Late-onset hemorrhagic cystitis after hematopoietic stem cell ... - Nature

Bone Marrow Transplantation, (1998) 22, 995–998  1998 Stockton Press All rights reserved 0268–3369/98 $12.00 http://www.stockton-press.co.uk/bmt

Late-onset hemorrhagic cystitis after hematopoietic stem cell transplantation in children M Kondo1,2, S Kojima1, K Kato1 and T Matsuyama1 1

Division of Hematology/Oncology, Children’s Medical Center, Japanese Red Cross Nagoya First Hospital; and 2Department of Pediatrics, Nagoya University School of Medicine, Nagoya, Japan

Summary: We analyzed the incidence, complications, and risk factors for late-onset hemorrhagic cystitis (HC) in 256 children undergoing hematopoietic stem cell transplantation (HSCT). Twenty-six recipients (10.2%) developed late-onset HC between 3 and 270 days (median, 33 days) after HSCT. In most patients, the severity of HC was mild to moderate, and spontaneous resolution occurred. Three children developed bladder tamponade, and one required suprapubic cystotomy. Four children died in the early post-transplant period without resolution of HC, but HC was not the direct cause of death in any patient. Twenty-two patients recovered within 6–86 days (median, 16 days) of onset. Three predisposing factors were identified for development of late-onset HC by multivariate analysis: allogeneic HSCT, older age (⭓7 years), and busulphan for pretransplant conditioning were significantly associated with late-onset HC (P ⴝ 0.022, P ⴝ 0.044 and P ⴝ 0.036, respectively). Excretion of adenovirus type 11 was demonstrated in six of 22 patients at the onset of cystitis. We suspect that reactivation of virus may be a major pathogenic factor in lateonset HC, but several clinical factors are also associated. Keywords: hemorrhagic cystitis; late-onset; hematopoietic stem cell transplantation; children

Hemorrhagic cystitis (HC) is a major complication in patients undergoing hematopoietic stem cell transplantation (HSCT).1–3 Although HC is not usually life-threatening, it causes significant pain and occasional bladder tamponade with urinary obstruction. In addition, a small percentage of patients require surgical intervention.4 High-dose cyclophosphamide (CY) as part of the pretransplant conditioning regimen in HSCT is well known as one of the major causes of HC.5 CY-induced HC usually occurs during or immediately after CY administration,6 but HC can also occur at later periods post-HSCT. Several factors, including viral reactivation,7–13 graftversus-host disease (GVHD),14,15 and chemotherapeutic agents,16,17 have been implicated as causes of HC. HowCorrespondence: Dr M Kondo, Department of Pediatrics, Nagoya University School of Medicine, 65, Tsurumai-cho, Showa-ku, Nagoya 466– 8550, Japan Received 9 April 1998; accepted 15 July 1998

ever, most reports included all cases developing HC, and few focused on late-onset HC alone. We followed children who underwent HSCT for malignant or non-malignant diseases, and examined the incidence, complications, and predisposing factors in late-onset HC. Patients and methods Two hundred and fifty-six children underwent HSCT for malignant or non-malignant diseases at the Japanese Red Cross Nagoya First Hospital between April 1982 and March 1996. Patient characteristics are summarized in Table 1. One hundred and eighty-three patients received allogeneic HSCT and 73 received autologous HSCT. Allogeneic recipients received cyclosporin A (CyA) or methotrexate or both for prophylaxis of GVHD. Conditioning regimens varied according to disease and clinical status. In patients with malignant diseases, conditioning regimens included melphalan (MEL, 180–210 mg/m2) or CY (120– 200 mg/kg) combined with busulphan (BU, 4 mg/kg or 140 mg/m2 ⫻ 4 days) or total body irradiation (TBI, 10– 12 Gy) or both. Agents such as etoposide or cytosine arabinoside were added to conditioning regimens in high-risk patients. In patients with aplastic anemia, CY with or without total lymphoid irradiation or TBI was administered as conditioning, and in those with metabolic disease/ immunodeficiency, BU and CY with or without etoposide were administered. Since 1993, anti-thymocyte globulin (ATG) has usually been added to conditioning regimens in patients with non-malignant diseases and in those receiving a transplant from an unrelated donor. All patients treated with CY also received 2-mercapto-ethane-sodium sulphonate (mesna) at 120% of the CY dose in three divided injections. Bladder irrigation was not used, and no patient received vesical irradiation before HSCT. A diagnosis of HC was made when macroscopic hematuria or sustained (⬎7 days) microhematuria with clinical signs of cystitis were present. Late-onset HC was defined as occurring beyond 48 h of CY administration.3 Severity of HC was graded according to the following criteria:1 mild HC was defined as sustained microhematuria or macrohematuria without clot; moderate HC as macrohematuria with clot; severe HC as macrohematuria with clot and urinary obstruction. Urinalyses and bacterial urine cultures were performed twice a week for at least 3 months after HSCT. Viral isolation from urine specimens was performed by standard methods,18 when symptoms developed. Excretion of BK virus was not examined.

