Introduction

Cyclophosphamide is a commonly used alkylating agent in the treatment of lymphoma, hemangiosarcoma, and other neoplasms.^1–5 It can also be used in combination with piroxicam as part of a metronomic chemotherapy protocol.^6,7 Cyclophosphamide is a prodrug metabolized in the liver. The majority of the drug is excreted as metabolites and unchanged drug in the urine.^8,9 Common side effects include myelosuppression and gastroenterocolitis. Sterile hemorrhagic cystitis (SHC), a specific side effect of cyclophosphamide, may occur because of the drug’s metabolite, acrolein, which causes a physical irritating effect on the bladder epithelium.¹⁰

Furosemide is a loop diuretic commonly used for the treatment of congestive cardiomyopathy, pulmonary edema, hypercalciuric nephropathy, and hyperkalemia.⁸ The adverse effects of furosemide include dehydration, hyponatremia, hypocalcemia, hypokalemia, and hypomagnesemia. Furosemide can also cause nephrotoxicity. In humans, diuretics are the second most common drug classes associated with acute kidney injury.¹¹ High-dose furosemide use is associated with nephrocalcinosis (mean 538 mg/day, range 120–2800 mg/day) and increased mortality for heart failure human patients (>80 mg/day).^12–14 It was shown that single IV doses of furosemide (1 mg, 3 mg, and 10 mg) increased urinary thiamine excretion.¹⁵ Severe thiamine deficiency was detected in 96% of human patients when furosemide was used at a dose ≥80 mg/day for long-term use.¹⁶

Concurrent administration of furosemide has been reported to decrease the incidence of SHC in dogs receiving the maximum tolerated dose (MTD) or metronomic cyclophosphamide.^6,17 In a report evaluating the incidence of SHC in dogs receiving cyclophosphamide with or without furosemide, 216 dogs were treated with cyclophosphamide at a dose of 200 mg/m² IV, and 83 of these dogs also received furosemide at a dose of 2.2 mg/kg IV. One dog in the furosemide treatment group (1.2%) developed SHC, whereas 12 of 133 (9%) dogs that did not receive furosemide developed SHC.¹⁷ Multi-agent protocol, called CHOP, is mainstay of treatment for lymphoma and it is composed of cyclophosphamide, doxorubicin, vincristine, and prednisone. Multiple CHOP-based protocols for treating canine lymphoma have been published using cyclophosphamide at a dose of 250 mg/m² IV or per os (PO), but furosemide dosing was not always reported and did not specifically assess for incidence of SHC.^1,18–21

The purpose of this study was to evaluate the incidence of SHC in dogs receiving a single MTD oral cyclophosphamide with oral furosemide at a dose of 1.0 mg/kg.

Materials and Methods

Medical records at Western College of Veterinary Medicine from January 2008 through January 2019 were reviewed. Dogs were eligible for inclusion in the study if they had a cytopathologic or histopathologic diagnosis of lymphoma and were treated with at least one dose of MTD cyclophosphamide. Information retrieved from the medical records included signalment, body weight, stage, substage, immunophenotype, date of diagnosis, concurrent disease, treatment protocol, cyclophosphamide dose, furosemide dose, and adverse events.

Lymphoma was staged based on criteria published by the World Health Organization.²² Staging diagnostic tests included complete blood counts (CBCs), serum chemistry, urinalysis, three-view thoracic radiographs, and abdominal ultrasound. The selection of specific staging tests was at the discretion of the attending clinician and owner. Dogs were classified as substage a if they did not have clinical signs and as substage b if they had clinical signs attributable to their disease.

Adverse events were characterized according to criteria set forth by the Veterinary Cooperative Oncology Group.²³ The following criteria was used to diagnose SHC: hematuria following treatment with cyclophosphamide, presence of red blood cells on urinalysis, negative urine culture, and absence of urolithiasis or urinary bladder mass on abdominal ultrasound. Incidence was presented as a rate and calculated by using the number of patients developing divided by the total number of patients treated.

