Radiation therapy as a treatment for urinary obstruction secondary to urothelial carcinoma in dogs

Authors: Madelyn Burtz, Mary-Keara Boss, Del Leary, Tiffany W Martin

Abstract

Background

Urothelial carcinoma (UC) is the most common urinary bladder tumor in dogs, often invading and involving the bladder wall, trigone, urethra, and prostate and can lead to obstruction of urine outflow at the level of the ureters or urethra.

Hypothesis/Objectives

To describe the outcome of dogs with UC presenting with urethral or ureteral obstruction treated with radiation therapy (RT).

Animals

Twenty-five client-owned dogs treated with RT.

Methods

A retrospective analysis was conducted. Overall survival, time to first event, and restoration of patency were evaluated by obstruction location and radiotherapy protocol.

Results

Overall obstruction relief was achieved in 15/25 (60%) dogs. Definitive intent protocols showed significantly longer time to obstruction relief (median 8 vs 4 days; range 2-10 vs 2-7 days; P = .02), however, had higher overall efficacy (83% vs 53%) and survival times compared to palliative intent protocols (P = .01). Dogs that had relief of obstruction regardless of RT intent received a significantly higher total dose compared to those that did not experience relief (median 36 Gy vs 8 Gy; range 12-57 vs 6-24 Gy; P < .001). Median overall survival time was 134 days (range 6-969 days) and time to first event was 107 days (range 4-695 days). Location of obstruction did not significantly improve survival or time to obstruction relief. Relief was temporary for 27% of unobstructed dogs, and re-obstruction was seen at a median of 14 days (range 5-33 days).

Conclusions and clinical importance

Radiation therapy is a viable treatment option for relieving urinary obstruction secondary to UC. Radiation therapy protocol should be determined based on the dog’s clinical signs, ability to medically manage obstruction, and therapeutic goals of the owner.

Introduction

Urothelial carcinoma (UC) is the most common malignant canine urinary bladder tumor, accounting for 2% of tumors in dogs.^1,2^ These tumors often localize to the trigone region and can invade into the lumen of the urethra and ureters causing urinary obstruction.

Dogs with UC can present with pollakiuria, stranguria, and hematuria; however, growth of the tumor might result in partial or complete obstruction which can lead to hydronephrosis and bladder rupture. Dogs can present with life-threatening obstruction at time of diagnosis, which can lead to euthanasia if not addressed immediately.^2^ Definitive diagnosis of UC in dogs is achieved via histopathology of biopsy acquired via surgery, cystoscopy, or catheterization.^3–5^ Detection of BRAF proto-oncogene has been reported in 80% of dogs with UC; however, concentration of BRAF mutation in circulating cell-free DNA has not shown to be a reliable marker of tumor stage or metastases.^6^ BRAF testing can be used both to aid diagnosis of UC, and might have potential utilization as a biomarker to detect tumor response using urine or tissue samples, as plasma levels of the mutation in cell-free DNA have correlated with tumor response to treatment.^6^

Complete surgical excision of UC is uncommon and reserved for tumors localized to the apex of the bladder. Dogs with UC localized to the trigone are poor candidates for aggressive surgical resection, especially with ureter or urethral involvement. Difficulties achieving complete surgical margins result in tumor recurrence between 3 and 10 months.^7^ Other surgical treatment options are considered palliative, as they do not significantly increase overall survival.^8^ Palliative treatment of obstructive UC includes stenting of the urethra or ureters, subcutaneous ureteral bypass, long-term catheter placement and nonsteroidal anti-inflammatory drugs (NSAIDs) protocols.^9^