Late-onset hemorrhagic cystitis M Kondo et al

996

Table 1

Patient characteristics

No. of patients Gender Male/Female Median age (range) Diagnosis ALL AML CML MDS NHL/HD Solid tumors Aplastic anemia Metabolic disease/Immunodeficiency Type of HSCT Allogeneic HLA-identical sibling donor HLA-mismatched related donor HLA-matched unrelated donor Autologous Bone marrow Peripheral blood Conditioning regimen TBI ⫹ MEL ⫾ (BU or VP-16) TBI ⫹ CY ⫾ (BU or VP-16 or Ara-C) TLI ⫹ CY BU ⫹ MEL ⫾ others BU ⫹ CY ⫾ others Others GVHD prophylaxis CyA CyA ⫹ MTX MTX

256 137/119 7 years (0–19 years) 90 57 9 12 14 24 40 10 183 122 31 30 73 58 15 93 53 22 55 10 23 7 100 76

HSCT = hematopoietic stem cell transplantation; ALL = acute lymphoblastic leukemia; AML = acute myelogenous leukemia; CML = chronic myelogenous leukemia; MDS = myelodysplastic syndrome; NHL = nonHodgkin’s lymphoma; HD = Hodgkin’s disease; TBI = total body irradiation; MEL = melphalan; BU = busulphan; CY = cyclophosphamide; VP-16 = etoposide; Ara-C = cytosine arabinoside; TLI = total lymphoid irradiation; CyA = cyclosporin A; MTX = methotrexate.

Predisposing factors for the development of late-onset HC were analyzed by univariate and multivariate analysis. Clinical factors including gender, age at HSCT, underlying diagnosis, type of HSCT, conditioning regimens, GVHD prophylaxis, presence of acute or chronic GVHD, and immunosuppressive therapy for post-HSCT complications were evaluated. Univariate analysis was performed by Fisher’s exact test. All factors associated with the risk of lateonset HC by univariate analysis were examined by multivariate analysis using a logistic regression procedure. Results Late-onset HC occurred in 26 of 256 patients (10.2%). Sixteen patients were boys and 10 were girls. Age at HSCT ranged from 2 to 16 years (median, 9 years). The underlying diagnoses were acute lymphoblastic leukemia (n = 14), acute myelogenous leukemia (n = 6), chronic myelogenous leukemia (n = 1), myelodysplastic syndrome (n = 2), nonHodgkin’s lymphoma (n = 1) and aplastic anemia (n = 2). The onset of HC ranged from 3 to 270 days (median, 33 days) after HSCT. Twelve patients developed mild HC and

