Objectives: BK virus-associated hemorrhagic cystitis is a common complication of allogeneic hematopoietic stem cell transplant. It is known to be associated with cyclophosphamide therapy and the intensity of the conditioning regimen as well as infection with the BK virus. Data are limited for BK virus-associated hemorrhagic cystitis in pediatric recipients of allogeneic hematopoietic stem cell transplant. Therefore, we aimed to identify the risk factors and etiology of BK virus-associated hemorrhagic cystitis and determine the factors that may improve the treatment efficacy.
Materials and Methods: Data from recipients of allogeneic hematopoietic stem cell transplant were retrospectively analyzed. These data included information about age, sex, underlying disease, the details of ablative conditioning, graft-versus-host disease prophylaxis, donor type, stem cell source, history of acute graft-versus-host disease, and cytomegalovirus reactivation.
Results: A total of 50 patients developed BK virus-associated hemorrhagic cystitis among 334 patients. Symptoms associated with BK virus-associated hemorrhagic cystitis manifested an average of 45.3 days after transplant. Most of the patients had grade 2 and grade 3 hemorrhagic cystitis. Risk factor analysis revealed that haploidentical donor type, treatment with busulfan and cyclophosphamide as part of conditioning regimen, and history of total body irradiation increased the risk of BK virus-associated hemorrhagic cystitis in the pediatric recipient population.
Conclusions: We found that, despite current conditioning regimens, BK virus-associated infection still leads to a considerable incidence rate of hemorrhagic cystitis in pediatric recipients of allogeneic hematopoietic stem cell transplant. Patients with a haploidentical donor and a history of busulfan and cyclophosphamide treatment or total body irradiation had a higher risk of BK virus-associated hemorrhagic cystitis. Thus, we suggest that patients with these factors should be followed closely after allogeneic hematopoietic stem cell transplant.
Key words : Allogeneic hematopoietic stem cell transplant, Cyclophosphamide therapy
Hemorrhagic cystitis (HC) is a common cause of morbidity and mortality among allogeneic hematopoietic stem cell transplant (HSCT) recipients. Patients with HC usually present with only microscopic hematuria, with or without clot formation, which may progress to urinary tract obstruction and renal failure. Early-onset HC that occurs during or shortly after (up to 48 hours) initiation of high-dose chemotherapy is mainly associated with the use of oxazaphosphorines, such as cyclophosphamide and ifosfamide, or busulfan.1,2
Hemorrhagic cystitis can also be caused by viral infections in severely immunosuppressed patients. The BK virus is known to cause nephropathy in recipients of kidney transplants and may also result in HC among recipients of HSCT.3,4 Various studies have suggested that infections with various other viruses including adenovirus (mainly type 2), cytomegalovirus (CMV), and human herpesvirus type 6 are also responsible for late-onset HC (later than 48 hours after completion of conditioning).5,6
Among these viruses, BK virus is the causal factor in the majority of cases, and BK virus-associated HC (BKV-HC) has been shown to affect 7.8% to 22.0% of pediatric recipients of HSCT.7-10 The BK virus is a nonencapsulated double-stranded DNA virus with a high prevalence in healthy adults and seropositivity up to 90.0%.
The transmission of BK virus likely occurs via the respiratory tract, especially in early childhood. The infection is usually asymptomatic, or fever and mild upper respiratory symptoms may occur. BK polyomavirus may remain silent in the urothelium, kidneys, or other organs after initial infection but may be reactivated by immunosuppression. In allogeneic HSCT recipients, immunosuppression due to aplasia or treatment for the prevention of graft rejection or graft-versus-host disease (GVHD) may often lead to BK virus reactivation.11
The management of HC is generally symptomatic and supportive and includes hyperhydration, forced diuresis, and continuous bladder irrigation.1,12 Local or systemic anti-inflammatory agents such as aluminum, formalin, prostaglandin E1, glycosaminoglycans (chondroitin sulfate or hyaluronate), hemostatic agents such as intravenous factor XIII concentrate, antiviral agents (such as leflunomide), mesenchymal cells, and estrogen are other treatment options reported in the literature, but their efficacy remains unclear. However, cystectomy should be considered in resistant cases and in cases for which there is deterioration of renal function and severe urinary tract obstruction.1
Various studies have shown the effectiveness of antiviral agents such as vidarabine, ribavirin, and cidofovir; however, these treatments are not standard therapeutic options in the setting of BKV-HC.1 To date, no antiviral drug has been proved effective against BK virus replication, but some case reports and case series have suggested that cidofovir could be beneficial in the treatment of BKV-HC.11 Some authors have recommended hyperbaric oxygen therapy or fibrin-glue application, both of which are procedures that may accelerate the healing of damaged urothelial lining.5,13
In this study, we aimed to identify possible risk factors for the development of BKV-HC after HSCT and to evaluate the clinical factors for the severity of HC based on our single-center experience with pediatric patients diagnosed with BKV-HC.