Piroxicam alone achieves a complete response in 17.6% of dogs and is well tolerated.^10^ When NSAIDs are used in conjunction with chemotherapy, remission increases to 58.3%.^11^ Doxorubicin chemotherapy protocols are well tolerated but results in modest progression-free survival times, and lower response rate than piroxicam alone.^12^ Currently, the mainstay of medical treatment involves a combination of piroxicam and mitoxantrone, vinblastine, or carboplatin.^11,13,14^ Urinary outflow obstruction secondary to the tumor is managed with locoregional tumor control, with the therapeutic aim of reducing tumor volume, and restoring urinary patency. Locoregional control of UC is best achieved with a combination of chemotherapy, NSAIDs, and radiation therapy (RT), as multimodal therapy has shown synergistic effects that improve overall treatment outcomes.^1^

The objective of this study was to report the outcome of dogs presenting with urethral or ureteral obstruction due to UC treated with RT. A secondary objective was to identify factors that can predict restoration of patency or result in longer survival times.

Materials and methods

Case selection

Dogs were included in this analysis if they were diagnosed with suspected or confirmed UC and obstruction of urinary outflow and were treated with RT at the Flint Animal Cancer Center at Colorado State University between February 2006 and March of 2023. A subset of these dogs were included in previous studies.^15,16^ Medical records were reviewed and clinical history, demographics, presenting complaint and clinical signs upon presentation were recorded as well as staging, including complete blood count, serum biochemistry, urinalysis, and imaging such as thoracic radiographs, abdominal ultrasonography, and computed tomography (CT) if available.

Dogs were included if they had clinical signs of blockage including stranguria, anuria, pollakiuria overflow incontinence or a combination of these, as well as the denotation of complete or partial blockage via imaging including cystoscopy, abdominal ultrasonography, and CT. Inclusion in this study also required treatment with radiotherapy before surgical intervention to relieve obstruction.

Urinary blockage was classified as either urethral or ureteral based upon cystoscopy, abdominal ultrasonography, CT evaluation, or a combination of these modalities. Dogs were grouped by location of their urinary outflow obstruction with the categories being urethral, ureteral, or both urethral and ureteral. Dogs were also grouped by radiotherapy treatment total dose and treatment intent. Radiotherapy protocols of less than 4 Gy per fraction were considered definitive intent, and those with 4 Gy or more per fraction were considered palliative intent protocols.

Treatment information was gathered including chemotherapeutic protocol, concurrent NSAID, and surgical intervention. Adverse radiation effects were retrospectively graded from medical records according to the Veterinary Radiation Therapy Oncology Group (VRTOG) toxicity grading scheme version 2 with acute effects occurring before 90 days and late effects occurring after 90 days.^17,18^ Radiotherapy treatment intent, as well as total dose, dose per fraction, and number of fractions was recorded. In all cases, radiotherapy was delivered using a Varian Trilogy Linear Accelerator (Varian Medical Systems, Inc., Palo Alto, CA).

For both definitive and palliative intent treatment, treatment planning and dose were not standardized, and no standard follow-up protocol was set after treatment concluded.

Radiation therapy

Computed tomography scans for radiation planning were obtained with an immobilizing cushion to ensure exact positioning upon treatment. Computed tomography was performed using either a Picker PQ2000 CT single slice helical scanner (Picker Medical Systems, Cleveland, OH), a Philips Gemini TF Big Bore 16-slice scanner (Philips Medical Systems, Nederland, B.V.), or Siemens Somaton Force 128-slice scanner (Siemens Medical Solutions, PA). Both a non-contrast and contrast helical dataset was obtained through the pelvis for radiation planning. Omnipaque 350 (GE Healthcare, Princeton, New Jersey) contrast media was administered IV (2 mL/kg) before the post-contrast series. Images were then reconstructed at 2-mm continuous intervals with a 512 × 512 matrix.