11 moderate HC. Although patients were suffering from serious micturition pain and pollakiuria, symptoms were well controlled by hydration and oral analgesia. Three patients developed severe HC, and all three required bladder irrigation. One went on to develop acute renal failure due to urinary obstruction and required suprapubic cystotomy. Two of the three patients recovered 66 and 86 days after the onset of HC, but the third died of interstitial pneumonia (IP) before cystitis resolved. The median duration of HC in 22 patients with resolution was 16 days (range, 6– 86 days). Four patients died before resolution of HC between 14 and 58 days after the onset of cystitis. Causes of death were IP (n = 2), veno-occlusive disease (n = 1) and acute renal failure caused by CyA (n = 1); HC was never the direct cause of death. In six of 22 patients whose urine specimens at the onset of HC were examined, adenovirus type 11 only was isolated. The association of various clinical factors with late-onset HC was examined by univariate analysis (Table 2). Because a significantly higher incidence of late-onset HC was observed in allogeneic than autologous HSCT (P = 0.043), we additionally analyzed other factors separately in allogeneic patients. In all patients, two other factors were identified as predisposing to HC besides allogeneic HSCT; the incidence of late-onset HC was significantly higher in older patients (⭓7 years at HSCT) (P = 0.047), and in patients receiving BU as part of the conditioning regimen (P = 0.013). Although administration of BU remained statistically significant (P = 0.022), age at HSCT was not a factor (P = 0.077) in allogeneic patients. Underlying diagnosis and administration of MEL as part of the conditioning regimen were also identified as risk factors in allogeneic patients (P = 0.032 and P = 0.036, respectively). Immunosuppressive agents including CyA, ATG, and methyl-prednisolone were often administered for prophylaxis of GVHD or treatment of post-HSCT complications. None were identified as risk factors for HC. Neither acute nor chronic GVHD correlated with the development of HC in allogeneic patients. By multivariate analysis, three factors were identified as predisposing to the development of late-onset HC in all patients (Table 3): type of HSCT (P = 0.022), age at HSCT (P = 0.044), and administration of BU (P = 0.036). Discussion HC is a well-recognized complication of HSCT with an incidence varying from 6.5 to 68%.2 The term ‘late-onset HC’ is used to exclude CY-induced HC in patients undergoing HSCT. It is well known that CY often induces early-onset HC,19 which usually occurs during or immediately after CY administered as part of the conditioning regimen. In the majority of patients, HC has been attributed to a direct toxic effect of CY metabolites on the urothelium. HC occurring late after HSCT is unlikely to be due to the conditioning regimen, and therefore probably has another pathogenesis. Associations with BK virus7–10 or adenovirus type 11,11–13 acute or chronic GVHD,14,15 and chemotherapeutic agents16,17 have been demonstrated in late-onset HC. In particular, some observations suggested an association


Table 2

997

Predisposing factors for late-onset hemorrhagic cystitis: univariate analysis

Factors

All patients (n = 256)

Variables

Gender Age at HSCT Underlying diagnosis Type of HSCT Conditioning regimen CY BU MEL TBI Immunosuppressive agents CyA ATG High-dose mPSL therapy Acute GVHD Chronic GVHD

Allogeneic patients (n = 183)

Incidence

P value

Incidence

P value

Male Female ⬍7 years ⭓7 years

17/137 9/119 7/116 19/140

0.20

15/104 8/79 6/79 17/104

0.39

Malignancy Non-malignancy

24/206 2/50

0.11

21/133 2/50

0.032

Allogeneic Autologous

23/183 3/73

0.043

—

—

Yes No Yes No Yes No Yes No

6/95 20/161 17/109 9/147 20/159 6/97 16/146 10/110

0.12

6/85 17/98 15/79 8/104 17/98 6/85 13/112 10/71

0.036

Yes No Yes No Yes No

10/110 16/146 4/41 22/215 8/50 18/206

0.62

10/110 13/73 4/41 19/142 8/47 15/136

0.082

None–1 2–4 Yes No

—

—

0.73

—

—

20/163 3/20 1/28 22/155

0.047

0.013 0.10 0.62

0.93 0.13

0.077

0.022 0.036 0.62

0.54 0.29

0.12

HSCT = hematopoietic stem cell transplantation; CY = cyclophosphamide; BU = busulphan; MEL = melphalan; TBI = total body irradiation; GVHD = graft-versus-host disease; CyA = cyclosporin A; ATG = anti-thymocyte globulin; mPSL = methyl-prednisolone.

