Demographics, staging, and clinical management

A single institution retrospective study was performed at a university-based veterinary teaching hospital in North America. Institutional animal care and use committee approval was not required. Oncology accession logs were searched for dogs with intranasal masses that had been treated with external beam RT between August 2013 and November 2020. Specifically, they had to have been treated using our institution’s standard 10 Gy X 3 SRT protocol.² Dogs were excluded if they had a lymphomatous mass.

For each included animal, standard written pet owner consent for treatment was obtained; clinical management was at the discretion of the medical team managing the individual dog at the time. Institutional norms dictate that within the month preceding SRT for an intranasal tumor, dogs be evaluated via the following: (a) physical examination; (b) CBC and serum biochemical profile; (c) pre- and postcontrast CT of the head and neck to determine the extent of the primary mass and evaluate regional lymph nodes; (d) fine-needle aspiration and cytology of the bilateral mandibular lymph nodes, with ultrasound-guided aspiration and cytology of the medial retropharyngeal lymph nodes if they are abnormal on CT; and (e) either 3-view thoracic radiographs or thoracic CT. It is also typical to offer urinalysis, and abdominal cavity imaging with ultrasound or CT, to evaluate comorbidities. At our institution, tumors are clinically staged according to previously published criteria (ie, modified Adams staging).^3,9 During the study period, our default treatment of choice for intranasal masses was this 10 Gy X 3 SRT protocol. Animals were generally treated on 3 consecutive business days but were permitted to span holidays and weekends as long as the scheduled break was no longer than 3 days. Treatments were intentionally split and given every other day as a strategy intended to reduce risk of fistula formation in patients where there was concern, based on physical examination, that the overlying skin or mucosa was thin or otherwise unhealthy. Full-course conventionally fractionated RT (eg, 42 Gy total in 10 daily fractions or 54 Gy total in 18 daily fractions) was considered or recommended when the tumor had extension outside of the nasal cavity and lacked conspicuity (as determined by the attending radiation oncologist). Less intensive palliative-intent irradiation protocols (eg, 6 to 8 Gy X 4 once weekly fractions or 4 Gy X 5 daily fractions) were generally reserved for patients where finances did not permit use of either SRT or full-course RT.

For this study, tumor stage was condensed into 3 groups: no calvarial involvement (T_1–3), cribriform lysis (T_4a), and tomographically evident intracalvarial mass effect (T_4b).³

Radiotherapy planning and delivery

Dogs were anesthetized for CT simulation and SRT. Anesthetic protocols varied but typically included an opioid premedication followed by propofol induction and maintenance with isoflurane gas. Radiation targets were delineated, and treatments were planned and delivered as previously described.²

Clinical follow-up

During the study period, follow-up examinations, ideally with the radiation oncology team at our institution, were recommended 2 to 3 weeks after treatment and then quarterly, with CT scans recommended at approximately 120-day intervals. Additional follow-up information for the included animals was also obtained at 90- to 120-day intervals via phone or email communication with at least one of the following: the pet owners, referring specialists, or primary care veterinarians.

Data abstracting

To facilitate data extraction from medical records, a data-abstracting spreadsheet was created; an anonymized copy of the complete spreadsheet has been deposited in the Harvard Dataverse.¹⁰ The following data were included: breed, sex, neuter status, date of birth, body weight at diagnosis, approximate date of first clinical signs, presenting clinical signs, date of diagnosis, whether rhinoscopy was performed, how the diagnosis was confirmed (histopathology, cytology, or presumptive imaging diagnosis), tumor type, mitotic index (when reported in the histopathology report), laterality of the tumor, whether or not there was bone lysis or extranasal extension of disease or cribriform lysis, status of lymph nodes (based on palpation, CT, and cytology), results of pre-SRT thoracic imaging, Hct, neutrophil and lymphocyte counts, treatments before and after SRT (anti-inflammatory drugs, radiotherapy, chemotherapy, surgery), types of comorbid conditions, overall survival data (date of death or last follow-up, and reason for death), results of necropsy (if applicable), clinical response to SRT (with regard to how quickly, completely, and durably clinical signs changed after SRT), whether follow-up CT was performed after SRT (and results of such testing), information about radiation toxicoses (acute and late), and results of any restaging tests that were performed. Data related to radiation dosimetry were also collected per published guidances.^11,12

