Vidarabine Therapy for Virus-Associated Cystitis After Allogeneic Bone Marrow Transplantation

Vidarabine therapy for virus-associated cystitis after allogeneic bone marrow transplantation

M Kawakami, S Ueda, T Maeda, T Karasuno, H Teshima, A Hiraoka, H Nakamura, K Tanaka and T Masaoka

The Fifth Department of Internal Medicine and Department of Structure Biology, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan

Summary: so, against double-stranded DNA viruses including human adenovirus and rodent polyomavirus in vitro.9,10 Vidarabine We describe a method of diagnosing virus-associated has been reported to be effective for hemorrhagic cystitis cystitis after allogeneic bone marrow transplantation caused by adenovirus and polyomavirus in a few cases after (BMT) and treatment with vidarabine therapy. At 7–10 BMT.11,12 However, clinical efﬁcacy of this agent still days post-BMT when cystitis was suspected, we remains unclear for hemorrhagic cystitis and related con- observed urinary sediments by the Papanicolaou stain ditions caused by adenovirus or polyomavirus after BMT. to detect virus inclusion bodies. When positive, we In this report, we describe our experience of vidarabine examined urinary sediments by transmission electron therapy for these conditions after BMT. microscope and measured the diameter of viral particles to determine the families. This process needed only 4 days. Among 16 consecutive cases, adenovirus and poly- Patients and methods omavirus were each detected in three. Adenovirus caused hemorrhagic cystitis in two cases and cystitis Patients without macroscopic hematuria in one case. Polyoma- virus caused cystitis without macroscopic hematuria in Among 16 consecutive patients who received BMT one case. Polyomavirus was also detected in two cases between November 1994 and October 1995 in our institute, without any symptoms. Vidarabine (10 mg/kg/day i.v.) a total of six patients suffered from virus-associated was administered for 5 days as one course. Soon after hemorrhagic cystitis, virus-associated cystitis or viruria. one course of vidarabine, most symptoms subsided and Patient characteristics are shown in Table 1. Two virus inclusion bodies disappeared in all cases except patients received BMT from HLA-matched unrelated for one with severe hemorrhagic cystitis. From these donors and others from HLA-matched sibling donors. Four experiences, vidarabine reduces excretion of adenovirus patients had a transplant for chronic myeloid leukemia in and polyomavirus in the urine of BMT recipients and chronic phase, one for myelodysplastic syndrome improves clinical symptoms in some cases of cystitis (FAB:RA), and one for acute myelogenous leukemia associated with these viruses. (FAB:M2 in 1st CR). All patients were conditioned with Keywords: vidarabine; hemorrhagic cystitis; adenovirus; high-dose cyclophosphamide (120 mg/kg) and total body polyomavirus; BMT; transmission electron microscope irradiation (12 Gy). MCNU (150 mg/body) and splenic irradiation (8 Gy) were added for CML patients. To avoid acrolein-induced hemorrhagic cystitis, infusion of mesna, alkalinization of urine and forced diuresis were performed Virus-associated hemorrhagic cystitis is one of the intrac- during the administration of cyclophosphamide. No urinary table complications after allogeneic bone marrow transplan- symptoms or hematuria occurred during and immediately tation (BMT), and is commonly caused by adenovirus or after administration of cyclophosphamide. Cyclosporine A polyomavirus.1–7 (6 mg/kg p.o. from day Ϫ1) and methotrexate (10 mg i.v. Adenovirus and polyomavirus are double-stranded DNA on days 1 and 3) were used for GVHD prophylaxis in all viruses, the diameters of which are about 80 and 42 nm, six patients. respectively.8 We have reported the use of transmission electron microscopy (TEM) on urinary sediments for differ- ential diagnosis of hemorrhagic cystitis caused by these Diagnosis 8 viruses as opposed to other causes. Every 7 to 10 days after BMT or when cystitis was sus- Vidarabine has been reported to be active, though weakly pected from symptoms, we performed urine cultures for bacteria and fungi and observed urinary sediments by Papanicolaou stain to detect virus inclusion bodies in the Correspondence: Dr A Hiraoka, 3-Nakamichi, 1-chome, Higashinari-ku, Osaka 537, Osaka Medical Center for Cancer and Cardiovascular Dis- nuclei and/or cytoplasm. When inclusion bodies were eases, Japan detected, we examined viral particles in urinary sediments Received 18 October 1996; accepted 29 May 1997 by TEM and speculated the family of the virus by measur- Vidarabine for cystitis after BMT M Kawakami et al 486 Table 1 Characteristics of patients

