recent study showed that TBI affects the incidence of HZ after ABMTI°; the investigators suggested that the difference in incidence of VZV infection between patients undergoing regimens with TBI and those undergoing regimens without TBI might be due to differences in the degree of immunosuppression induced. Similarly, in our study, the non-TBI regimen for patients with solid tumors was significantly less immunosuppressive than the regimen with TBI. The median duration from BMT to engraftment (neutrophils >0.5 x 109/L) was 25.1 days for the TBI regimen and 13.8 days for the non-TBI regimen. However, a recent study of children undergoing peripheral blood stem-cell transplantation for hematologic malignancy showed that HZ developed in 44%, which is similar to that for our patients. H Their non-TBI conditioning regimen was less immunosuppressive and their engraftment was earlier than ours. These findings, taken together, suggest that the underlying hematologic mafignancy may profoundly affect the incidence of HZ. All of our patients received acyclovir at the time of the presumptive diagnosis of HZ. The good outcome that we obtained after HZ differs markedly from the 18% incidence of visceral dissemination and the 10% mortality rate in the Seattle series.5 These differences in findings may be explained by differences in the era of study and the impact of acyclovir treatment; only a minority of the Seattle patients were treated with acyclovir. Early recognition of HZ and prompt acyclovir therapy should prevent
The Journal of Pediatrics March 1996
significant morbidity and death in children undergoing BMT. 12 We conclude that children undergoing BMT are predisposed to the development of HZ at an earlier stage than are adults, and that hematologic malignancy is associated with a higher incidence of HZ than are solid tumors. The use of antiviral agents may be beneficial in reducing the morbidity and mortality rates associated with HZ; however, the role of such agents in prophylaxis for HZ remains to be determined. 13 REFERENCES
1. Nader S, Arvin AM. Varicella-zoster virus infections. In: Forman SJ, Blume KG, Thomas ED, eds. Bone marrow transplantations. Boston: Blackwell Scientific Publications, 1994:41228. 2. Atkinson K, Meyers JD, Storb R, Prentice RL, Thomas El). Varicella-zoster virus infection after marrow transplantation for aplastic anemia or leukemia. Transplantation 1980;29:4750. 3. Schuchter LM, Wingard JR, Piantadosi S, Bums WH, Santos GW, Saral R. Herpes zoster infection after autologous bone marrow transplantation. Blood 1989;74:1424-7. 4. Han CS, Miller W, Haake R, Weisdorf D. Varicella-zoster infection after bone marrow transplantation: incidence, risk factors and complications. Bone Marrow Transplant 1994; 13:27783. 5. Locksley RM, Floumoy N, Sullivan KM, Meyers JD. Infection with vaficella-zoster virus after marrow transplantation. J Infect Dis 1985;11:72-81. 6. Wacker P, Hartmann O, Benhamou E, Salloum E, Lemerle J. VariceUa-zoster virus infections after autologous bone marrow transplantation in children. Bone Marrow Transplant 1989; 4:191-4. 7. Novelli VM, Brunell PA, Geiser CF, Narkewics S, Frierson L. Herpes zoster in children with acute lymphocytic leukemia. Arch Pediatr Adolesc Med [Am J Dis Child] 1988;142:71-2. 8. Wilson A, Sharp M, Koropchak CM, Ting SF, Arvin AM. Subclinical varicella-zoster virus viremia, herpes zoster, and T-lymphocyte immunity to varicella-zoster viral antigens after bone marrow transplantation. J Infect Dis 1992;165:119-26. 9. Christiansen NP, Haake RJ, Hurd DD. Early herpes zoster infection in adult patients with Hodgkin's disease undergoing autologous bone marrow transplantation. Bone Marrow Transplant 1991;7:435-7. 10. Callum JL, Brandwein JM, Sutcliffe SB, Scott JG, Keating A. Influence of total body irradiation on infections after autologous bone marrow transplantation. Bone Marrow Transplant 1991;8:245-51. 11. Takaue Y, Okamoto Y, Kawano Y, et al. Regeneration of immunity and varicella-zoster virus infection after high-dose chemotherapy and peripheral blood stem-cell autografts in children. Bone Marrow Transplant 1994;14:219-23. 12. Meyers JD, Wade JC, Shepp DH, Newton B. Acyclovir treatment of varicella-zoster virus infection in the compromised host. Transplantation 1984;37:571-4. 13. Selby PJ, Powels RL, Eaton D, et al. The prophylactic role of intravenous and long-term oral acyclovir after allogenic bone marrow transplantation. Br J Cancer 1989;59:434-8.