Table 3 Predisposing factors for late-onset hemorrhagic cystitis: multivariate analysis Factors All patients (n = 256) Type of HSCT (allogeneic vs autologous) Age at HSCT (⬍7 years vs ⭓7 years) Administration of BU (yes vs no)

OD

CI

P value

4.4

1.2–15.8

0.022

2.6

1.0–6.6

0.044

2.7

1.1–6.8

0.036

OD = odds ratio; CI = 95% confidence intervals; HSCT = hematopoietic stem cell transplantation; BU = busulphan.

between late-onset HC after HSCT and BK virus reactivation. The difficulties in BK virus isolation by cell culture methods are well known; therefore, a DNA hybridization assay or the polymerase chain reaction were usually used for the detection of viral urinary shedding.9 Unfortunately, excretion of BK virus was not demonstrated in any patients, because we did not examine for the presence of BK virus by these methods. As previously reported, we often isolated adenovirus type 11 in patients with late-onset HC after HSCT. To date, only a few reports have focused on HC in pediatric stem cell transplant patients. In this study, we

reviewed the clinical records of 256 consecutive children receiving HSCT in order to find predisposing factors for late-onset HC. In our series, late-onset HC occurred in 26 of 256 patients (10.2%), an incidence lower than that of previous reports.1–3,6–17 Most previous reports did not distinguish early-onset HC from late-onset HC, and this could be one reason for the relatively low incidence in our series. However, we suspect that patient age is more important with regard to this point. Brugieres et al1 reported that the incidence of HC was 6.5% in their childhood series. This figure is lower than those of adult series, and they also showed that young age correlated with a lower incidence of HC. Sencer et al2 also reported a significantly higher rate of HC in patients older than 10 years by univariate analysis, but this finding was not identified by multivariate analysis. In our series, we identified a lower incidence of late-onset HC in younger children by multivariate analysis. Brugieres et al1 have suggested that the low incidence of HC in young children may be due to frequent voiding which leads to brief contact between CY metabolites and the bladder mucosa. In patients with late-onset HC, however, the prevalence of virus seems to be more important than the habit of frequent voiding. Dei et al20 have reported that prevalence rates of BK virus increased proportionally from birth until the age of 12 years. Because the incidence


998

of latent infection with adenovirus is lower in younger children than in older children,21 we speculate that the lower incidence of latent infection with these viruses may lead to a lower incidence of late-onset HC in younger patients. Administration of BU was also reported to be a risk factor for HC.16,17 Our study confirmed that patients receiving BU as pretransplant conditioning developed HC more frequently than those who did not. BU itself produces epithelial atypia and is thought to contribute to the pathogenesis of late-onset HC.22,23 Another possibility is that BU damages the uroepithelium and predisposes to infection with some viruses. This could explain the relatively long duration between time of BU administration and the onset of HC. Sencer et al2 have reported that higher rates of HC were seen in an allogeneic population, and that factors such as adenoviruria and underlying diagnosis were implicated in the high incidence of HC. We also identified allogeneic HSCT as a predisposing factor for HC by multivariate analysis, but no other features were identified as significant factors in allogeneic patients. Immunosuppression after HSCT is usually stronger in allogeneic recipients than in autologous recipients. In order to prevent graft failure and GVHD, the pretransplant conditioning regimen is intensified, and immunosuppressive agents are used. We speculate that, although no immunosuppressive factor correlated directly with the development of HC, immunosuppression in allogeneic recipients may be associated with reactivation of virus, leading to higher rates of late-onset HC. In summary, three independent factors predisposing to late-onset HC were identified. Although we suspect that the main pathogenesis of late-onset HC may be reactivation of virus, several clinical factors were also associated with the development of HC. A prospective study is necessary to clarify the association between viral reactivation and other clinical factors in the development of late-onset HC following HSCT.

References 1 Brugieres L, Hartmann O, Travagli JP et al. Hemorrhagic cystitis following high-dose chemotherapy and bone marrow transplantation in children with malignancies: incidence, clinical course, and outcome. J Clin Oncol 1989; 7: 194–199. 2 Sencer SF, Haake RJ, Weisdorf DJ. Hemorrhagic cystitis after bone marrow transplantation. Risk factor and complications. Transplantation 1993; 56: 875–879. 3 Russell SJ, Vowels MR, Vale T. Haemorrhagic cystitis in paediatric bone marrow transplant patients: an association with infective agents, GVHD and prior cyclophosphamide. Bone Marrow Transplant 1994; 13: 533–539. 4 Baronciani D, Angelucci E, Erer B et al. Suprapubic cystotomy as treatment for severe hemorrhagic cystitis after bone marrow transplantation. Bone Marrow Transplant 1995; 16: 267–270.