The pre- and post-contrast series were imported into the treatment planning software and used for contouring with the plan created on the pre-contrast series. Gross tumor volume (GTV) was determined by delineating the grossly evident tumor from the CT scan or cystoscopy if available. The clinical target volume (CTV) was defined as the entire bladder plus 2 cm down the urethra or distal to the GTV if within the urethra to account for microscopic extension of disease. Planning target volume (PTV) was determined by adding an isotropic 0.3 or 0.5 cm to the CTV to account for variations in target position, shape, and size. Contoured organs at risk included colon, femoral head, cauda equina, and kidneys. One dog had sublumbar lymph nodes contoured due to suspected metastasis. The CT scan was utilized to create intensity-modulated RT (IMRT) and three-dimensional conformal radiation therapy (3D-CRT) plans. These plans were created using Varian Eclipse treatment planning software (versions 8, 10, 11, 13, or 15) utilizing the AAA or Acuros algorithm.

Intensity modulated radiation therapy plans were created using coplanar, isocentrically placed 6 MV dynamic multileaf collimation in a sliding-window technique with static gantry or coplanar 6 or 10 MV volumetric arc therapy with sliding-window technique. Three-dimensional conformal radiation therapy plans were created with coplanar static beams and multileaf collimation or wedges to aid in dosimetry goals. The goal of the computer-based plans was to deliver 95% of the prescription to 100% of the PTV.

Quality assurance (QA) was measured with γ analysis using the portal dosimetry system equipped on the linear accelerator for each field or arc for IMRT plans. A passing score required a minimum of 95% γ for a 3 mm distance to agreement and a 3% absolute dose difference. No individual plan QA was required or measured for 3D-CRT or manually calculated plans.

The conformity index (CI) was calculated retrospectively for all computer-based plans. Conformity index (CI) was calculated as CI = (TVPIV)^2^/TV × PIV, where TVPIV is the volume of the tumor (PTV) receiving 100% of the prescription dose, TV is the target volume, and PIV is the volume of the body receiving 100% of the prescription dose. Dosimetry parameters by treatment intent and modality are found in Table 1.

Manually calculated treatment plans were generated using diagnostic abdominal radiographs and verified with an MV port film before treatment. Beams were delivered in a parallel opposed approach with dogs positioned in lateral recumbency. The goal of treatment was to include the bladder with most of the urethra in the field. Manual calculations were performed under the discretion of a board-certified American College of Veterinary Radiology (ACVR) radiation oncologist.

Follow-up

Time to relief of obstruction was determined from medical records, including owner reporting relief of clinical signs between treatments (via communication logs), normalization of chemistry values consistent with post-renal azotemia, as well as communication with owners and clinical signs reported at recheck appointments. Medical records were reviewed and follow-up information was collected from owners and referring veterinarians including events post RT and date and cause of death. Dogs were censored if lost to follow-up or alive at the time of analysis. There was insufficient data to evaluate tumor control using Response Evaluation Criteria in Solid Tumors (RECIST).^19^

Statistical analysis

Survival time was calculated as time from first fraction of RT to time to obstruction relief, first event (time to first event [TFE]) or death (overall survival time [OST]). Events were defined as local disease progression, recurrence of clinical signs, metastatic disease or death from any cause. Dogs alive at the time of analysis were censored from data analysis. Kaplan–Meier survival curves were created, and median survival time (MST), median time to event, and median time to relief of obstruction were determined. A Mantel–Cox log-rank test was used to evaluate the impact of tumor location, and treatment intent on survival and obstruction duration. Comparison of dose parameters by outcome was analyzed using unpaired 2 tailed t-tests. Chi-square analysis was used to determine the effect of location on obstruction relief. Statistical significance was determined using a P-value of .05. All statistical analysis were generated using PRISM v.10.1.1 (GraphPad, San Diego, CA).

Results

During the study, 179 dogs were treated with RT for UC. Of those, 31 dogs met the inclusion criteria. Six dogs were excluded due to surgical intervention before RT resulting in 25 dogs for analysis in this study. Dogs included 16 (64%) females, 1 of which was intact and 9 (36%) castrated males dogs. The median age of dogs was 10 (7-15) years of age, and median weight was 15.4 (2.93-58.1) kg.