Objective: To determine the frequency of, risk factors for, and clinical course of herpes zoster (HZ) after bone marrow transplantation (BMT) in children. Study design: A total of 107 children with hematologic malignancy or solid tumor who underwent allogeneic or autologous BMT were studied retrospectively. Results:.Of the 107 patients, HZ developed in 35 (33%) after BMT; 31 (89%) of these 35 patients had localized HZ. The median onset of infection was day 96 after BMT, and 89% of cases of HZ occurred before day 365 after BMT. HZ developed in 26 (58%) of 45 patients (13/21 [62%] allogeneic and 13/24 [54%] autologous patients) with hematologic malignancy; most of these patients had undergone total body irradiation. Of 33 patients with solid tumor, HZ developed in 9 (27%). All patients with HZ were treated with acyclovir, and no patients died of complications directly resulting from HZ. Conclusion: Herpes zoster occurred earlier after BMT than in adults, and it occurred frequently in children who had hematologic malignancy and/or had undergone total body irradiation. Prompt antiviral therapy reduced the mortality rate and significant morbidity associated with HZ. (J PEDIATR1996;128:353-6) Allogeneic and autologous bone marrow transplantation is associated with variable periods of severe neutropenia, and morbidity and death in patients after BMT are largely due to infectious complications. Herpes zoster is a common disease caused by reactivation of the varicella-zoster virus in immunosuppressed patients. 1 Some reports have described the occurrence of HZ infection in adults after BMT, and have demonstrated several risk factors for it, including age, 2 underlying disease,3 conditioning regimen including total body irradiation,4 and the presence of acute or chronic graft-versus-host disease. 5 However, there are few reports 6 concerning the occun'ence of HZ after BMT in children, and HZ after BMT in children is poorly documented. We reviewed the overall incidence and timing of HZ after BMT and examined the risk factors for HZ in children undergoing BMT.
Submitted for publication July 24, 1995; accepted Oct. 20, 1995. Reprint requests: Hirohide Kawaski, MD, Fumizonocho 10-15, Moriguchi, Osaka 570, Japan. Copyright © 1996 by Mosby-Year Book, Inc. 0022-3476/96/$5.00 + 0 9/20/70011
METHODS Patients. This study included all 107 children (60 boys) undergoing allogeneic BMT or ABMT at the National Cancer Center Hospital Department of Pediatrics, Tokyo, Japan, between September 1982 and June 1993. Case identification was retrospective, and information was collected by chart review. The mean age was 9.6 years (range, 1.0 to 18.0 ABMT BMT GVHD HZ TBI VZV
Autologousbone marrow transplantation Bone marrow transplantation Graft-versus-hostdisease Herpes zoster Total body irradiation Varicella-zostervirus
years). Serologic tests for VZV were performed with the enzyme-linked immunosorbent assay before BMT for 65 patients. A total of 59 patients had a hematologic malignancy: 23 had acute lymphoblastic leukemia, 22 acute myeloid leukemia, 8 chronic myelogenous leukemia, 4 non-Hodgkin lymphoma, and 2 other malignancies. Of these 59 children, 28 underwent allogeneic BMT and 31 underwent ABMT. Forty-eight patients had solid tumors: 11 had rhabdomyo-
353
354
Kawasaki, Takayama, and Ohira
The JournalofPediatrics March 1996
Table. Underlying diseases AIIo-BMT
Hematologicmalignancy Acute lymphoblasticleukemia Acute myeloid leukemia Chronic myelogonousleukemia Non-Hodgkin lymphoma Other hematologic malignancies Solid tumor Neuroblastoma Rhabdomyosarcoma Osteosarcoma Brain tumor Ewing sarcoma Retinoblastoma Others solid tumors
28
TOTAL
30
Auto-BMT
31 8 9 7 2 2
2
Total
59 15 13 1 2 0
46 2 0 0 0 0 0 0
48 7 11 7 5 4 4 8
77
107
Allo, Allogeneic; Auto, autologous.