Materials and Methods
We studied data for pediatric recipients who underwent allogeneic HSCT between September 2013 and December 2020 at the Pediatric Hematology/Oncology and Bone Marrow Transplantation Unit. All patients provided informed consent for this retrospective study, which was approved by the Ethics Committee of Acibadem University. We reviewed the medical records and clinical data to identify patients with evidence of BKV-HC after allogeneic HSCT.
Files of the patients with BKV-HC were further reviewed for data regarding the onset of HC after transplant, signs and symptoms, duration of symptoms, grade and severity of HC, and presence of concurrent systemic infections, as well as GVHD, treatment options, relapses, and outcomes. These data were recorded andanalyzed.
Data obtained from patients diagnosed with BKV-HC and confirmed via polymerase chain reaction (PCR) assay of urine were compared with data from the control group (patients without BKV-HC). The urine samples from patients who developed HC were examined for BK viremia by PCR assays performed in the Central Virology Laboratory of our university in Istanbul, Turkey. None of the patients with HC had a history of documented urinary tract infection, bladder irradiation, or radiological evidence of renal or bladder calculi.
Hemorrhagic cystitis prevention and management
Hyperhydration (2.5 to 3 L/m2/d) with forced alkaline diuresis was administered to all patients who underwent HSCT. All patients who received cyclophosphamide or busulfan were treated prophylactically with 2-mercaptoethane sulfonate sodium during the conditioning regimen. As part of the conditioning regimen, all patients were given acyclovir to prevent infection with herpes simplex and varicella-zoster viruses. Fluconazole was given for antifungal prophylaxis, and trimethoprim-sulfamethoxazole was prescribed for Pneumocystis prophylaxis. Recipients of grafts from matched related donors or matched unrelated donors were given cyclosporine with low-dose methotrexate to prevent GVHD.
The standard treatment for BKV-HC has not been established. We implemented different treatment options to address the various levels of severity for BKV-HC. Patients with mild BKV-HC were given supportive treatment with intravenous hydration, analgesics (ibuprofen, acetaminophen, or tramadol), and spasmolytics (such as oxybutynin). Patients with grade 3 or grade 4 HC were given additional treatment with hyperbaric oxygen therapy, intravenous factor XIII concentrate, estrogen administration, or cystoscopic catheterization. All patients were treated in a dedicated ward in a single room isolated by a high-efficiency particulate air filter.
Different antiviral treatment regimens (high-dose cidofovir combined with probenecid, or low-dose cidofovir) were used to address the various grades of HC severity and tolerance of the patients. Treatment was continued until hematuria and other urinary symptoms resolved. Urinary symptoms, renal function tests, and BK viral loads were monitored during treatment. Serum creatinine levels were measured at least twice a week during cidofovir treatment. Nephrotoxicity was defined as an increase in serum creatinine of 0.3 mg/dL or greater at any time during and up to 2 weeks after cidofovir treatment.
Hemorrhagic cystitis was defined as microscopic or macroscopic hematuria combined with other irritative urinary symptoms such as dysuria, pollakiuria, urgency, or sensation of residual urine in the absence of bacterial growth in urine culture and systemic coagulopathy. The BKV-HC status was defined as HC associated with BK viremia or viruria. Patients with HC present with variable degrees of hematuria, and the Droller grading system has been developed to assess the severity of HC.2-6 With this system, patients who present with only microscopic hematuria are assigned grade 1, patients with macroscopic hematuria are assigned grade 2, patients with gross hematuria with blood clots are assigned grade 3, and patients with urinary obstruction or renal dysfunction in addition to hematuria and clots are assigned grade 4.2-6,14
All patients with HC symptoms underwent a laboratory workup to rule out viral and other causes. This workup included real-time PCR to diagnose possible viremia or viruria with CMV, BK virus, adenovirus, and JC polyoma virus.