Results

Eighty-one dogs met the inclusion criteria. The median age at diagnosis was 8 yr (range 4–14 yr). Median body weight was 28.2 kg (range 4.8–62 kg). Thirty-nine dogs were spayed female, 38 dogs were castrated male, and 4 dogs were intact male. Mixed-breed dogs (n = 26), golden retrievers (n = 5), boxers (n = 4), and German shepherd dogs (n = 4) were the most commonly affected dogs. Other dog breeds included beagle (n = 3), border collie (n = 3), Labrador retriever (n = 3), Shetland sheepdog (n = 3), cocker spaniel (n = 2), miniature poodle (n = 2), Pit bull terrier (n = 2), rottweiler (n = 2), Airedale terrier (n = 1), Alaskan malamute (n = 1), American Eskimo (n = 1), Australian shepherd (n = 1), basset hound (n = 1), Bernese mountain dog (n = 1), Australian cattle dog (n = 1), bulldog (n = 1), Boston terrier (n = 1), Doberman pinscher (n = 1), Dogue de Bordeaux (n = 1), English setter (n = 1), goldendoodle (n = 1), Great Dane (n = 1), Jack Russell terrier (n = 1), mastiff (n = 1), pug (n = 1), Rhodesian ridgeback (n = 1), shihtzu (n = 1), Staffordshire bull terrier (n = 1), Maltese (n = 1), and miniature schnauzer (n = 1).

At the initial presentation, all dogs had a CBC evaluation. Serum chemistry panel was evaluated in 73 dogs, and urinalysis was evaluated in 52 dogs. Thoracic radiographs were performed in 66 dogs, and abdominal ultrasound was performed in 38 dogs. Cytopathology of the liver was evaluated in 6 dogs, and cytopathology of the spleen was evaluated in 4 dogs. The most common anatomic form of lymphoma was multicentric (n = 71), followed by hepatosplenic (n = 3), gastrointestinal (n = 3), mediastinal (n = 2), cutaneous (n = 1), and ocular (n = 1). The stage and substage information for the 71 cases with multicentric lymphoma are summarized in Table 1. Staging information was not available in 1 dog because of the lack of the information of initial visit in the medical record. However, this dog was included in the study because the cyclophosphamide dose and outcome information were available. Immunophenotype was evaluated by flow cytometry in 5 dogs. Three dogs had B-cell lymphoma and 2 dogs had T-cell lymphoma. Hypercalcemia was noted in 8 dogs. The following concurrent diseases were reported: heart murmur (n = 7), mast cell tumor (n = 4), hypothyroidism (n = 2), lung mass (n = 1), idiopathic epilepsy (n = 1), immune-mediated hemolytic anemia (n = 1), immune-mediated thrombocytopenia (n = 1), Cushing’s disease (n = 1), hemangioma (n = 1), urinary tract infection (n = 1), heart-based mass (n = 1), and mammary mass (n = 1).

TABLE 1 Staging of Multicentric Lymphoma Patients

View Fullscreen

Patients were initially treated with a 19 wk CHOP protocol (59 dogs; Table 2), CMOP (cyclophosphamide, mitoxantrone, vincristine, and prednisone) protocol (12 dogs; Table 2), L-CHOP (L-asparaginase and CHOP) protocol (6 dogs), and CHOP protocol with half-body radiation therapy (1 dog). One dog was treated with each doxorubicin, mitoxantrone, or cyclophosphamide single agent as a first-line treatment and then treated with a CHOP protocol at the time of relapse. Eleven dogs were treated with CHOP or CMOP at the time of relapse. Other drugs used for rescue therapy included L-asparaginase, lomustine, doxorubicin, mitoxantrone, rabacfosadine, and chlorambucil.