To account for breed-related differences in longevity, dogs were stratified into life stages using categories based on American Animal Hospital Association guidelines¹³ combined with average breed life span published on the American Kennel Club website.¹⁴ The categories utilized were mature adult (n = 50 dogs), senior (51), and end of life (28). Briefly, mature adulthood spans the time from completion of physical and social maturation until the last quarter of the estimated life span, the senior life stage includes the last 25% of the estimated life span, and end of life is defined as the terminal stage, when animals are alive past the average estimated life span for their size and breed.¹³

Statistical analysis

Overall survival time was defined as the time from the first fraction of SRT (10 Gy X 3), until death. When animals were lost to follow-up or alive at the time of analysis (December 2021), they were censored at the date of last contact. Event-free survival time was defined as the time from the first fraction of SRT until an event occurred. An event was defined as local, regional, or distant progression of the nasal tumor. When available, local progression was based on CT findings and assessed per the Veterinary Cooperative Oncology Group’s response evaluation criteria for solid tumors in dogs¹⁵; otherwise, it was based upon progression of clinical signs. Subjects were censored at the time of death if they died due to non-nasal tumor-related causes; they were also censored if they were alive without evidence of tumor progression at the time of data analysis. Survival analyses were performed using the Kaplan-Meier product-limit method. To determine whether any patient, tumor, or dosimetric features were predictive of outcome, log-rank analysis was performed for categorical variables, and univariate Cox proportional hazards regression analysis was used for both categorical and continuous variables. The variables we assessed are listed in the relevant tables in the present report. Factors with P < 0.20 on univariable analysis were moved into a multivariable Cox proportional hazards regression model. Backward selection was applied with the predictor, with the largest Wald P value being considered for removal at each stage. Reported results reflect P values for the Wald statistic and unit hazard ratio. All statistical analyses were performed using commercial software (JMP Pro version 15.2.0; SAS Institute Inc; and GraphPad Prism version 6; GraphPad Software Inc). The threshold for statistical significance was set at P < 0.05.

Patient demographics

During the study period (August 2013 to November 2020), 180 dogs with intranasal mass lesions were treated with external beam radiotherapy. Fifty-one dogs were excluded from this study because they had lymphomatous tumors (n = 3) or were treated with a protocol other than 10 Gy X 3 SRT (48). In total, 129 dogs met criteria for inclusion in this study. Data related to twenty-nine of these dogs were previously reported.² There were 67 castrated males, 3 sexually intact males, and 59 spayed females. Mean body weight was 24.2 kg (SD, 11.8 kg; range, 3.1 to 54.6 kg). The most common types of dogs were as follows: mixed breed (n = 18), Labrador Retriever (16), Golden Retriever (14), German Shepherd Dog (6), Border Collie (4), Siberian Husky (4), and 3 each of Beagle, Boxer, Dachshund, Jack Russell Terrier, Maltese, Pug, West Highland White Terrier. Clinical signs included epistaxis (n = 104), sneezing (88), nonhemorrhagic nasal discharge (53), facial deformity (19), seizures (7) and other neurologic signs including gait abnormalities or vestibular signs (4), head tilt (3), changes in behavior or mentation (1), or facial nerve dysfunction (1).