The date of onset for BKV-HC was defined as the first day when patients presented with urinary symptoms.5,14 The complete recovery from HC was defined as the resolution of microscopic or macroscopic hematuria with the disappearance of HC symptoms.
The positive virological response was defined as a reduction in BK viral load shown by quantitative PCR during weekly measurements. Urine viral loads (≥109-1010 copies, or ≥3-log increase from baseline) and the presence of plasma viremia greater than 104 copies/mL are diagnostic evidence for hematuria in HSCT patients. The classification of GVHD was assigned according to the Glucksberg criteria.
Categorical variables were compared by the chi-square test. The qualitative variables with normal distribution were compared by t test, and qualitative variables that were not normally distributed were compared by the Mann-Whitney U test. Variables with significant differences (P < .05) were entered into a backward stepwise logistic regression model for multivariate analysis.
A total of 334 HSCT procedures were performed for 314 patients during the study period at our center. Demographics and clinical characteristics of HSCT patients with or without BKV-HC are defined in Table 1.
Our analyses revealed that HC developed in 50 of 334 patients who underwent HSCT (14.9%), and all patients who developed HC had BK viruria, which was diagnosed by urine PCR analysis.
Among these patients with BKV-HC, 7 were grade 1, 22 were grade 2, 17 were grade 3, and 4 were grade 4. Among those patients with grade 3 and grade 4, 6 had concurrent CMV viremia (28.5%), but none of the patients with HC in the entire study group was positive for adenovirus. Two patients died from severe GVHD and concurrent BKV-HC (Table 2; Figure 1 and Figure 2).
The mean onset time of BKV-HC (ie, the time interval between the date of HSCT and onset of BKV-HC) in our patients was 45.3 days (minimum-maximum, 3-160 days), and 71.4% of these cases were determined to be late-onset BKV-HC, whereas 28.6% were determined to be early-onset BKV-HC.
The mean duration of HC was 43.7 days (minimum-maximum, 10-114 days). We observed GVHD in 38 of 50 patients with BKV-HC (76.0%). Among these 38 patients with GVHD, 16 had grade 2, 16 had grade 3, and 6 had grade 4. The mean onset time of GVHD was 33.6 days (minimum-maximum, 15-80 days) (Table 2). Table 1 shows the data for patients with and without BKV-HC.
The mean age of the patients diagnosed with BKV-HC was 11.82 ± 5.2 years (minimum-maximum, 5-23 years), whereas patients who did not develop HC (control group) had a mean age of 6.2 ± 3.8 years (minimum-maximum, 1-22 years). The age difference between the 2 groups was statistically significant
(P = .022). The male-to-female ratio was similar in both groups.
Comparison of the 2 groups revealed that BKV-HC was more common in patients with thalassemia major, Fanconi aplastic anemia, solid cancer/immunodeficiency, and acute lymphoblastic leukemia. Also, the incidence rate of BKV-HC was found to be higher in recipients with a matched unrelated donor and a haploidentical transplant. Furthermore, analyses of the medications and therapies of the conditioning regimen revealed that BKV-HC was more common among those treated with VP-16, busulfan, cyclophosphamide, and total body irradiation (TBI) (P < .05).
