Abstract
Case series summary
Salivary gland carcinoma is uncommon in cats. We report the outcome of radiation therapy in six cases (four salivary gland adenocarcinomas, one tubulopapillary adenocarcinoma, one carcinoma). Five were treated after surgical excision of the primary tumour, but four had gross disease (primary or metastatic) at the time of starting radiotherapy. Exact progression-free interval from the start of radiotherapy in the two cats where this was known was 120 and 144 days, respectively. One cat was signed off at 766 days with no evidence of recurrence. Another cat was in remission at 202 days (when last seen by the referring practice) but subsequently developed recurrence (date uncertain). Survival time was known for three cats (55 days, 258 days and 570 days from initiation of radiotherapy, respectively). In two cases, locoregional progressive disease (PD) was confirmed, and the other presumed as the cause of death. Two cats, known to have developed PD, were alive at the time of writing (at 206 and 549 days, respectively). No cat died as a result of distant metastatic disease.
Relevance and novel information
There is a paucity of information on the treatment of salivary gland tumours. In humans, as in cats, there is no optimised standard of care for malignant tumours. It is accepted that, for surgical candidates (even with large tumours), surgery and radiotherapy is superior to radiotherapy alone. However, the benefits of postoperative radiotherapy compared with surgery alone are only clear in patients with high-risk tumours (ie, those with large and invasive primary tumours, close or incomplete margins, high histopathological grade, histological evidence of neural or vascular invasion, or positive lymph nodes). This population is analogous to the population reported here, and likely to most cats presented in practice. Thus, radiation therapy may help improve locoregional control and survival in cats.
Introduction
Salivary gland carcinomas in cats are rare: 1 the overall incidence in dogs and cats has been reported to be 0.17%. 2 There are very limited treatment data, and much of the literature focuses on pathological findings,3–6 or single- or low-number case reports.7–12 Aetiology is unknown. One study (in the USA), which investigated prognostic factors in surgical cases, suggested that Siamese cats and male cats may be at increased risk, 13 but this has not been further investigated. Simple adenocarcinomas (tubular or papillary) are most common, but epithelial and myoepithelial carcinomas and duct carcinomas have also been reported,4–6,12 although not all reports have applied the World Health Organization’s (WHO) classification system. The limited data and lack of consistency of categorisation makes it difficult to assign prognostic significance to different subtypes. The limited clinical information available suggests that feline salivary gland tumours are variably locally invasive but that lymph node metastases are common (39% of cats in a series of 30).2,8,13 Distant metastases seem to be slow to develop, 5 although cats often present with advanced disease.2,8,13
Although there are few data reporting treatment, the recommended treatment for primary tumours is surgical excision.1,2,14 As many of these tumours are invasive, complete excision may be difficult to achieve. Where tumours are deemed unresectable, or incomplete excision is achieved, radiotherapy (RT) has been recommended but with little evidence.1,2 Good outcomes were reported in three dogs treated in a minimal residual disease setting, 15 but data for cats are lacking. In the largest review paper, which reported on 30 cats, treatment type (surgery alone vs surgery and radiation or surgery and chemotherapy) did not affect outcome, but, in fact, only five cats received RT. 13 In that review paper, median survival time was 516 days, 13 but there is an obvious bias as all cats were surgical candidates. There are no studies specifically reporting RT outcomes.
The aim of the present study is to report outcomes in a series of cats treated for salivary gland carcinoma using external beam RT.
Case series description
The study design was approved by the University of Liverpool Veterinary Research Ethics Committee. Cats with histologically confirmed salivary gland carcinoma/adenocarcinoma treated with external beam RT between 2008 and 2017 were identified, using the University of Liverpool Small Animal Teaching Hospital (SATH) Radiotherapy Database. All identified patients treated with RT were included, including those that also received other treatment modalities. In addition, all feline patients with salivary gland carcinoma/adenocarcinoma over the same period were identified from the University of Liverpool SATH Clinical Database.