The first fraction of SRT was administered an average of 175 days after clinical signs were reported by the family to have been initially identified (SD, 214 days; range, 26 to 1,781 days). The average age at the time of SRT was 9.8 years (SD, 2.5 years; range, 4.0 to 14.8 years). Clinical data as a function of life stage were compiled (Table 1). Subjectively, the dogs treated earlier in life were more likely to have received other tumor-directed therapies (eg, chemotherapy, radiotherapy, or both) either before or after SRT, and such therapies are detailed below. Radiation dosimetry was similar for the 3 life stages. The mean value (and SD) for the near-maximum dose (D_2%) in the planning target volume (PTV) was 39.0 ± 1.4 Gy for mature adults, 39.5 ± 1.9 Gy for senior dogs, and 38.2 ± 3.9 Gy for those at end of life. The near-minimum dose (D_98%) in the gross tumor volume (GTV) was 30.3 ± 2.3 Gy for mature adults, 30.0 ± 2.0 Gy for seniors, and 29.4 ± 2.4 Gy for those at end of life. Tumor size was also similar; the ratio of GTV to body weight was 3.0 cm³/kg (cm³/kg) ± 4.8 cm³/kg for mature adults, 2.5 ± 1.3 cm³/kg for seniors, and 2.4 ± 1.7 cm³/kg for those at end of life.

Pathologic diagnoses were confirmed cytologically (n = 4) or histopathologically (116). In 9 dogs, an imaging diagnosis was made because either tumor tissue samples were not obtained for pathologic assessment (n = 1) or the pathologic diagnosis was inconclusive (8). Tumor stage was defined via assessment of postcontrast CT in all included animals. Of the 129 dogs, 83 had carcinomas, 30 had sarcomas, and 16 had other nonlymphomatous intranasal tumors; additional tumor staging and histopathology data are summarized (Supplementary Table S1).

Mandibular lymphadenomegaly was reported on initial (pretreatment) physical examination in 41 dogs; on the simulation CT scan, either mandibular or medial retropharyngeal lymphadenomegaly was identified in 72 dogs. Fine-needle aspiration and cytology of the regional lymph nodes was performed in 110 dogs, including the ipsilateral mandibular lymph node (n = 29), bilateral mandibular lymph nodes (73), or a combination of mandibular and medial retropharyngeal lymph nodes (8). Metastasis was cytologically confirmed in 9 dogs; in all 9 instances, the dogs had lymphadenomegaly identified on CT, and in 7 of those 9, they also had palpable lymphadenomegaly. The other 101 samples had cytologically normal or reactive lymph nodes.

Evaluation for distant metastases was performed using radiographs in 52 dogs and CT in 72; no thoracic imaging was performed in 5 dogs. An imaging diagnosis of intrathoracic (pulmonary) metastasis was made in 2 dogs. In this study, we did not record whether screening tests were performed to evaluate for metastasis or comorbid disease in other anatomic locations (eg, via abdominal ultrasound or CT).

Results of a CBC performed within the month preceding SRT were available for review in 126 dogs. The mean (± SD) Hct was 45.7% (± 6.3%; reference range, 40.2% to 60.2%). The mean neutrophil count was 10,257/µL (± 5,692/µL; reference range, 2,841 to 9,112/µL). The mean lymphocyte count was 1,533/µL (± 1,058/µL; reference range, 594 to 3,305/µL). A neutrophil-to-lymphocyte ratio was calculated, and the mean value was 10.6 (± 12.8).

Tumor-directed therapies given before SRT

A variety of tumor-directed therapies (surgical, radiotherapeutic, chemotherapeutic, or a combination thereof) were used before SRT. In many instances, referral records were sparse, with missing details particularly related to rationale for each given intervention (eg, was chemotherapy administered after SRT for potential additive effects or to manage progressive clinical signs or progressive disease?); nonetheless, the available data are summarized (Supplementary Table S2).

Systemic antineoplastic drug therapy had been given before SRT in 7 dogs (carboplatin, n = 3; toceranib phosphate, 3; masitinib mesylate, 1); there was inconsistent reporting of when and why these medications were stopped relative to the timing of SRT. Beyond taking small biopsies for diagnosis, a total of 7 dogs had undergone nasal surgery before referral for radiotherapy; this included biopsy, exenteration or debulking via rhinotomy or sinusotomy (n = 6), debulking and exenteration via rhinotomy (3), and debulking via rhinoscopy (1). Prior to SRT at our institution, 5 dogs had previously undergone an initial course of external beam radiotherapy for their nasal tumor; in each instance, they had received a full-course definitive-intent protocol, involving delivery of 48 to 60.2 Gy total in 16 to 19 daily (Monday through Friday) fractions. One additional dog had previously undergone a course of 54 Gy, given in daily 3-Gy fractions via intensity-modulated radiotherapy for an incompletely excised high-grade maxillary fibrosarcoma; 10 months after completing that set of treatments, the dog developed a nasal carcinoma that was treated with SRT given over 5 days (every-other-day treatment schedule). Additionally, at the time of SRT, 71 dogs were taking an orally administered NSAID; in 66 dogs, it had been prescribed for the nasal tumor, and in 5 instances, the dog had been taking it prior to the nasal tumor diagnosis, to manage comorbid disease (eg, osteoarthritis). It is beyond the scope of this report to provide a detailed reporting of other comorbid conditions and concomitant medications that were used before, during, or after SRT.