Case No. (UPN) Age Sex Diagnosis Transplant type Conditioning regimen

1 (143) 25 F CML-CP R MCNU ϩ SI ϩ CY ϩ TBI 2 (142) 46 M CML-CP U MCNU ϩ SI ϩ CY ϩ TBI 3 (152) 36 F CML-CP U MCNU ϩ SI ϩ CY ϩ TBI 4 (137) 23 M AML (M2) 1st CR R CY ϩ TBI 5 (139) 44 M MDS (RA) U CY ϩ TBI 6 (141) 30 F CML-CP U MCNU ϩ SI ϩ CY ϩ TBI

R = HLA matched sibling donor; MCNU = ranimustine; CY = cyclophosphamide; U = HLA matched unrelated donor; SI = splenic irradiation; TBI = total body irradiation.

ing the diameter of particles, as we reported previously.8 teria and fungus were negative or when they were positive, Urine samples were centrifuged at 1800 r.p.m. for 5 min, and antibiotics and antifungal agents did not improve the and supernatants discarded. Buffy coats layered onto red symptoms even after bacteria and fungus had cleared. blood cells were fixed in Saccommano solution, smeared ‘Viruria’ was diagnosed when no urinary symptoms nor on the slide, and stained by the Papanicolaou method.13 For hematuria occurred but viral particles were detected in TEM, buffy coats were resuspended in 2% glutaraldehyde urinary sediments. at pH 7.4 and fixed for 45 min at 4°C. They were postfixed in 1% osmium tetroxide, dehydrated in ethanol and embed- Treatment ded in epon. Sections were cut with a glass knife on a LKB microtome (LKB, Bromma, Sweden) and stained with ura- Patients with urinary symptoms were treated with forced nyl acetate and lead nitrate. A minimum of 100 cells per alkaline diuresis. Analgesics and antibiotics were used as sample selected at random from two epon blocks was necessary. In cases with macroscopic hematuria we perfor- examined. Photographs were taken by Jeol electron micro- med continuous bladder irrigation with prostaglandin E1 scope (Jeol-100CX; Jeol, Tokyo, Japan).8 bladder instillation,15 in addition to forced alkaline diuresis. The diameters of adenovirus and polyomavirus particles All six cases were given vidarabine at a dose of 10 mg/kg are 80 and 42 nm, respectively. When adenovirus infection per day for 5 days as a 2–3-h intravenous infusion, as one was suspected from TEM examination, we performed viral course. Another course was given when necessary. culture to confirm the diagnosis. Polyomavirus is difficult to isolate by the conventional culture method. We previously reported a case of hemorrhagic cystitis after Results alloBMT in which TEM examination of urinary sediments revealed many viral particles of 42 nm diameter, and urine Diagnosis culture with human embryonal kidney (HEK) cells produced polyomavirus BK. Immunocytochemistry of urinary The diagnosis and clinical features of each case are summa- sediments using a polyclonal antibody against simian virus rized in Table 2. Adenovirus was detected in three cases 40 (SV40) virion raised in rabbits was found to react (cases 1–3) and polyomavirus in the remaining three (cases mainly with the major coat protein (VP1) of SV40 and was 4–6). In the former three cases, viral cultures of the urine confirmed to crossreact with both human polyomaviruses also produced adenovirus. Cases 1 and 2 were diagnosed BK and JC by immunofluorescence which detected many to have hemorrhagic cystitis, cases 3 and 4 cystitis, and VP1-positive cells with staining in the nuclei.14 In this cases 5 and 6 viruria. Urine cultures for bacteria and fungus study we considered viral particles of 42 nm diameter as were negative in all cases except case 2. In case 2, cultures polyomaviruses without viral culture and immunocytoch- of urine produced E. coli and urinalysis showed an increase emistry because HEK cells and the antibody against SV40 of leukocytes, indicating the co-incident presence of bac- virion were not routinely available.8 Figure 1 shows terial cystitis. As severe hematuria with blood clot forma- electron micrographs of these viral particles in urinary tion continued even after the administration of antibiotics, sediments. a diagnosis of virus-associated hemorrhagic cystitis was made. Definition Treatment with vidarabine In this report, virus-associated hemorrhagic cystitis and related conditions are defined as follows. ‘Virus-associated Case 1 (Figure 2): Micturition pain and macroscopic hem- hemorrhagic cystitis’ was diagnosed when viral particles aturia developed on days 26 and 28 after BMT, respect- were detected in urinary sediments and macroscopic hemat- ively. The diagnosis of adenovirus-associated hemorrhagic uria with urinary symptoms were present. ‘Virus-associated cystitis was made after TEM examination of urinary sedi- cystitis’ was diagnosed when viral particles were detected ments. Continuous bladder irrigation and urine alkaliniz- in urinary sediments and urinary symptoms developed, but ation, and bladder instillation of prostaglandin E1 at a dose no macroscopic hematuria occurred. The diagnosis of these of 500 ␮g per day for 7 days were instigated but the symp- two conditions was made only when urine cultures for bac- toms did not resolve. On day 34 vidarabine therapy was Vidarabine for cystitis after BMT M Kawakami et al 487