TABLE 2 CHOP or CMOP Chemotherapy Protocol

View Fullscreen

At the end of study, 13 dogs were still alive, 23 dogs were lost to follow-up, and 45 dogs died or were euthanized. The median time to lost to follow-up was 232 days (range of 44–1405 days). A total of 252 doses of cyclophosphamide were administered PO. The median number of cyclophosphamide treatments was 3 (range 1–8 treatments), and the median dose of each cyclophosphamide treatment was 239.3 mg/m² (range 160–270 mg/m²). Furosemide was administered PO immediately after cyclophosphamide. The median dose of furosemide was 1.08 mg/kg (range 0.59–2.35 mg/kg). Two dogs developed SHC (2.46%), and the cumulative risk was 0.79%.

The first dog developed pollakiuria and hematuria 3 days after the third treatment with cyclophosphamide with a dose of 200.8 mg/m². The cumulative dose of cyclophosphamide at this time was 595.3 mg/m². Furosemide was administered with 0.7 mg/kg PO. CBC and serum chemistry panel were unremarkable. Urine was obtained by catheterization. Urine color was yellow, the pH was 6.0, and the urine specific gravity was 1.058. Erythrocytes were present on urinalysis (4+ blood; 20–30 erythrocytes/high-power field). Urine culture was negative. An abdominal ultrasound revealed thickening of the urinary bladder wall with no evidence of masses or urolithiasis. No other significant findings were noted in the remainder of the abdomen. Meloxicam (0.1 mg/kg PO once a day) was prescribed, and clinical signs improved within 2 wk.

The second dog was reported to have developed stranguria and hematuria 2 wk after the third dose of cyclophosphamide with a dose of 234 mg/m². The cumulative dose of cyclophosphamide was 715.9 mg/m². The furosemide dose was 1.25 mg/kg. Urine was obtained by cystocentesis. Urine color was cloudy yellow with a pH of 9.0. Urine specific gravity was 1.025. Proteinuria and erythrocytes were noted on the urinalysis (2+ protein, 3+ blood; 0–2 erythrocytes/high-power field). The urine dipstick and urine sediment do not always directly correlate with each other. A plausible explanation for this discordance is that some degree of in vitro erythrocyte lysis may have occurred, which could contribute free hemoglobin to the sample without overtly causing gross discoloration. An abdominal ultrasound showed a thickened ventral wall and apex of the urinary bladder. Amoxicillin/clavulanic acid was prescribed initially pending urine culture results. Urine culture was negative for bacterial infection. Clinical signs improved without additional therapy.

Neither of the dogs had any previous concurrent disease. They also did not have abdominal ultrasound before their episodes of urinary signs. Other adverse effects of cyclophosphamide in all dogs included neutropenia, vomiting, diarrhea, nausea, lethargy, and weight loss. No dogs died or were euthanized because of the side effects of oral cyclophosphamide.

Discussion

In this study, we report the development of SHC in 2 of 81 dogs (2.46%) after treatment with cyclophosphamide and concurrent administration of furosemide. The reported incidence of SHC after cyclophosphamide treatment is 0–22%.^{17,20,21,24–28} The dosage of cyclophosphamide and furosemide vary in each study (Table 3). Moreover, reporting the incidence of SHC with cyclophosphamide was not the primary purpose of most of these studies, and the true incidence of SHC may have been underreported. Two studies looked primarily at the incidence of SHC following cyclophosphamide treatment.^6,17 Charney et al. reported an incidence of 1.2% of SHC when using a dose of 2.2 mg/kg of IV furosemide with 200 mg/m² of IV cyclophosphamide. This incidence appears similar to slightly lower than the 2.46% incidence we reported in this study. The lower dose of furosemide in our study may have resulted in this small difference. However, the dose of cyclophosphamide in this study was also higher than the dose used in the study of Charney et al. that may have resulted in an increased incidence of SHC rather than the lower dose of furosemide. It is difficult to simply compare the incidence of SHC between our study and the study of Charney et al. However, Charney et al. reported an incidence of 9% of dogs developing SHC when treated with cyclophosphamide at a dose of 200 mg/m² IV without furosemide.¹⁷ Another report by Chun, Garrett, and Vail mentioned that the incidence of SHC was 22% when cyclophosphamide was given at a dose of 200–250 mg/m² IV without furosemide.²⁰ When using these historical controls, the dose of 1.0 mg/kg of furosemide appears to decrease the incidence of cyclophosphamide associated SHC in our population.