Stereotactic radiotherapy

SRT was administered on 3 consecutive days (96 dogs) or over 5 (23 dogs), 6 (4 dogs), or 7 calendar days (3 dogs). Complete details of the individual radiation treatment plans were available and reviewed for 127 of 129 animals, with some details available for the remaining 2. An average (mean) of 10 beams (range, 7 to 18) each with unique gantry angles was used. The beam arrangement was coplanar for 62 plans and noncoplanar for 66. Photons were used in all plans; for 127 dogs, the plan was constructed with 6-MV beams, and in 2 additional dogs, the plan included a mix of both 6- and 10-MV beams. Plan dosimetry for the study population is summarized (Supplementary Table S3), and details of the individual radiation treatment plans are available in the raw data file posted to the Harvard Dataverse.¹⁰

Acute radiation toxicoses were assessable in 113 dogs that underwent recheck evaluation at our institution or with another veterinarian or whose families were contacted by our team within the first 90 days following SRT. On the basis of Veterinary Radiation Therapy Oncology Group criteria for grading toxicoses,¹⁶ none of the dogs had grade 3 toxicoses, whereas late grade 3 toxicoses involving the eyes (n = 1), skin (3), or CNS (10), alone or in combination, was observed (Supplementary Table S4).

Clinical response to stereotactic radiotherapy

After SRT, there were 9 dogs for whom clinical response to treatment could not be assessed. Of the remaining dogs, 20 were reported to have had clinical signs that improved for the duration of clinical follow-up (20 of 120 dogs; 17%), 95 had initial improvement and then later experienced signs of a progressive nasal tumor (95 of 120; 79%), 5 had an unchanged clinical status (5 of 120; 4%), and no dogs were reported to get worse without initial clinical benefit. Of the dogs with a known outcome and client- or clinician-reported benefit, we were able to retrospectively categorize that benefit as follows: (a) subtle improvement that started within 3 weeks of SRT (28 of 115 dogs; 24%), (b) substantial improvement that started within 3 weeks of SRT (65 of 115; 57%), (c) subtle improvement that took longer than 3 weeks to notice (9 of 115; 8%), (d) substantial improvement that took longer than 3 weeks to notice (7 of 115; 6%), or (e) substantial improvement that occurred with an uncertain time frame (6 of 115; 5%). The most conservative estimate of overall clinical response rate was 89% (115/129 dogs, including those for whom clinical response could not be ascertained from the medical record review). Of the 95 dogs reported to have signs of progressive nasal disease at some point after SRT, testing was not performed to determine the cause for progressive signs in 63 of the dogs. In the other 32 dogs, diagnostic testing was pursued and confirmed the cause of progressive signs as one of the following: tomographically documented progressive tumor growth (15 of 32 dogs; 47%), bacterial rhinitis (7 of 32; 22%), fungal rhinitis (6 of 32; 19%), oronasal fistula (2 of 32; 6%), or nasocutaneous fistula (1 of 32; 3%); in one dog, there was rhinitis with no identifiable infectious component, thus leading to the diagnosis of sterile radiation-induced rhinitis (1 of 32; 3%).