Dogs initially presented for anuria (n = 10), stranguria (n = 6), hematuria (n = 2), pollakiuria (n = 3), or progression of previously diagnosed UC unresponsive to medication (n = 4). The median duration of clinical signs before presentation for RT was 16 days (2-90). Eleven dogs (44%) presented with azotemia before RT. Of dogs with azotemia, median creatinine was 2.4 mg/dL (1.77-4.33 mg/dL), median urine-specific gravity was 1.009 (1.007-1.016), and 4 (36%) were proteinuric.

Eleven dogs (44%) had a temporary urinary catheter placed before, or at the time of RT, that was in place during initiation of treatment. Of male dogs, 4 (44%) had prostate involvement, one of which resulted in external compressive obstruction of the urethra. Eleven (34%) dogs had evidence of metastases at the time of treatment including local lymph nodes (n = 6), lung (n = 5), and proximal femur (n = 1) that was confirmed or suspected on imaging.

Twenty-four dogs (96%) had abdominal or focal urinary ultrasonography for staging. Fourteen dogs (56%) had a mass visualized with ultrasonography. Of dogs without a visible mass on ultrasonography, 7 (29%) had a mass visualized with CT, 3 (12.5%) had a mass visualized with cystoscopy (1 of these was also seen on CT), and 1 (4%) was diagnosed at the referring primary care veterinarian by an undocumented modality. Eleven dogs (44%) had a CT scan and 11 (44%) had radiographs of the abdomen and pelvis for RT planning. One dog (4%) received only a cone beam CT before RT for 3D-CRT planning. Three (12%) had CT of the thorax while 6 (24%) had chest radiographs. Eleven (39%) dogs had cystoscopy performed and 7 (28%) had a cystoscopic tumor biopsy. Five biopsies (71%) confirmed carcinoma and 2 were nondiagnostic. Nine dogs (36%) were tested for BRAF and 7 (78% of dogs tested, 28% of total cohort) were positive. The 2 dogs who tested negative for BRAF had confirmation of UC via cytology. One dog had a confirmed diagnosis from the rDVM, but it was unclear from records how this diagnosis was achieved. Three (12%) dogs did not have confirmation of UC and had a presumptive diagnosis.

Eighteen (72%) dogs, regardless of treatment intent, received NSAIDs before or concurrent to RT. Fifteen (60%) dogs received chemotherapy in addition to RT. Eleven received chemotherapy during RT, 2 dogs received neoadjuvant chemotherapy before RT, and 2 received adjuvant chemotherapy after RT completion. All dogs who had concurrent chemotherapy during RT received mitoxantrone. Eight dogs (32%) received single agent mitoxantrone at a dose of 5 (n = 5), or 5.5 (n = 3) mg/m^2^ for a median of 2 doses (1-5). One dog received 2 doses of mitoxantrone at 5.5 mg/m^2^, followed by 1 dose of carboplatin at 300 mg/m^2^. One dog received 5 doses of 5 mg/m^2^ mitoxantrone, followed by 4 doses of 2 mg/m^2^ of vinblastine. One dog received 6 doses of 4.5 mg/m^2^ mitoxantrone, followed by 5 doses of 1.8 mg/m^2^ vinblastine, and 6 doses of 227 mg/m^2^ carboplatin. Of the dogs who received neoadjuvant chemotherapy before RT, 1 dog received 2 doses of 2 mg/m^2^ vinblastine and 1 unknown dose of carboplatin, and the other received 3 doses of vinblastine at 2.2 mg/m^2^. Of the 2 dogs who received adjuvant chemotherapy after RT, both received single-agent vinblastine at 16 doses of 2.25 mg/m^2^ and 4 doses of 2 mg/m^2^.