Seven pediatric patients were treated only with hydration and supportive treatment, 30 patients received bladder irrigation with chondroitin sulfate solution, and 40 patients received ciprofloxacin in addition to supportive treatment. Antiviral treatment was administered to 24 patients. Six patients received high-dose cidofovir (5 mg/kg/wk) combined with probenecid, and 18 patients received low-dose cidofovir (1 mg/kg, 3 times a week). Although probenecid was used to prevent nephrotoxicity in the high-dose cidofovir group, severe renal insufficiency developed in 2 patients during treatment. In the low-dose cidofovir group, no signs of nephrotoxicity were observed except for a slight increase in serum creatinine levels. The BK viruria levels measured by PCR at 15-day intervals decreased in patients who received antiviral treatment. Twelve patients were given hyperbaric oxygen therapy, and 6 patients were given estrogen therapy combined with hyperbaric oxygen therapy for 30 days as an adjunct to their primary treatment. Six patients underwent cystoscopic catheterization. Coagulation factor Xa was administered to 1 patient to treat massive bleeding; however, this patient did not benefit from this treatment and died from invasive fungal infection and secondary sepsis. Two patients died from severe GVHD, but all other patients recovered completely and had no episodes of recurrent bleeding during long-term follow-up.
The BK virus is integral to the etiopathogenesis of HC after HSCT.1-5 This virus remains latent in the uroepithelium and causes HC by reactivation when the immune system of the recipient is suppressed.1-3 Therefore, it is essential to identify the patients at risk for developing BKV-HC, either de novo or via reactivation.
Routine screening for this virus in the HSCT population is difficult and expensive. Therefore, the determination of the risk factors for BKV-HC is a reasonable alternative.
Various studies have reported incidence rates of BKV-HC of 7.8% to 22.0% in HSCT recipients.1,5,9-12,14-16 In agreement with these reports, the rate of incidence of BKV-HC in our HSCT recipients was 14.9%. Also, older age of HSCT recipients has been reported to be a risk factor.17,18 In our study, the patients in the BKV-HC group were older than the patients in the control group. Therefore, in pediatric HSCT recipients, older age may be considered a risk factor for BKV-HC.19,20 Uhm and colleagues and Rorije and colleagues reported that patient sex was not associated with post-HSCT BKV-HC.7,10 Similarly, we did not find a relationship between sex and BKV-HC in our study.
The mean onset time of BKV-HC was 45.3 days (minimum-maximum, 3-160 days) in our study, and 71.4% of these were late-onset cases. The early-onset BKV-HC rate was lower in our study compared with the previous reports. This difference may be associated with the BKV-HC prophylaxis that we included in our conditioning regimen. The BK virus loads in the urine of all of our patients were significantly high at the time of initial diagnosis.
Although BK viruria is present in 50.0% to 90.0% of HSCT recipients, several studies have shown that BK virus is detectable via real-time quantitative PCR in patients who develop BKV-HC.1,6 Our analyses revealed that BK viremia was present in 60.0% of our cases.
Gorczynska and colleagues and Cesaro and colleagues reported higher rates of BK viremia in their studies (81.0% and 87.5%).1,13 The rate of BK viremia in our patients who developed BKV-HC was lower than the rates reported in the literature, which indicates reactivation of BK virus in our study population rather than de novo BK virus infection.
A study by Uhm and colleagues identified CMV as a risk factor for BKV-HC.7 In our study, 22.9% of the patients with BKV-HC had concurrent CMV and BK viremia. Of note, the severity of BKV-HC was higher (grade 3 and grade 4) in 75.0% of the patients with CMV viremia. Patients with CMV viremia were treated with ganciclovir for 10 to 20 days (mean, 14 days), and those patients without BK viremia benefited from this treatment. The presence of CMV viremia may indicate increased immunodeficiency in this group of patients, and therefore these patients may be more susceptible to reactivation of BK virus.
Some studies have investigated the relationship between underlying disease and BKV-HC.7,10 Uhm and colleagues7 reported no significant relationship between BKV-HC and underlying disease. However, Rorije and colleagues10 found that the rate of BKV-HC was lower in patients with acute myeloid leukemia, but they did not find a relationship between other diseases and BKV-HC. In our study, the incidence of BKV-HC was higher in patients with acute lymphoblastic leukemia and thalassemia major.
Rorije and colleagues10 reported that recipients of transplants from mismatched donors were more likely to develop BKV-HC, whereas other types of transplants were not associated with BKV-HC. Our findings were similar, as we found BKV-HC was more frequent among recipients of allogeneic transplants from matched unrelated donors and haploidentical donors. However, allogeneic transplants received from matched related donors were not found to affect the rate of BKV-HC.