Data were collected from the medical records, including signalment, histological diagnosis and information on margins of excision (if available), tumour site, staging methods and results, treatment prior to RT, tumour volume at the start of RT, RT parameters (dose and fractionation, photon or electron and energy, planning method, field size, inclusion of lymph nodes in field), acute adverse effects of RT scored according to the Veterinary Radiation Therapy Oncology Group (VRTOG) acute morbidity scoring scheme (Table 1; 16 at the end of treatment and at the first revisit, 10–14 days after completion of RT), any medications prescribed for RT adverse effects, best response to RT (as defined by Response Evaluation Criteria in Solid Tumors [RECIST] criteria), administration of chemotherapy or tyrosine kinase inhibitors (drug, dosage, frequency and duration), any chemotherapy toxicity scored according to the Veterinary Cooperative Oncology Group (VCOG) scoring scheme, 17 and survival data. Significant comorbidities were also noted (any concurrent diagnoses found in the clinical notes).
Veterinary Radiation Therapy Oncology Group (VRTOG) acute radiation morbidity scoring scheme, 16 as used to assess six cats receiving radiotherapy for salivary gland carcinomas
For planning CT scan and treatment delivery, patients were anaesthetised and immobilised using vacuum cushions (Oncology Imaging Systems), foam blocks (JAK Marketing) or thermoplastic mask (Aquaplast U-frame standard; Oncology Imaging Systems) and a customised head support (MoldCare pillow; Oncology Imaging Systems), secured to a plastic head frame with four points of fixation (Bionix Development), as appropriate.
All treatments were delivered using a linear accelerator (Clinac 2100; Varian Medical Systems). Two fractionation schemes were used: either 12 × 4 Gy (on a Monday–Wednesday–Friday basis for four consecutive weeks) or 4 × 8 Gy or 9 Gy (one fraction per week for four consecutive weeks).
Most treatments were manually planned to include the scar or gross tumour plus a 3–5 cm margin when possible. Tissue-equivalent bolus material (Superflab [Imaging Equipment] and Superstuff [JRT Associates]) was used to improve dose distribution (by moving Dmax closer to the skin surface) and reduce air gaps as required. For photon plans, additional field shaping was carried out using multi-leaf collimator (MLC) beam modification. For patients treated with electrons, standard rectangular cut-outs were used with additional lead blocks for the first treatment, then individual, patient-specific cut-outs were made to block non-target tissues, if required.
Computerised plans were generated from CT images using Pinnacle version 8/9 (Philips Radiation Oncology Systems; Philips Healthcare) with the intent to include at least 95% of the planning treatment volume in the 95–105% isodose. The gross tumour volume (GTV) was delineated by the oncologist responsible for treatment planning (LB) using co-registered contrast-enhanced CT images, and clinical target volume (CTV), accounting for subclinical microscopic disease extension of 3 cm, was defined. The CTV margin was extended three-dimensionally by 5 mm to define the planning target volume (PTV) accounting for internal physiological movements, patient motion and set-up uncertainties. Organs at risk were segmented (tympanic bullae, brain, spinal cord, base of tongue, etc) and PTV was adjusted appropriately to allow preferential protection of the spinal cord or other organs at risk. Plans utilised two or three coplanar beams, with beam collimation using MLC beam modification using dynamic wedges when appropriate. Portal imaging was carried out at least twice during the treatment protocol, to verify position.
In all cases, the local lymph node(s) were irradiated by including the nodes in the treatment field for the primary tumour (either manually or computer planned). Margins around lymph nodes were 1 cm (combined CTV of 5 mm and PTV of 5 mm).
Field size was determined from the RT records. For planned patients, the largest two-dimensional field size was used.
When possible, progression-free survival (PFS) was determined: this was the time from the start of RT until progression was documented. Survival time was the time from the start of RT to death or euthanasia. Cause of death was recorded.
Descriptive analysis alone was performed owing to the small sample size, which precluded statistical analysis.
Six cats were identified from the RT database that were included in the study. Interestingly, the hospital database search revealed only one additional cat with confirmed salivary gland carcinoma presented over the study period, which did not receive RT. The data of the six cats that were irradiated are summarised in Table 2. There were five domestic shorthair cats and one Siamese cat; three were male neutered and three female neutered. Age ranged from 6 years to 15 years 9 months, with a mean of 9 years 7 months, and median of 9 years 4 months. Two cats (cats 2 and 4) had been diagnosed with hyperthyroidism: both were well controlled (one after radioiodine treatment and the other on carbimazole [Vidalta; MSD Animal Health]).
