GCF Chan, SY Ha, JSM Peiris, DCK Chiu, WL Lim, YL Lau

Abstract

Varicella infection in paediatric cancer patients could result in significant morbidity and mortality. We reviewed our patients cohort from January 89 to June 96 and found 5.6% (10 of 180) developed chickenpox. Nine had mild clinical disease and responded to acyclovir treatment and one died of disseminated varicella. Majority of the infected patients were receiving relatively mild chemotherapy and seven acquired the disease via community exposure. Interruption of the chemotherapy schedule was a concern with a median delay of 3 weeks (range 7 to 60 days). Another 7 patients developed zoster and all recovered with acyclovir. The median delay of chemotherapy was 7 days (range 5 to 9 days). With the introduction of varicella vaccine, an elective vaccination regimen has been developed for children with cancer. This practice is cost-effective and will minimize the use of VZJG and acyclovir.

Introduction

Varicella infection in immunocompromised children has been associated with significant morbidity and mortality.¹ After the introduction of acyclovir as treatment and Varicella Zoster immunoglobulin (VZIG) as prophylaxis for those with early exposure, varicella infection appears to be less severe in this group of patient.² But even with acyclovir, fatal treatment failures have been reported.³ An additional concern is the frequent interrupted chemotherapy schedule during the varicella infection period which may predispose to recurrence of the malignant disease. The live attenuated varicella vaccine was licensed for general use in Hong Kong in August 1996. We would like to review our experience and examine the role of varicella vaccine in paediatric oncology patients.

Methods

From January 1989 to June 1996, 180 children (<15-yr-old) with various malignancy were diagnosed and treated in our unit. The medical records of those who developed chickenpox and herpes zoster were retrospectively reviewed. Patients with contact history of varicella infection and received VZIG but with no evidence of disease were excluded. The diagnosis of chickenpox and herpes zoster were by characteristic clinical features. If available, the condition was confirmed by virology study. The virology study included electron-microscopy (EM) followed by immunofluroscent study (if EM positive), or by viral culture of the vesicular fluid obtained from the patients. Previous varicella exposure was documented by history and serology study. The serology would be done by complement fixation (CF) test first. If CF showed negative result, ELISA test would be performed to confirm the serology status.

Recovery time was counted from the day of onset of vesicles up to the time when there was no more new lesions with resolution of old lesions. For community acquired cases, transmission history was obtained from the interview with the patients' parents. For hospital acquired cases, a presumptive incubation period of 21 days (or 28 days if patients received VZIG) would be used to establish the contact history with an index case. Patient's primary disease and treatment status, lymphocytes and neutrophils count were included in the analysis.

Results

Ten of our 180 (5.6%) paediatric oncology patients developed chickenpox during this 7.5 years period. Their median age was 6 years (range 1-12 years) and 5 of them were male. Eight of 10 patients suffered from acute leukaemia [7 lymphoblastic (ALL) and 1 myeloid (AML)]. The other 2 patients had neuroblastoma and hepatoblastoma.

Chickenpox did not preferentially occur during the intensive treatment period; 5 of the 7 ALL patients in fact developed chickenpox during their maintenance therapy period. One patient with hepatoblastoma developed chickenpox just prior to his next course of chemotherapy. Only 4 of these 10 patients had a lymphocyte count less than 0.5x10⁹/l while they developed chickenpox.

Seven of the 10 patients with chickenpox infection acquired the infection via community exposure and none received VZJG due to absence of contact history. The other 3 patients acquired their infection in the hospital but due to no previous identifiable index case (2 patients), or with positive anti-varicella antibody (recurrent chickenpox infection), they also did not receive VZJG.

There was a case of recurrent chickenpox infection. The patient is an Indonesian boy with ALL who had a positive history of chickenpox 2 years prior to the onset of his malignancy. Chickenpox scar could be identified from his trunk and pre-therapy serology by ELISA clearly demonstrated an elevated anti-varicella titer. He developed chickenpox during his interim maintenance therapy period (UKALL XI intermediate risk protocol, status-post third high dose methotrexate for CNS prophylaxis). Vesicles came out from his scalp, trunk and 4 limbs in crops and were crossing midline. His recurrent chickenpox status was confirmed by EM and immunofluroscent study of his vesicular fluid. Concerning his transmission history, due to his foreigner status with no local residence for discharge, he had never been out of our ward during this early treatment period. An index case with chickenpox who was admitted to the hospital 19 days before the onset of his disease was identified. Since he had positive history of chickenpox infection with a positive anti-varicella antibody, VZIG was not given. There was no direct contact between these 2 patients and they resided in 2 separate isolation rooms. His recurrent chickenpox infection was uncomplicated and he responded to acyclovir therapy well.