Figure 1 Electron micrographs of urinary sediments. Many viral particles are observed in the nucleus of the urotherial cell. (a) The diameter of viral particles is about 80 nm, corresponding to adenovirus. (b) The diameter of viral particles is about 42 nm, corresponding to polyomavirus. started. On the next day the pain improved. Virus inclusion per day for 5 days was undertaken. On day 268, fever bodies in urinary sediments and macroscopic hematuria dis- improved but the pain and macroscopic hematuria appeared on day 39. Microscopic hematuria disappeared on continued for about 20 days after the start of vidarabine. day 41. Microscopic hematuria ceased 10 days later.

Case 2 (Figure 3): About 250 days after BMT urinary fre- Case 3: On day 55 after BMT dysuria occurred and on day quency occurred and urinary sediments revealed virus 56 vidarabine therapy, intravenous hydration and alkaliniz- inclusion bodies in the nuclei. Oral hydration and alkalinization were started. On day 59 dysuria improved and on day ation were begun. On day 261 high-grade fever and mictur- 63 virus inclusion bodies disappeared. TEM examination of ition pain occurred. On day 262, gross hematuria with urinary sediments revealed adenovirus-associated cystitis. blood clots developed. TEM examination revealed adenovirus in urinary sediments. On day 263 vidarabine, anti- Case 4: On day 59 after BMT high-grade fever developed biotics, continuous bladder irrigation and alkalinization and continued even after administration of antibiotics. On were started. On day 265, virus inclusion bodies in urinary day 69, micturition pain and microscopic hematuria sediments disappeared but the symptoms did not improve. developed. TEM examination allowed the diagnosis of Bladder instillation of prostaglandin E1 at a dose of 500 ␮g polyomavirus-associated cystitis. On the same day vidarab- Vidarabine for cystitis after BMT M Kawakami et al 488 275 mg 250 mg CsA p.o. 140 mg div. p.o. 200 150 175 150

GVHD grade I

Continuous bladder irrigation BMT

PGE1 500 µ g × 7 days bladder instillation 24.2.95

vidarabine 10 mg/kg div × 5 days

hematuria

virus inclusion bodies (+) (–) (–)

TEM: corresponding to adenovirus viral culture: adenovirus WBC Plt (µµ I) (× 104/ I) 10000 40 WBC G-CSF

5000 20 Plt

day 0 10 20 30 40 50 60 Figure 2 Clinical course of a case (case 1) with adenovirus-associated hemorrhagic cystitis. Macroscopic hematuria improved immediately after vidarabine therapy.