Six dogs (24%) received definitive intent IMRT, and 19 (76%) received palliative intent RT. Of the 19 dogs who received palliative intent RT, 1 (5%) had IMRT, 5 (26%) had 3D-CRT, and 13 (68%) had manual setup. Of the 6 dogs who had had definitive intent radiation, 2 (33%) received only chemotherapy, 3 (50%) were treated with both NSAID and chemotherapy, and 1 (16%) had no medications at time of RT. Of the 19 dogs who had palliative intent radiation, 3 (16%) had only chemotherapy, 7 (37%) had only NSAIDs, 8 (42%) had chemotherapy and NSAIDs, and 1 (5%) had no other treatment at time of RT.

One dog receiving definitive intent treatment (17%) received 1 dose of 5.4 Gy, followed by 11 fractions of 2.85 Gy for a total dose of 36.75 Gy. Three (16%) palliative intent dogs received a clinical setup protocol of 1 fraction of 8 Gy, followed by 4 weekly fractions of 6 Gy for a total dose of 32 Gy. Three (16%) palliative intent dogs were treated with a “Quad Shot” protocol of 4 Gy × 4 fractions over 2 consecutive days. Of these dogs 2 received one round of this protocol and 1 dog repeated this protocol one month apart for a total of 3 sessions.^20,21^

Of all dogs included in the study, 15 (60%) were successfully unobstructed with RT (Figure 1a). While 77% of BRAF + dogs unobstructed, BRAF testing was not a significant indicator (P = .32) of obstruction relief from radiotherapy (Figure 1b). Of dogs with pretreatment azotemia, 5 (45%) were successfully unobstructed and 2 (40%) of those had resolution of azotemia on post-RT urinalysis. Relief was temporary for 4 dogs (27%) who experienced obstruction recurrence at a median of 14 days (5-33 days) after initiation of RT. All 4 dogs received palliative intent radiation, with a median total dose of 16 Gy (range 12-32 Gy). Two of these dogs went on to surgical stent placement and 2 were euthanized.

Definitive intent protocols successfully relieved obstruction in 83% of dogs, while palliative protocols successfully relieved obstruction in 53% of dogs (Figure 1c and d). All 3 dogs which received quad shot RT and all 3 who received 8 Gy × 1 followed by 6 Gy × 4 experienced obstruction relief. Of these dogs, one that received one round of quad shot and one that received 8 Gy × 1 followed by 6 Gy × 4 did not experience durable relief. Dogs receiving definitive intent protocols that lead to obstruction relief had a significantly longer median obstruction time of 8 days (2-10 days) compared to 4 days (2-7 days) with palliative protocols (P = .02). On its own, quad shot therapy relieved obstructed dogs with a median of 3 days (2-3 days). Of dogs who unobstructed, those who received palliative intent radiation with concurrent chemotherapy unobstructed significantly sooner compared with definitive intent protocols with chemotherapy (2 days vs 7 days; P = .04; Figure 2a).

Of the 10 dogs that did not have relief of obstruction with RT, 5 underwent surgical intervention to relieve obstruction, 4 were humanely euthanized before completion of RT, and 1 pursued hospice care at completion of RT. A total of 7 dogs (28%) received surgical intervention after RT due to unsuccessful relief of obstruction (n = 5) or recurrence of obstruction (n = 2). Of these dogs, 6 (86%) pursued stent and 1 (14%) pursued cystostomy tube. Of dogs that were humanely euthanized or pursued surgical intervention before completion of treatment, 3 (33%) discontinued after 1 fraction (median 8 Gy), 1 (14%) discontinued after 2 fractions of 8 Gy, and 1 dog (14%) discontinued after 4 fractions of 6 Gy. Surgical intervention was pursued a median of 5 days (range 2-7 days) post initiation of RT, and euthanasia was pursued a median of 9 days (range 8-10 days) after start of RT. All dogs with relief of obstruction completed the full course of radiation regardless of protocol.