sarcoma, 9 neuroblastoma, 7 osteosarcoma, 5 brain tumor, 4 Ewing sarcoma, 4 retinoblastoma, and 8 other tumors. Of these 48 children, 46 underwent ABMT; two children with neuroblastoma underwent allogeneic BMT. Of the 59 patients with hematologic malignancy, 43 underwent conditioning regimens, including TBI; six patients underwent non-TBI regimens. Of the 48 patients with solid tumors, 44 underwent non-TBI regimens and four had TBI (Table). Definitions of VZV infection. Varicella was defined as vesicular skin lesions with generalized onset and without dermatomal distribution, irrespective of previous VZV infection. Herpes zoster was defined as the presence of vesicular lesions beginning in a dermatomal distribution in a patient with a history of varicella, a positive resuk on a VZV enzyme-linked immunosorbent assay, or both. Statistical analysis. The incidence of HZ was determined by the Kaplan-Meier method with 95 % confidence intervals. The curves were compared by the log-rank test. Potential risk factors for HZ included age, gender, underlying disease, conditioning regimen, duration of post-BMT aplasia, presence of acute or chronic GVHD, and antibody indexes as determined by enzyme-linked immunosorbent assay. Comparisons between selected patient groups with infection were made with the univariate chi-square test and the two-sided Student t test.
RESULTS Herpes zoster. Of the 107 patients, 78 (72.9%) had a history of varicella, a positive VZV test result, or both, before BMT. Herpes zoster occurred in 35 (32.8%) of the 107 patients; all 35 had a history of varicella (44.9% of such patients) (Fig. 1). The patients with HZ included 17 boys and 18 girls with a mean age of 10.4 years (range, 2.3 to 17.8 years). The median onset of HZ was 96 days after BMT (range, 0 to 1021 days). HZ occurred in a total of 20 patients
(57.1%) and 31 patients (88.6%), respectively, within the first 100 days and 365 days after BMT. Of the 35 patients with varicella, 26 (74%) had a hematologic malignancy and 9 had solid tumors. Thirteen (37%) underwent allogeneic BMT and 22 (63%) underwent ABMT. Thirty-one patients (89%) had localized HZ. The locations of cutaneous infection were as follows: cranial dermatome, 6 (17%); cervical, 3 (9%); thoracic, 17 (49%); lumbar, 1 (3%); and sacral, 4 (11%). Three patients had disseminated cutaneous HZ based on dermatomal presentations, one patient had varicelliform reactivated latent zoster (cutaneous dissemination), and no patient had visceral dissemination. All patients were treated with acyclovir (15 to 30 mg/kg per day). Postherpetic neuralgia occurred in four patients (11%). However, none of the complications were either permanent or disabling. No death occurred in relation to HZ. Risk factors for HZ. There were significant differences in the incidence of HZ, depending on the underlying disease. Herpes zoster occurred in 26 children (58%) with hematologic malignancy, in comparison with 9 children (27%) with a solid tumor (p <0.01) (Fig. 2). The use of TBI in the conditioning regimen was associated with a significantly higher rate of HZ (p <0.01). At our institution, however, the choice of a conditioning regimen was based on the underlying disease. Fifty-three children with hematologic malignancy (96%) underwent conditioning regimens including TBI, and 44 children with a solid tumor (92%) received non-TBI regimens. The results of multivariate analysis indicated that there was no independent association between TBI and HZ. There was no significant difference in the frequency of HZ between patients undergoing allogeneic BMT (59%) and those undergoing ABMT (39%). Even among patients with hematologic malignancy, there was no significant difference in the frequency of HZ between those who underwent allo-
The Journal of Pediatrics Volume 128, Number 3
Kawasaki, Takayama, and Ohira
355
% 100"
90" N 80' -r "6 70' ID
g 60, --
varicella (+)
50'
:>
40' O
,
total
30' 2010" j'""
)3
=
6 g
1'2
24
1'8
"
:30
36
42
4'8
i
54
60
months after BMT
Fig. 1. Cumulative incidence of HZ among all children undergoing BMT was 32.8% (solid line), whereas that among 78 children with a history of varicella was 44.9% (dotted line).