Table 2 Diagnosis and clinical features at the onset of hemorrhagic cystitis and related conditions

Case No. Onset day after BMT Viral particles identiﬁed Virus isolation Clinical symptoms (UPN) by TEM Hematuria Pain on Fever micturition

1 (143) 26 adeno adeno macroscopic severe low-grade 2 (142) 261 adeno adeno macroscopic severe high-grade 3 (152) 55 adeno adeno — mild — 4 (137) 69 polyoma not done microscopic mild high-grade 5 (139) 42 polyoma not done — — — 6 (141) 64 polyoma not done — — low-grade

ine was started. On day 74 virus inclusion bodies in urinary nausea and general fatigue developed in case 6. Other sediments disappeared and fever improved. Ten days after adverse effects including myelosuppression did not occur. the start of vidarabine, pain and microscopic hematuria disappeared. On day 82, the viruria recurred and vidarabine was re-administered, and the virus inclusion bodies Discussion disappeared on day 87. To make a diagnosis of virus-associated hemorrhagic cys- Case 5 and 6: Polyomavirus viruria developed on days 42 titis and related conditions, we use Papanicolaou stain and and 64 after BMT, vidarabine was started on days 49 and TEM on urinary sediments taking 1 day for the Papanico- 82, and the virus inclusion bodies in urinary sediments dis- laou stain and about 4 days for the TEM. Virus isolation appeared on days 52 and 84, respectively. in cell cultures takes more time, and it is difﬁcult to isolate Regarding the adverse effects of vidarabine therapy, mild polyomavirus by the conventional culture method.8 Our Vidarabine for cystitis after BMT M Kawakami et al 489 CsA 175 mg p.o. 125

Mizoribin 100 mg

Continuous bladder irrigation

PGE1 500 µ g × 5 days bladder instillation

vidarabine 10 mg/kg div × 5 days

fever

hematuria

virus inclusion bodies (++) (–) (–) (–) (–)

TEM: corresponding to adenovirus viral culture: adenovirus WBC Plt 4 (/µ I) ( × 10 /µ I)