Follow-up CT was performed in 74 dogs (74 of 129; 57%). In 21 of those dogs, a single routine restaging CT was performed within 6 months of SRT to document response to therapy (21/120; 16%). In 8 dogs (6%), a single CT was performed when the dog developed progressive nasal signs. In 17 dogs (23%), an initial restaging CT was performed within 6 months of SRT, and another CT was subsequently performed when nasal signs progressed. In 9 additional dogs (7%), 2 CT scans were performed, but on a different (nonstandardized) schedule. And in 19 dogs (15%), more than 2 recheck CT scans were performed. Of those 74 dogs with CT-evaluable responses, the best overall response was classified as follows: stable disease (7 dogs; 9.5%), partial response (47 dogs; 63.5%), or complete response (16 dogs; 21.6%) for a best overall measurable response rate (partial plus complete responders) of 85.1%. Progressive disease was documented in 4 of 74 dogs (5.4%).

Late adverse effects that were potentially attributable to SRT were able to be evaluated in 108 dogs and are summarized (Supplementary Table S4). Thirteen dogs developed either an oronasal (n = 10) or nasocutaneous fistula (3; reported as grade 3 late skin toxicosis) at a median of 475 days after SRT (range, 80 to 1,247 days). Late neurologic sequelae occurred in 18 dogs (18 of 108; 17%). Provision of a detailed description of these late neurologic sequelae is beyond the scope of this manuscript, but a summary for each animal is provided in the raw data file.¹⁰ There was no apparent increase in risk of neurologic events in dogs with advanced stage tumors, nor did use of other tumor-directed therapies before or after SRT appear to be correlated with risk. The only clinical factor for which a potential link was identified was life stage. Scorable neurologic complications were more common in dogs treated with SRT at an end-of-life stage (5/20 [25%]) versus at a mature adult (8/47 [17%]) or senior (5/41 [12%]) stage; grade 3 late neurologic effects were recorded for 10 dogs (10 of 108; 9%); in 5 instances, those dogs were mature adults or seniors (5 of 88; 5.7%), and in 5 instances, the dogs were at an end-of-life stage (5 of 20; 25%; Supplementary Table S5).

New (post-SRT) metastasis to regional lymph nodes was either suspected (based on physical examination or CT; n = 2 dogs) or confirmed (cytologically; 3 dogs) in 5 dogs at a median of 294 days (range, 191 to 633 days). New pulmonary metastasis that was attributed to the nasal tumor was documented radiographically in 7 dogs at a median of 237 days (range, 68 to 352 days). Three dogs were suspected to have developed extrapulmonary metastasis after SRT at 116 days (multifocal lytic bone lesions), 191 days (multifocal hepatic nodules), and 326 days (hepatic and spinal masses).

Tumor-directed therapies given after SRT

After SRT, 94 dogs were known to have been prescribed an NSAID (94 of 124 evaluable animals; 76%); in 20 of those dogs, the NSAID was prescribed for short-term (< 30-days’ duration) use, whereas in 68 dogs, it was prescribed long term, and in 6 dogs, it was only used when the animal developed signs of rhinitis.

Systemic antineoplastic therapy was utilized after SRT in 16 dogs (Supplementary Table S2). In 4 of these dogs, carboplatin was prescribed concurrent with SRT and then given every 3 weeks for at least 4 total doses. Beginning in 2014, this combination of SRT with carboplatin is something that was offered to all families considering SRT for nonlymphomatous sinonasal tumors as a treatment that was likely to be well tolerated and was given with hope for both radiosensitization and prolongation of EFST but described to pet owners as having an unknown likelihood of achieving either goal. Due to logistical challenges and a low interest level among our client base, we stopped offering this in 2020. Some dogs received more than 1 drug; agents utilized included carboplatin (9 dogs), doxorubicin (1 dog), mitoxantrone (1 dog, for concurrent bladder cancer), vinblastine and lomustine (1 dog, for concurrent metastatic mast cell tumor), metronomic cyclophosphamide (1 dog), toceranib phosphate (10 dogs), and masitinib mesylate (1 dog). Four other dogs received chemotherapy, but the timing and indication were unclear from review of the medical records. In fact, for most dogs, there were substantial amounts of missing data related to specific drug indication, drug dose, frequency, toxicoses, and response; for that reason, additional details are omitted from this manuscript. However, a summary of all available data is provided in the raw data file that was uploaded to the Harvard Dataverse.¹⁰