% 100" 90" N 80" '-r~6 Q) 70O t--
Hematologic malignancy
o~ 60-g_ O
.=-- 50._> _N 40-
,,
I-"
|
P
E
Solid tumor
= 30O 2010,,,
0
; 1'2
1'8
,,
44
,t2
4'8
d4
months after BMT Fig, 9. Cumulative incidence of HZ among children with hematologic malignancy was 57.8% (solid line), whereas that
among children with solid tumor was 27.3% (dotted line). HZ was significantlymore frequent in children with hematologic malignancy than in those with solid tumor (p <0.01).
geneic BMT (62%) and those who underwent ABMT (54%). Eight children had varicella after an initial diagnosis of malignancy or during treatment, and underwent BMT within a comparatively short period after varicella (<150 days). Their cases of HZ occurred at a significantly higher rate (100%) and occurred earlier (range, 0 to 80 days; median, 40 days)
than did those in the other children. There were no differences in median pretransplantation liters of IgG antibody to VZV between patients who did and those who did not have HZ after BMT. Shnilarly, age, gender, and the presence of acute or chronic GVHD were each unassociated with an increase in the incidence of HZ after BMT,
356
Kawasaki, Takayama, and Ohira
DISCUSSION This study showed that herpes zoster occurred with a high level of frequency after BMT in children. In our study, HV developed in one third of children undergoing BMT, in contrast to 16.5% of children with leukemia who underwent only chemotherapy. 7 For adults who have undergone BMT, the reported incidence of HZ has ranged from 30% to 62%2-5; the reported time to onset was 5 months. 8 For patients with Hodgkin disease, a history of primary varicella or HZ after the initial diagnosis is a significant risk factor for the development of HZ early after ABMT. 9 In our population, HZ occurred significantly earlier in those with a history of primary varicella after the initial diagnosis of malignant disease than in other patients. Thus the early development of HZ in children was the result of the comparatively short interval between primary varicella infection and BMT. Reported risk factors for I-LZinclude patient age, 2 underlying disease, 3 conditioning regimen including TBI, 4 and acute and chronic GVHD. 5 In this study the only risk factor found to be significantly associated with the occurrence of HZ after BMT was underlying disease. In our hospital, however, most children with hematologic malignancy undergo a pre-BMT regimen, including TBI, and most children with solid tumor undergo a non-TBI regimen. A recent study showed that TBI affects the incidence of HZ after ABMTI°; the investigators suggested that the difference in incidence of VZV infection between patients undergoing regimens with TBI and those undergoing regimens without TBI might be due to differences in the degree of immunosuppression induced. Similarly, in our study, the non-TBI regimen for patients with solid tumors was significantly less immunosuppressive than the regimen with TBI. The median duration from BMT to engraftment (neutrophils >0.5 x 109/L) was 25.1 days for the TBI regimen and 13.8 days for the non-TBI regimen. However, a recent study of children undergoing peripheral blood stem-cell transplantation for hematologic malignancy showed that HZ developed in 44%, which is similar to that for our patients. H Their non-TBI conditioning regimen was less immunosuppressive and their engraftment was earlier than ours. These findings, taken together, suggest that the underlying hematologic mafignancy may