Concerning the treatment and outcome, while none of the 10 patients received VZIG prophylaxis, all received acyclovir therapy. Nine of them had a mild clinical course with no visceral involvement. But one patient with relapsed ALL status-post reinduction therapy developed disseminated varicella infection. She was severely neutropenic and lymphopenic. Severe visceral involvement including gastrointestinal haemorrhage and liver derangement was noted. Despite acyclovir therapy, she succumbed to the infection 15 days after the onset of the chickenpox infection.

Delay in the chemotherapy schedule due to the infection was common. The median delay was 3 weeks and ranged from 1 week to as long as 2 months in one patient.

Within the same time period, 35 patients was diagnosed to have herpes zoster in our department and 7 of them were oncology patients. Five of these 7 patients had ALL and the other 2 had brain tumour (n=1) and neuroblastoma (n=1). Their median age was 6 years (range 3 to 17 years). Five of them were male.

Interestingly, herpes zoster was also more commonly found in our patients while they were not on intensive chemotherapy. Six episodes of herpes zoster were found in the 5 ALL patients including a recurrent attack in one infant ALL patient. All ALL patients were either receiving maintenance therapy or shortly off therapy when they developed herpes zoster. One solid tumour patient (neuroblastoma) developed zoster while she was off therapy. All of them were treated with acyclovir and none developed complication after the zoster attack. Similar to chickenpox, herpes zoster also commonly delayed the chemotherapy schedule but at a much shorter duration. The median delay was one week and ranged from 5 to 9 days.

Discussions

Up to early 1980, varicella infection in immunocompromised children still carried a very significant mortality and morbidity. In a review by Morgan et al, of 600 children who received immunosuppressive therapy, 3 1(5%) developed chickenpox infection and 15 of them experienced life threatening complications with visceral involvement.¹ Towards 1990, the severity of chickenpox infection seems to be less intense. In a study by the Children Cancer Group, of 472 patients with ALL, 44 (9%) developed chickenpox but only one had visceral involvement.² Our review agreed with this recent report. Severe complication was uncommon in our patients cohort. Effective treatment with acyclovir is believed to be a significant factor in reducing the severity of infection.^3,4 Despite the improvement in the outcome of chickenpox infection in this group of patients, the incidence of chickenpox infection remains relatively constant over the past 2 decades. And even with acyclovir treatment, fatal disseminated varicella infection though uncommon, does occur.

Chickenpox was more commonly found in our patients receiving maintenance chemotherapy, a relatively mild form of therapy for ALL, or shortly after chemotherapy was stopped. This paradox can be explained by the fact that patients receiving more intensive chemotherapy are usually strictly isolated hence the chance of contacting infectious host is low. In contrast, maintenance therapy for ALL are basically out-patient based and the patients are encouraged to resume their ordinary activity including schooling. The chance of exposure to varicella in this setting is definitely higher. That was also reflected in our review that majority of the patients acquired their chickenpox in the community rather than in the hospital. On the other hand, our only patient with fatal varicella infection was infected during her reinduction therapy period (which is a form of intensive treatment). So even chickenpox is rare in patients receiving intensive chemotherapy, when it happens, the outcome can be grave.

Although with the current management regimen, chickenpox in immunocompromised children is a less fatal illness, the interruption of the chemotherapy schedule which may predispose to cancer treatment failure is a concern. In addition, the high cost of treatment including hospitalization requires proper attention. If we include the days lost from school, parental loss of work, and possible VZIG prophylaxis for other exposed patients then the cost will even be higher. In the cost effective study by Buda et al.,² the average medical cost per susceptible patient with chickenpox was US$1933 (HK$15077) without counting the much higher parental costs.