10000 40 WBC 5000 20 Plt

day 260 270 280 290

Figure 3 Clinical course of a severe case (case 2) with adenovirus-associated hemorrhagic cystitis. Symptoms did not improve enough by vidarabine therapy. methods are useful for early diagnosis of viruria. Enzyme- of vidarabine, but in case 2 severe symptoms, namely gross linked immunosorbent assay (ELISA), DNA hybridization hematuria accompanied by blood clot formation, severe and PCR methods are also used to identify virus in the pain that needed morphine, and high-grade fever, did not urine after BMT.5–7 These methods are quicker and cheaper improve immediately and hematuria continued for about 30 than TEM. Therefore, the combination of these techniques days after the start of vidarabine. In this case the start of with routine urinalysis by Papanicolaou stain may be an vidarabine therapy was thought to be too late. In two cases alternative to our method using TEM. diagnosed to have cystitis (cases 3 and 4) symptoms disap- In this study of 16 cases after BMT three cases (19%) peared 3 and 10 days after the start of vidarabine, respect- developed cystitis or hemorrhagic cystitis associated with ively and hemorrhagic cystitis did not develop. It was adenovirus, a rate which is higher compared with studies uncertain whether or not vidarabine was effective in these 1 in the USA. Shields et al recovered 22 adenovirus isolates two cases, because viral cystitis improves spontaneously from urine of 1051 BMT recipients, and eight of these suf- in some cases. Miyamura et al3 revealed that adenovirus- 2 fered from hemorrhagic cystitis. Ambinder et al cultured associated hemorrhagic cystitis tends to be of long duration, adenovirus type 11 in the urine from five of 502 BMT with a median of 42 days. Vidarabine may therefore have recipients, and four of these five were associated with 3 shortened the clinical course of case 3 (adenovirus-associa- hemorrhagic cystitis. Otherwise, Miyamura et al reported ted cystitis). In the two cases diagnosed to have polyomavi- that 12 of 50 Japanese BMT recipients developed ruria (cases 5 and 6) with no symptoms, cystitis did not hemorrhagic cystitis and eight of these 12 recipients develop although in case 6 viruria relapsed. 36–76% of excreted adenovirus type 11 in the urine at the onset of cases of polyomaviruria after BMT improve spontaneously cystitis. In Japan, adenovirus may be excreted in the urine without hemorrhagic cystitis.5 Therefore the significance of more often than in USA. vidarabine therapy for polyomaviruria without symptoms, Within a few days (2–7 days) after the start of including cases 5 and 6 in this study, is uncertain. vidarabine, virus inclusion bodies in urinary sediments disappeared in all six cases in this study. Vidarabine This study suggests that vidarabine reduces excretion of reduces excretion of these two viruses in the urine. adenovirus and polyomavirus in the urine after BMT and Clinical improvement of each case with vidarabine ther- improves clinical symptoms in a proportion of cases of cys- apy was as follows. Of two cases of adenovirus-associated titis and hemorrhagic cystitis. Also, this agent may prevent hemorrhagic cystitis (case 1 and 2) vidarabine was effective progression from viruria into hemorrhagic cystitis, but in one case (case 1), and not in the other (case 2). In case observation of many more cases is required. 1, hematuria disappeared immediately after administration Vidarabine for cystitis after BMT M Kawakami et al 490 Acknowledgements hybridization assay and by polymerase chain reaction: an approach to assess the relationship between BK viruria and This study was in part supported by the Grant-in-aid for Cancer hemorrhagic cystitis. Bone Marrow Transplant 1994; 14: Research, No. 7–3, The Ministry of Health and Welfare, Japan. 235–240. 8 Hiraoka A, Ishikawa J, Yamagami T et al. Hemorrhagic cystitis after bone marrow transplantation: importance of a thin sectioning technique on urinary sediments for diagnosis. Bone References Marrow Transplant 1991; 7: 107–111. 9 Schabel FM Jr. The antiviral activity of 9-␤-d-arabinofuranos- 1 Shields AF, Hackman RC, Fife KH et al. Adenovirus infec- yladenine (Ara-A). Chemotherapy 1968; 13: 321–338. tions in patients undergoing bone marrow transplantation. New 10 Shannon WM. Adenine arabinoside: antiviral activity in vitro. Engl J Med 1985; 146: 529–533. In: Pavan-Langston D, Buchman RA, Alford Jr CA, (eds). 2 Ambinder RF, Burns W, Forman M et al. Hemorrhagic cystitis Adenine Arabinoside: An Antiviral Agent. Raven Press: New associated with adenovirus infection in bone marrow trans- York, 1975, pp 1–43. plantation. Arch Intern Med 1985; 146: 1400–1401. 11 Kitabayashi A, Hirokawa M, Miura AB. Successful vidarabine 3 Miyamura K, Takeyama K, Kojima S et al. Hemorrhagic cys- therapy for adenovirus type 11-associated acute hemorrhagic titis associated with urinary excretion of adenovirus type 11 cystitis after bone marrow transplantation. Bone Marrow following allogeneic bone marrow transplantation. Bone Mar- Transplant 1994; 14: 853–854. row Transplant 1989; 4: 533–535. 12 Chapman C, Flower AJE, Durrant STS. The use of vidarabine 4 Appeley JF, Rice SJ, Bishop JA et al. Late-onset hemorrhagic in the treatment of human polyomavirus associated acute cystitis associated with urinary excretion of polyomavirus hemorrhagic cystitis. Bone Marrow Transplant 1991; 7: after bone marrow transplantation. Transplantation 1987; 43: 481–483. 108–112. 13 Papanicolaou GN. Atlas of Exfoliative Cytology. Harvard Uni- 5 Arthur RR, Shah KV, Baust SJ et al. Association of BK virus versity Press: Cambridge, MA, 1963, pp 3–6. with hemorrhagic cystitis in recipient of bone marrow trans- 14 Akura K, Hatakenaka M, Kawai K et al. Use of immunocyto- plants. New Engl J Med 1986; 43: 230–234. chemistry on urinary sedements for the rapid identification of 6 Arthur RR, Shah KV, Charache P, Saral R. BK and JC virus human polyomavirus infection – a case report. Acta Cytolog- infection in recipients of bone marrow transplants. J Infect Dis ica 1988; 32: 247–251. 1988; 158: 563–569. 15 Trigg ME, O’Reilly J, Rumelhart S et al. Prostaglandin E1 7 Azzi A, Fanci R, Bosi A et al. Monitoring of polyomavirus bladder instillations to control severe hemorrhagic cystitis. J BK viruria in bone marrow transplantation patients by DNA Urol 1990; 143: 92–94.