profoundly affect the incidence of HZ. All of our patients received acyclovir at the time of the presumptive diagnosis of HZ. The good outcome that we obtained after HZ differs markedly from the 18% incidence of visceral dissemination and the 10% mortality rate in the Seattle series.5 These differences in findings may be explained by differences in the era of study and the impact of acyclovir treatment; only a minority of the Seattle patients were treated with acyclovir. Early recognition of HZ and prompt acyclovir therapy should prevent
The Journal of Pediatrics March 1996
significant morbidity and death in children undergoing BMT. 12 We conclude that children undergoing BMT are predisposed to the development of HZ at an earlier stage than are adults, and that hematologic malignancy is associated with a higher incidence of HZ than are solid tumors. The use of antiviral agents may be beneficial in reducing the morbidity and mortality rates associated with HZ; however, the role of such agents in prophylaxis for HZ remains to be determined. 13 REFERENCES
1. Nader S, Arvin AM. Varicella-zoster virus infections. In: Forman SJ, Blume KG, Thomas ED, eds. Bone marrow transplantations. Boston: Blackwell Scientific Publications, 1994:41228. 2. Atkinson K, Meyers JD, Storb R, Prentice RL, Thomas El). Varicella-zoster virus infection after marrow transplantation for aplastic anemia or leukemia. Transplantation 1980;29:4750. 3. Schuchter LM, Wingard JR, Piantadosi S, Bums WH, Santos GW, Saral R. Herpes zoster infection after autologous bone marrow transplantation. Blood 1989;74:1424-7. 4. Han CS, Miller W, Haake R, Weisdorf D. Varicella-zoster infection after bone marrow transplantation: incidence, risk factors and complications. Bone Marrow Transplant 1994; 13:27783. 5. Locksley RM, Floumoy N, Sullivan KM, Meyers JD. Infection with vaficella-zoster virus after marrow transplantation. J Infect Dis 1985;11:72-81. 6. Wacker P, Hartmann O, Benhamou E, Salloum E, Lemerle J. VariceUa-zoster virus infections after autologous bone marrow transplantation in children. Bone Marrow Transplant 1989; 4:191-4. 7. Novelli VM, Brunell PA, Geiser CF, Narkewics S, Frierson L. Herpes zoster in children with acute lymphocytic leukemia. Arch Pediatr Adolesc Med [Am J Dis Child] 1988;142:71-2. 8. Wilson A, Sharp M, Koropchak CM, Ting SF, Arvin AM. Subclinical varicella-zoster virus viremia, herpes zoster, and T-lymphocyte immunity to varicella-zoster viral antigens after bone marrow transplantation. J Infect Dis 1992;165:119-26. 9. Christiansen NP, Haake RJ, Hurd DD. Early herpes zoster infection in adult patients with Hodgkin's disease undergoing autologous bone marrow transplantation. Bone Marrow Transplant 1991;7:435-7. 10. Callum JL, Brandwein JM, Sutcliffe SB, Scott JG, Keating A. Influence of total body irradiation on infections after autologous bone marrow transplantation. Bone Marrow Transplant 1991;8:245-51. 11. Takaue Y, Okamoto Y, Kawano Y, et al. Regeneration of immunity and varicella-zoster virus infection after high-dose chemotherapy and peripheral blood stem-cell autografts in children. Bone Marrow Transplant 1994;14:219-23. 12. Meyers JD, Wade JC, Shepp DH, Newton B. Acyclovir treatment of varicella-zoster virus infection in the compromised host. Transplantation 1984;37:571-4. 13. Selby PJ, Powels RL, Eaton D, et al. The prophylactic role of intravenous and long-term oral acyclovir after allogenic bone marrow transplantation. Br J Cancer 1989;59:434-8.