Eighteen dogs underwent reirradiation at the time of nasal tumor progression (Supplementary Table S2). A variety of protocols were utilized, including a second course of 3-fraction SRT delivering another 30 Gy total (n = 12 dogs; 8 of these 12 reirradiations are detailed elsewhere),¹⁷ a 5-fraction SRT protocol delivering 30 Gy total (1 dog), a palliative-intent course of 4 Gy X 5 daily fractions (2 dogs) and a palliative-intent course of 6 Gy X 5 weekly fractions (3 dogs). The median time to reirradiation was 311 days (range, 149 to 1,316 days). The median time from reirradiation until death was 210 days (range, 29 to 1,868 days).

One dog had a carotid artery ligation to manage severe epistaxis 217 days after SRT; another dog had palliative surgery to repair a nasocutaneous fistula approximately 1.5 years after SRT (additional details not available). One dog had frontal sinus trephination for clotrimazole infusion to manage fungal rhinitis 555 days after SRT. There were no dogs that underwent tumor-directed surgery after SRT.

Survival analysis

The overall median EFST was 237 days (95% CI, 197 to 266 days; Figure 1); events are described above. The assessment of OST included 106 deaths and 23 censored patients. The median follow-up time for censored patients was 576 days (95% CI, 427 to 777 days). Of the censored patients, 14 were alive at the time of data collection, and 9 were lost to follow-up (known to be deceased, but without a date of death). The median OST was 542 days (95% CI, 402 to 622 days). The cause of death was recorded as being unknown in 35 dogs, and in the remainder, death was attributable to recurrence or worsening of nasal tumor clinical signs (47 of 80 dogs; 59%), metastasis of the nasal tumor (12 of 80 dogs; 15%), or unrelated illness (21 of 80 dogs; 26%). Necropsy was performed and reports were available for review in 4 dogs. In each animal, there was evidence of tumor within the nasal cavity or sinuses, with variable degrees of histologically evident necrosis. Consistent with antemortem clinical data, necropsy revealed no evidence of metastasis in those animals. In a dog euthanized 28 days after irradiation of a brain-compressive nasal chondrosarcoma, there was evidence of moderate multifocal lymphoplasmacytic meningoencephalitis and neuronal intracytoplasmic inclusions. This dog had been euthanized due to rapidly declining neurologic status characterized by severe ataxia and weakness without seizures. Another dog was euthanized 1,255 days after irradiation of low-stage nasal adenocarcinoma; at necropsy, cribriform lysis and brain compression by the tumor were identified, with histologically evident marked regionally extensive encephalomalacia, gliosis, and rare neuronal necrosis. This dog had developed seizures 15 months after SRT, had intracalvarial tumor progression documented via MRI 35 months after SRT, and was euthanized 7 months later.

View Full Size

Kaplan-Meier curve depicting event-free survival time (EFST; gray line) and overall survival time (OST; black line) for 129 client-owned dogs with nonlymphomatous intranasal tumors treated with 10 Gy X 3-fraction stereotactic radiotherapy (SRT) between August 2013 and November 2020. The median EFST was 237 days (95% CI, 197 to 266 days). The median OST was 542 days (95% CI, 402 to 622 days). Tick marks represent dogs that were censored, each step in the EFST curve represents the onset of an event (local, regional, or distant progression of the nasal tumor) for ≥ 1 dog, and each step of the OST curve represents the death of ≥ 1 dog.