TABLE 3 Previous Report for the Incidence of SHC in Dogs

View Fullscreen

Setyo et al. reported the incidence of SHC in dogs treated with metronomic cyclophosphamide and furosemide at a dose of 0.5–1.0 mg/kg PO.⁶ In this study, the incidence of SHC was 10.2%. However, the cyclophosphamide in this study was administered daily at a dose of 10–15 mg/m² PO. Daily administration may have led to prolonged exposure of the bladder wall to acrolein resulting in the increased incidence of SHC compared with our study using a similar dose of furosemide. Because the incidence of SHC may be different between MTD and metronomic dosing, it appears important to evaluate the effects of dosing furosemide for each protocol separately.

Best and Fry reported a protocol of MTD cyclophosphamide administration divided over 3 days. In this study, 57 dogs received cyclophosphamide PO over 3 consecutive days at a median dose of 236.0 mg/m² without concurrent furosemide; and no dogs developed SHC.²⁹ This could be because of the decreased acrolein concentration in urine. Although this study showed a lower incidence of SHC, the efficacy of this protocol was not compared with a single MTD dose administration. Single dosing of cyclophosphamide presents the advantage of not relying on client compliance for treatment administration and decreased client handling of chemotherapy drugs. Lee, Liao, and Wang recently evaluated the incidence of SHC between the groups of single MTD dose administration and divided doses of cyclophosphamide. In their study, there was no significant difference of the incidence.³⁰ Further evaluation is needed for over 3 consecutive days course of cyclophosphamide administration in dogs.

The diagnosis of SHC is not well defined and is often a diagnosis of exclusion associated with a recent history of administration of cyclophosphamide. In human medicine, diagnosis of cyclophosphamideinduced hemorrhagic cystitis is based on (a) a history of gross hematuria, (b) laboratory findings of microscopic hematuria, (c) platelet counts of >50,000/mm³, and (d) a lack of significant bacterial growth on urine culture.³¹ The two patients in this study were diagnosed with SHC based on clinical signs, microscopic hematuria, a negative urine culture, and the absence of concurrent masses or urolithiasis on abdominal ultrasound. CBCs were checked weekly during the chemotherapy protocol, and neither patient developed thrombocytopenia. Common causes of hematuria in dogs include urinary tract infection, urolithiasis, trauma, neoplasia, and bleeding disorders.³² These diseases were ruled out in two dogs with SHC in our study. It is possible that one or both patients did not have SHC. This would result in the incidence of SHC being overestimated in this study. Because the incidence of SHC reported in this study is low, an overestimation of the incidence does not appear to be a significant limitation.

Other limitations of this study include its retrospective nature and the lack of a control group. Urinalysis was not performed routinely in patients without clinical signs. Microscopic hematuria could have been missed, and episodes of SHC may have gone unreported leading to an underestimation of the incidence of SHC. In this study, 23 dogs were lost to follow-up, and it is possible that other dogs developed SHC at a later time. This study had no control group of dogs treated with cyclophosphamide without concurrent furosemide, and the data were compared with historical controls.

Clinical signs in both dogs developing SHC in this study resolved within 1 month. Abdominal ultrasounds were not repeated after resolution of the clinical signs. One dog was treated with meloxicam, and the second one received only antibiotics. There is no gold standard for treatment of SHC. In previous reports, discontinuation of cyclophosphamide, treatment with analgesic drugs, spasmolytic drugs, steroids, or antibiotics have been used.^6,27,28 Some dogs also required hospitalization and IV administration of opioids.¹⁷ Although furosemide at a dose of 1.0 mg/kg did not prevent SHC, the dogs that experienced this adverse event had only mild clinical signs that resolved in both cases.

Conclusion

The incidence of SHC when dogs were treated with bolus MTD oral cyclophosphamide and 1.0 mg/kg oral furosemide was similar with previous reports.^17,24 This dose of furosemide appears effective in decreasing the risk of SHC in dogs treated with cyclophosphamide.