Dogs with relief of obstruction regardless of RT treatment intent received a significantly lower median dose per fraction with a median of 4 Gy (2.7-6 Gy) compared to the median dose per fraction of 8 Gy (4-8 Gy) for dogs that remained obstructed (P = .005; Figure 3a). The median total dose of all dogs that unobstructed was 36 Gy (12-57 Gy), which was significantly higher than the median total dose of 8 Gy (6-24 Gy) given to dogs that remained obstructed (P < .001; Figure 3b).

Twelve dogs had urethral obstruction, 6 (50%) of which had relief of obstruction. Nine dogs had ureter obstruction, 6 (66.6%) of which were relieved. Four dogs had obstruction to both the urethra and ureters, 3 (75%) of which were relieved (Figure 4a). Location of obstruction (urethral vs ureteral) had no significant association with overall survival (Figure 4b), event-free (Figure 4c) survival time, or with the duration of obstruction (Figure 4d).

Acute radiation toxicity occurred in 14 (56%) dogs. All dogs who experienced acute toxicoses were associated with the skin, gastrointestinal (GI) tract, and bladder. All dogs (n = 6) receiving definitive intent radiation developed acute toxicoses, while 8 (42%) dogs receiving palliative intent radiation developed acute radiation toxicoses. Of the acute toxicoses, 9 (36%) dogs developed diarrhea a median of 15 days (4-41 days) after the start of radiation. Three (12%) dogs developed grade 3 GI toxicities at a median of 21 days, consisting of progressive diarrhea containing frank blood that was managed on an outpatient basis. Of the 9 dogs which developed diarrhea, 6 were receiving chemotherapy at the time of signs. All 3 dogs who experienced grade 3 diarrhea received definitive intent RT and concurrent chemotherapy. Due to concurrent radiation and chemotherapy treatment, these toxicoses were attributed as possible secondary to radiation. Radiation effects to GI for dogs not receiving concurrent chemotherapy were limited to grade 1 and attributed as probable or definitive. Two dogs (8%), 1 receiving definitive intent and 1 receiving palliative intent RT, developed acute grade 3 skin toxicoses consisting of moist desquamation of the vulva at 21 and 17 days, respectively. Three dogs (12%) developed acute grade 3 urogenital toxicoses. Of these dogs, 1 received definitive intent treatment and developed hematuria at 7 days and 2 palliative intent dogs developed hematuria and pollakiuria at 9 and 10 days (Table 2).

Four dogs (16%) developed late radiation effects. Three dogs (12%) developed suspected fibrosis and stricture of the urinary tract, 2 at the level of the ureter, and 1 at the level of urethra a median of 180 (167-260) days after treatment. Fibrosis was suspected from ultrasonographic findings of urinary bladder and urethral wall thickening inconsistent with progressive neoplasia. Fibrosis was grade 3 in all dogs and attributed as possible to probable secondary to radiation given the lack of concern for progressive disease based on imaging. Of these dogs, 2 received palliative intent treatment, and 1 received definitive intent treatment. One dog, who received palliative intent treatment, developed suspected cystitis secondary to radiation at 157 days. Cystitis was a diagnosis of exclusion from a history of stranguria, presence of white blood cells without bacteria on urinalysis, and ultrasonographically detected bladder wall thickening not associated with the urethral tumor; however, tumor progression could not be definitively ruled out. Three of 4 dogs (75%) who experienced late radiation toxicoses of the urinary system also developed acute urogenital effects. Although not considered a radiation toxicoses, 4 dogs developed urinary tract infections during or after treatment.

The median OST for all dogs was 134 days (range 6-969 days) and the median TFE was 107 days (range 4-695 days). Euthanasia was elected due to progression of local disease in 9 dogs; of these dogs 1 received definitive intent treatment. Two dogs (8%) were euthanized secondary to metastasis. Four dogs (16%) ultimately succumbed to diseases that cannot be directly attributed to neoplasia including pneumonia (n = 2), GI disease (n = 1), and cardiac disease (n = 1). The remaining dogs did not have recorded cause of death/euthanasia.