5 Thomas ED, Storb R, Clift RA et al. Bone marrow transplantation. New Engl J Med 1975; 292: 835–843, 895–902. 6 DeVries CR, Freiha FS. Hemorrhagic cystitis: a review. J Urol 1990; 143: 1–9. 7 Arthur RR, Shah KV, Baust SJ et al. Association of BK viruria with hemorrhagic cystitis in recipients of bone marrow transplants. New Engl J Med 1986; 315: 230–234. 8 Apperley JF, Rice SJ, Bishop JA et al. Late-onset hemorrhagic cystitis associated with urinary excretion of polyomaviruses after bone marrow transplantation. Transplantation 1987; 43: 108–112. 9 Azzi A, Fanci R, Bosi A et al. Monitoring of polyomavirus BK viruria in bone marrow transplantation patients by DNA hybridization assay and by polymerase chain reaction: an approach to assess the relationship between BK viruria and hemorrhagic cystitis. Bone Marrow Transplant 1994; 14: 235–240. 10 Bedi A, Miller CB, Hanson JL et al. Association of BK virus with failure of prophylaxis against hemorrhagic cystitis following bone marrow transplantation. J Clin Oncol 1995; 13: 1103–1109. 11 Ambinder RF, Burns W, Forman M et al. Hemorrhagic cystitis associated with adenovirus infection in bone marrow transplantation. Arch Intern Med 1986; 146: 1400–1401. 12 Miyamura K, Takeyama K, Kojima S et al. Hemorrhagic cystitis associated with urinary excretion of adenovirus type 11 following allogeneic bone marrow transplantation. Bone Marrow Transplant 1989; 4: 533–535. 13 Miyamura K, Minami S, Matsuyama T et al. Adenovirusinduced late onset hemorrhagic cystitis following allogeneic bone marrow transplantation (Correspondence). Bone Marrow Transplant 1987; 2: 109–110. 14 Ost L, Lonnqvist B, Eriksson L et al. Hemorrhagic cystitis – a manifestation of graft versus host disease? Bone Marrow Transplant 1987; 2: 19–25. 15 Ruutu T, Ruutu M, Volin L et al. Severe cystitis as a manifestation of chronic graft-versus-host disease after bone marrow transplantation. Br J Urol 1988; 62: 612–613. 16 Thomas AE, Patterson J, Prentice HG et al. Haemorrhagic cystitis in bone marrow transplantation patients: possible increased risk associated with prior busulphan therapy. Bone Marrow Transplant 1987; 1: 347–355. 17 Atkinson K, Biggs J, Noble G et al. Preparative regimens for marrow transplantation containing busulphan are associated with haemorrhagic cystitis and hepatic veno-occlusive disease but a short duration of leucopenia and little oro-pharyngeal mucositis. Bone Marrow Transplant 1987; 2: 385–394. 18 Horwitz MS. Adenovirus disease. In: Fields BN (ed). Virology. Raven Press: New York, pp 477–495. 19 Stillwell TJ, Benson RC. Cyclophosphamide-induced hemorrhagic cystitis. A review of 100 patients. Cancer 1988; 61: 451–457. 20 Dei R, Marmo F, Corte D et al. Age-related changes in the prevalence of precipitating antibodies to BK virus in infants and children. J Med Microbiol 1982; 15: 285–291. 21 Numazaki Y, Shigeta S, Kumamoto T et al. Acute hemorrhagic cystitis in children. New Engl J Med 1968; 278: 700–704. 22 Millard RJ. Busulfan-induced hemorrhagic cystitis. Urology 1981; 18: 143–144. 23 Pode D, Perlberg S, Steiner D. Busulfan-induced hemorrhagic cystitis. J Urol 1983; 130: 347–348.