The association between conditioning agents and the occurrence of BKV-HC has been investigated in several studies.21-24 These studies have suggested that the level of overall immunosuppression was more critical than the use of specific chemotherapeutic agents with regard to the risk of post-HSCT BKV-HC.21
However, some medications used after HSCT are known to be associated with HC.22-27 Some studies have proposed that busulfan conditioning may a risk factor for BKV-HC.22-25 However, Rorije and colleagues10 evaluated the effects of the conditioning regimen on the risk of BKV-HC in 491 allogeneic HSCT recipients and reported that patients treated with fludarabine or busulfan had a lower risk of BKV-HC, whereas patients treated with cyclophosphamide or TBI had a higher risk. Furthermore, Rorije and colleagues reported that cyclophosphamide and TBI were significant risk factors for post-HSCT BKV-HC. In our study, the incidence rate for BKV-HC was 91.4% for patients who received antithymocyte globulins and cyclophosphamide, 85.7% for those who received busulfan, and 100.0% for those who received thiotepa and TBI. These findings agree with the results previously reported in the literature.
The presence of GVHD higher than grade 1 has been reported to be a potential risk factor for post-HSCT BKV-HC in the literature.15 Uhm and colleagues7 suggested that grade 3 or grade 4 GVHD was a strong risk factor for BKV-HC. In our study, 82.8% of the patients had grade 2 to grade 4 GVHD. This high rate of GVHD in our series is similar to the rates reported in other series.1,2,5,6 In addition, a relationship between GVHD/TBI and BKV-HC has been reported.20,28 As such, in our study, BKV-HC was diagnosed in both patients who received TBI treatment. We believe the higher rate of BKV-HC in patients with GVHD is the result of the aggressive immunosuppressive treatments given to these patients, which may lead to accelerated BK and CMV viral replication.
Gardner and colleagues17 reported 6 deaths (2 after recurrence) among 10 patients who developed BKV-HC. However, a retrospective study by Kwon and colleagues9 did not report any recurrence of HC among recipients of HSCT. In our study, all patients recovered except for 2 who died from severe GVHD, and no recurrent bleeding episodes were encountered during long-term follow-up.
Our study has some limitations. It is a retrospective study and therefore is susceptible to the inherent weakness of this design. Also, this is a single-center study with a small sample size.
BK virus infection is an important cause of mortality and morbidity in pediatric transplant recipients. Despite the limitations of our study, we conclude that the widely accepted major risk factors for post-HSCT BKV-HC in the adult population are also significant risk factors in pediatric patients; these risk factors include transplant type (allogeneic unrelated and haploidentical), the medications in the conditioning regimen such as busulfan and cyclophosphamide, and procedures such as TBI. Therefore, pediatric HSCT recipients with these risk factors should be considered at high risk for post-HSCT BKV-HC and should be followed closely and treated immediately after diagnosis, with awareness that such patients may require hemostatic treatments or interventions. Our study was conducted with a significant number of patients, is an important contribution to the literature, and will be a helpful resource for the future preparation of guidelines on this subject.
DOI : 10.6002/ect.2021.0104
From the 1Department of Pediatric Hematology/Oncology and Bone Marrow Transplantation Unit, the 2Department of Urology, and the 3Department of Radiology, Faculty of Medicine, Acibadem University, Adana Hospital; and the 4Department of Pediatric Nephrology, Faculty of Medicine, Cukurova University, Adana, Turkey
Acknowledgements: We thank the entire Acibadem University physician group for contributing to this challenging collaborative work. The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Barbaros Sahin Karagun, Department of Pediatric Hematology/Oncology and Bone Marrow Transplantation Unit, Faculty of Medicine, Acibadem University, Adana Hospital, 01130 Adana, Turkey
Phone: +90 506 2345145
Table 1. Clinical Characteristics of Pediatric Recipients With and Without BK Virus-Associated Hemorrhagic Cystitis After Hematopoietic Stem Cell Transplant
Figure 1. Suprapubic Ultrasonography and Confirmatory Computed Tomography
Figure 2. Computed Tomography and Ultrasonography of the Right Kidney
Table 2. Demographic and Clinical Characteristics of Pediatric Recipients With BK Virus-Associated Hemorrhagic Cystitis After Hematopoietic Stem Cell Transplant