Definitive intent protocols with adjuvant chemotherapy had significantly longer OST in comparison to palliative intent treatments without chemotherapy (270 days vs 25 days; P = .01; Figure 1b). Palliative intent protocols with chemotherapy had a significantly longer OST than those without chemotherapy (135 days vs 25 days; P = .05). There were no significant differences between TFE between treatment intents with and without chemotherapy. Median TFE for dogs receiving multimodal definitive intent and palliative intent therapy was 139 and 70 days, respectively (P = .22). Dogs with confirmed or suspected metastases to the lungs had an MST of 20 days (5-182 days), which was significantly lower when compared to dogs with no metastasis (135 days; P = .03). No significant differences were observed between dogs with suspected or confirmed metastasis to lymph nodes and those without.

Discussion

Radiotherapy is a viable option for relieving urinary obstruction with an overall 60% success rate. Choice in protocol is dependent on clinician and owner goals regarding tumor control and palliative care. Definitive intent protocols had a higher overall success rate in relieving obstruction, and increased survival time especially when combined with chemotherapeutics; however, these dogs were obstructed for longer and often needed hospitalization during their treatment with concurrent urinary catheter placement. Median TFE and MST for dogs receiving multimodal definitive intent protocols were 139 and 270 days respectfully. These survival times are less than previously reported event-free survival times of 260-317 days, and OST of 510-654 days in multimodal definitive intent protocols.^15,16^ The discrepancy in survival times is likely due to patient selection, as dogs who present with urinary obstruction have a more guarded prognosis with disease in more compromising locations.^22,23^

Location of obstruction (urethral vs ureteral) had no relevant effect on TFE or OST, nor did it affect if the dog would unobstruct. A previous IMRT study on dogs with urogenital carcinoma showed that location of the primary tumor had a significant effect on local tumor control or OST; however, this study primarily included dogs with patent urinary systems and did not factor presence of urinary obstruction into local tumor control and survival time.^16^ In a number of cases, there was a lack of identification of tumor on ultrasonography particularly within the distal urethra yet seen on CT, cystoscopy, or a combination of these modalities. Ultrasound waves are completely reflected by high-density tissues, such as the pelvic bone, creating an acoustic shadow deep to the bone where the urethra is located.^24^ This natural artifact can be beneficial in many circumstances; however, here presents a diagnostic challenge when trying to treat a specific area within the urethra. Should the lesion be difficult to identify, it might be necessary to pursue a CT or cystoscopy to further locate the area of obstruction before RT to assure that the appropriate area is being included. If further diagnostics are not possible, RT to the entire urogenital tract might be warranted so as not to miss the key region.

In all dogs, higher doses per fraction with lower total dose, variables associated with palliative intent protocols, had a significant association with shorter time to relief of obstruction; however, these dogs had a decreased TFE and OST. Palliative intent protocols relieved obstruction faster, but a higher percentage of these dogs did not improve enough clinically and were euthanized before treatment completion or required surgical interventions. In contrast, definitive intent protocols with lower dose per fraction and higher total dose had higher success rates in relief of obstruction, with longer TFE and OST, but longer times to relieve the obstruction with a median of 8 days. An earlier study with a protocol of 10 daily fractions of 2.7 Gy in 3 dogs had urethral patency restored after 3, 5, and 8 days, and one dog had ureteral patency restored at 6 week recheck on ultrasonography; however, the protocol was defined as palliative intent.^25^ Dogs in this study that received palliative intent RT experienced relief of urinary obstruction at a median of 4 days. The quad shot protocol performed in 3 dogs in the current study, provided obstruction relief in all 3 dogs in a median of 3 days. This protocol delivers 16 Gy in a 48-h period and, if all 3 rounds are delivered, delivers a total dose of 48 Gy. As palliative intent protocols result in more rapid relief of obstruction, and definitive intent protocols provide a higher relief efficacy with prolonged survival time, a protocol that combines aspects of both protocols might provide the most optimal results. With the quad shot protocol, the optimal long-term control is likely to occur when all 3 rounds are delivered; however, delivering the first round might give the rapid relief needed for owners to make further decisions regarding continued therapy or moving to hospice care.