Citation: Journal of the American Veterinary Medical Association 260, 12; 10.2460/javma.22.03.0141

• Download Figure
• Download figure as PowerPoint slide

On univariable analysis, advancing life stage at onset of SRT was associated with significantly (P < 0.001) shorter median OST for senior dogs (542 days; n = 51) and end-of-life stage dogs (354 days; 28) versus mature adult dogs (741 days; 50; Figure 2). There were no statistically detectable differences in EFST or OST for dogs grouped on the basis of tumor type (chondrosarcoma, squamous cell or transitional carcinoma, or other nonlymphomatous confirmed intranasal malignancies). However, advancing tumor stage^3,9 was associated with significantly (P = 0.028) shorter median EFST for dogs with tumor stages T_4a (213 days; n = 36) or T_4b (195 days; 29) versus T_1–3 (265 days; 64) and significantly (P = 0.032) shorter median OST for dogs with tumor stages T_4a (541 days) or T_4b (375 days) versus T_1–3 (650 days). Median EFST was significantly (P = 0.046) shorter for dogs with lymph node metastasis (173 days; n = 9), compared with those that had no lymph node metastasis (237 days; 101). Dogs that completed SRT in 3 days had significantly (P = 0.024) longer median OST (575 days; n = 97) than did dogs with SRT of a greater duration (422 days; 32; Figure 3). Use of tumor-directed therapies (chemotherapy, radiotherapy, or surgery) given before SRT (Supplementary Figure S1) or after SRT (Supplementary Figure S2) had no measurable effect on EFST or OST; as such, their use was not considered in the multivariable analysis. However, when these neoadjunctive and adjunctive therapies were combined and assessed as a single factor, it was found that dogs that received any form of tumor-directed treatment before or after an initial course of SRT had significantly (P = 0.022) longer median OST (622 days), compared with those that did not receive such other treatment before or after SRT (472 days); because the combined single factor was significantly associated with OST, it was included in the multivariable model. Overall survival times were longer with higher near-maximum doses (D_2%) in the PTV (hazard ratio, 0.904; P = 0.031; Supplementary Table S6). Similarly, higher near-minimum doses (D_98%) existed in the GTV (hazard ratio, 0.887; P = 0.019) and PTV (hazard ratio, 0.850; P = 0.033); since the D_98% for the GTV and PTV are similar dosimetric factors and because they had similar impacts on survival in this univariable model, we opted to only advance the D_98% for the GTV into the multivariable model.

View Full Size

Kaplan-Meier curves depicting EFST and OST for the dogs described in Figure 1, grouped on the basis of life stage¹³ (mature adult, senior, or end-of-life stage) at the onset of SRT (A and B) or on the basis of tumor type (chondrosarcoma, squamous cell carcinoma [SCC], or transitional carcinoma; C and D), tumor stage^3,9 (T_1–3, T_4a, or T_4b; E and F), or the presence versus absence of lymph node metastasis (G and H). See Figure 1 for remainder of key.

Citation: Journal of the American Veterinary Medical Association 260, 12; 10.2460/javma.22.03.0141

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Kaplan-Meier curves depicting EFST and OST for the dogs described in Figure 1 grouped on the basis of SRT duration (3 vs ≥ 4 days; A and B) or whether (no or yes) they received any other tumor-directed treatment before or after SRT. See Figure 1 for remainder of key.

Citation: Journal of the American Veterinary Medical Association 260, 12; 10.2460/javma.22.03.0141

• Download Figure
• Download figure as PowerPoint slide

Multivariable Cox proportional hazards modeling demonstrated that senior dogs that did not receive (vs did receive) other tumor-directed treatments before or after SRT had greater risk (hazard ratio, 2.404; P = 0.037) of shorter EFST, whereas receiving higher near-minimum radiation doses (GTV D_98%) was protective (hazard ratio, 0.686; P = 0.003; Table 2). For senior dogs, chondrosarcoma (vs other tumors) was a risk factor for shorter OST (hazard ratio, 7.232; P = 0.007), whereas higher GTV D_98% was protective (hazard ratio, 0.743; P = 0.035), as was tumor stage T_4a (vs T_4b; hazard ratio, 0.288, P = 0.047) for longer OST (Table 3). For dogs at an end-of-life stage when SRT was initiated, squamous cell or transitional carcinoma was associated with significantly (P = 0.042) shorter EFST (hazard ratio, 6.462). Use of other tumor-directed treatments before or after SRT was associated with longer OST in both mature adult dogs (hazard ratio, 0.2410; P = 0.024) and senior dogs (hazard ratio, 0.348; P = 0.022).