Patient's previous immune status to varicella, disease epidemics in the general population, and the practice of VZIG prophylaxis are important determinants if we have to calculate the cost benefit ratio for this group of patients. The number of susceptible hosts to chickenpox varies according to different age group. Based on our local seroepidemiology data, around 17% of our paediatric oncology patients were susceptible to varicella according to the age distribution.⁵ In Hong Kong, an average of 176 new cases of paediatric (<15 yrs old) cancer was diagnosed each year.⁶ That means approximately 30 of them would be susceptible to chickenpox infection. The incidence rate of chickenpox is 8.3 to 9.1% per year in children 1 to 9 years of age.⁷ With this estimation, 3 paediatric oncology patients in Hong Kong may develop chickenpox each year. Around HK$45231 (US$5800) would be spent based on Buda's calculation. The number of susceptible patients with history of contact and requiring VZJG prophylaxis will even be higher. VZIG is expensive (HK $1180 or US$150 per 5 ml) and its efficacy is around 75% only.² In addition, history of contact is not always reliable. As our review illustrated, VZIG was not given in most of our patients developed chickenpox because of absence of contact history.

Currently, another option seems feasible and may be more cost effective. Elective live attenuated varicella vaccination for immunocompromised patients has been proposed. Based on the study on leukaemic children, cessation of chemotherapy for 1 week before and after vaccination while the patient's lymphocyte count is > 500/mm would be effective for them to mount immune response to the vaccine.⁸ Varicella vaccination had no deleterious effect on the outcome of the leukaemia.⁹ On the other hand, it has also been demonstrated that chemotherapy, except prednisolone, can be continued during the immediate pre and post-vaccine period so interruption of the therapy can be minimized. Based on the unit cost of vaccine is HK$460 (US$60), the estimated cost for vaccination of the susceptible paediatric oncology patient per year is HK$13800 (US$1800) for the whole Hong Kong. It is much cheaper than the cost of treating infected patients. And this measure will also drastically decrease, if not eliminate our expenses in using VZIG as prophylaxis.

In summary, though varicella infection is uncommon in our children with cancer, it can be fatal and requires expensive hospital care once developed. Majority of our patients acquired their infection via community contact and the sources could hardly be identified, early intervention with VZIG prophylaxis was not possible. Early treatment with acyclovir could decrease the severity of the infection. A significant delay in the schedule of chemotherapy was noted in our patients and the impact on their primary disease control remains unknown. Varicella vaccine may decrease the incidence or severity of varicella infection and minimize the interruption of chemotherapy schedule. And this practice seems to be cheaper than either early intervention or treatment of varicella infection. Future study is needed to evaluate the cost effectiveness of elective varicella vaccination for paediatric oncology patients and their susceptible siblings.

References

1. Morgan ER, Smalley LA. Varicella in immunocompromised children. Am J Dis Child 1983;137:883-5.

2. Buda k, Tubergen DG, Levin MJ. The frequency and consequences of varicella exposure and varicella infection in children receiving maintenance therapy for acute lymphoblastic leukaemia. J Pediatr Hematol Oncol 1996;18(2): 106-12.

3. Prober CG, Kirk LE, Keeney RE. Acyclovir therapy of chickenpox in immunosuppressed children-A collaborative study. J Pediatr 1982;101:622-5.

4. Meszner Z, Nyerges G, Bell AR. Oral acyclovir to prevent dissemination of varicella in immunocompromised children. J Infect 1993;26:9-15.

5. Kangro HO, Osman HK, Lau YL, Heath RB, Yeung CY, Ng MH. Seroprevalence of antibodies to Human Herpesviruses in England and Hong Kong. J Med Virol 1994;43:91-6.

6. Unpublished data; Hong Kong Paediatric Haematology & Oncology Study Group: census of childhood cancer from 1982 to 1991.

7. Weibel RE, Neff BJ, Kuter BJ, Guess HA, et al. Live attenuated varicella virus vaccine: Efficacy trial in healthy children. N Engl J Med 1984;310(22):1409-15.

8. Gershon AA, Steinberg SP, Gelb L, Galasso G, et al. Live attenuated varicella vaccine efficacy for children with leukaemia in remission. JAMA 1984;252(3):355-62.

9. Gershon AA, Steinberg SP, Gelb L. Live attenuated varicella vaccine use in immunocompromised children and adults. Paediatrics 1986;78(suppl):757-62.

10. Arbeter AM, Granowetter L, Starr SE, Lange B, et al. Immunization of children with acute lymphoblastic leukaemia with attenuated varicella vaccine without suspension of chemotherapy. Paediatrics 1990;85:338-44.