Palliative and definitive intent treatment protocols demonstrated significantly reduced obstruction times with concurrent chemotherapy, suggesting that integration should be considered to enhance therapeutic outcomes. Combination therapy with surgical intervention, such as urinary stent, followed by radiation therapy is not well studied. Dogs that received surgical intervention before RT were excluded and, therefore, no conclusions could be drawn regarding the integration of this modality in this study. One advantage of avoiding surgical intervention for urinary obstruction is fewer quality of life adverse events (AEs). In one study on temporary urethral stents in dogs, 75% of dogs experienced urinary tract infections and 94% of dogs experienced urinary incontinence where diapering was recommended.^26^ Here, 16% of dogs developed urinary tract infections and there were no reports of urinary incontinence. While RT isn’t without possible AEs, most resolved and few required lifestyle changes. If possible, starting with RT to determine if dogs can be successfully unobstructed before moving to surgical intervention might lead to fewer long-term AEs. Future studies are needed to determine the most effective multimodal treatment, integrating medical, surgical, and radiation therapies to achieve rapid obstruction relief and prevent tumor progression.

Carcinoma cells die secondary to ionizing radiation via mitotic catastrophe.^27^ This process results in latent gross tumor shrinkage that takes days to weeks after RT depending on mitotic phase during treatment.^28^ Both definitive and palliative intent protocols in this study resulted in relief of obstruction, and therefore presumed gross tumor shrinkage, at median intervals of 8 and 4 days, respectively. Time to gross tumor shrinkage could be due to different cellular pathways resulting in cell death or inflammation secondary to radiation. The cumulative effects of NSAID, chemotherapy, or a combination of these modalities during treatment might also play a role in this timeline. The role of gross tumor shrinkage in comparison to decrease in urinary tract inflammation in relieving obstruction has not been characterized. Biopsy to determine the percent of inflammatory cells in relation to carcinoma cells might provide a more objective methodology to distinguish this difference.^29^

Mild acute radiation adverse effects (grades 1 and 2) occurred in 56% of dogs and were primarily localized to the skin and GI tract. This severity and location of signs is consistent with previous studies characterizing side effects of definitive intent radiation therapy.^15,16^ Of the 9 dogs who experienced acute radiation effects to the GI tract, 8 received concurrent chemotherapy, which likely contributed to the high percentage of acute effects.^30^ Owners should be counseled on the possibility of radiation AEs and careful consideration and comparison of radiation AEs and AEs associated with concurrent chemotherapy, stent, or surgical procedures should be discussed.

Limitations of this study revolve around its retrospective nature. Radiation protocols and follow-up were not standardized within treatment intent. Data obtained were limited by reporting of clinicians, owners, and referring DVMs. Because of this limitation, time to relief of obstruction might have been over- or underestimated due to inconsistent reporting and inconsistent documentation.

For radiation oncologists treating urinary obstructed dogs due to UC, it is important to relay to owners the more guarded prognosis with these dogs compared to those presenting without obstruction. From these data, the suggested clinical approach would be to use the full protocol quad shot (3 monthly rounds) with concurrent chemotherapy. A CT scan at the time of cystoscopy, if possible, would be ideal to understand the location of the obstruction as it is not always evident on ultrasonography. Without a CT, inclusion of the entirety of the urogenital tract should be considered if ultrasonography cannot definitively locate the obstructive lesion. With this approach, it is anticipated that dogs will be hospitalized for 2-3 days for obstruction with discharge occurring near the time of completion of the first round of quad shot.