Summary data for six cats with salivary gland carcinoma treated with radiation therapy
LN = lymph node; RT = radiotherapy; TKI = tyrosine kinase inhibitor; PFI = progression-free interval; MN = male neutered; L = left; MSG = mandibular salivary gland; SM = submandibular; RP = retropharyngeal; MV = megavolts (photon energy); R = right; Carbo = carboplatin; Eu = euthanased; PD = progressive disease; DSH = domestic shorthair; FN = female neutered; FNA = fine-needle aspirate; MeV = mega electron volts (electron energy); SD = stable disease; CR = complete remission; LTFU = lost to follow-up; Cyclo = cyclophosphamide; y = years; m = months; d = days
All cats were presented to the referring veterinarian because the owner had detected a mass, and time between mass detection and initial presentation to the primary care vet ranged from 3 days to approximately 1 month.
All tumours were diagnosed histologically: there were four salivary gland adenocarcinomas, one tubulopapillary adenocarcinoma (which under the WHO classification would be referred to as adenocarcinoma), and one carcinoma. Unfortunately, tissue was not available for review. All pathology reports stated salivary gland or likely salivary gland origin. The diagnosis of carcinoma in cat 5 was thought to refer to carcinoma in the broad sense, rather than reflecting application of the WHO criteria for salivary gland carcinoma (where the term carcinoma refers to undifferentiated tumours) had been applied. The histopathological description was highly suggestive of glandular origin/adenocarcinoma and did not describe an undifferentiated tumour. The reports for cats 1 and 3 were written by non-boarded pathologists, for case 3 by a boarded pathologist, for case 6 by a non-boarded pathologist but verified by a boarded pathologist, and for cases 2 and 4, the pathologists’ qualifications were unknown.
Four tumours affected the mandibular salivary gland (three left, one right); one arose in the left buccal cavity and the remaining tumour arose in the labial mucosa (presumed to be from minor salivary gland tissue) (Table 2). The salivary gland of origin was identified based on history, clinical examination and imaging findings at presentation to the SATH.
All patients had clinical staging prior to RT: five had CT of head/neck, thorax and abdomen, and one had inflated thoracic radiographs (this patient had only clinical evaluation of the primary site as the primary tumour had been excised by the referring veterinarian, leaving only microscopic residual disease). All had fine-needle aspirates (FNAs) of one or more locoregional lymph nodes (as deemed appropriate by the clinician): retropharyngeal nodes were aspirated under ultrasound guidance, and submandibular nodes were identified by palpation.
In cat 1, the ipsilateral submandibular and retropharyngeal nodes were assumed to be affected by metastases due to enlargement on CT and were not aspirated, and the ipsilateral pre-scapular (superficial cervical) node was negative on aspiration. In two cases (cats 5 and 6), FNAs confirmed nodal metastases, and in two cases (cats 3 and 4) aspirates were negative. In cat 2, lymph node FNAs and subsequent histology after resection revealed no evidence of metastases to submandibular or retropharyngeal nodes. No distant metastases were confirmed at diagnosis: cat 4 had a focal, poorly marginated, approximately 5 mm, rounded area of moderately increased attenuation in the dorsal aspect of the left cranial lung lobe of uncertain significance. This remained unchanged on CTs 3 weeks and 5 months later, so was presumed not to be metastatic. This cat also had a small adrenal nodule (suspected adenoma based on CT findings [well defined and non-invasive] and lack of progression when re-scanned 5 months later).
Five of the six cats (cats 2, 3, 4, 5 and 6) had surgical excision of the primary tumour prior to RT: cat 1’s tumour had been deemed non-resectable and it was presented with gross disease. For cat 2, the owner delayed initial surgery for 2 months after its first visit to the hospital. This patient had repeat lymph node FNAs at re-presentation, which were negative. The primary and two enlarged nodes (submandibular and retropharyngeal) were resected despite this result; histopathology of the lymph nodes revealed no metastases, consistent with the results of the previous FNAs. However, when the cat was presented to start radiation therapy (29 days after surgery), two nodules were present in the subcutaneous tissue of the submandibular area: FNA revealed only carcinoma cells, so it was unclear if this tissue was recurrent tumour or effaced node. Cat 5 had undergone two surgeries prior to RT (244 days apart), and the tumour had been present for 228 days prior to the first surgery. RT started 38 days after the second surgery. The other cats started RT 30, 31 and 52 days postoperatively, respectively.
At the time of starting RT, cats 1 and 4 had measurable primary disease (tumour volume 42.07 cm3 and 2.25 cm3, respectively, based on outlined GTV for computer planning). Cat 2, as described above, had two small carcinoma nodules of 6 mm diameter each, of unclear origin. The other three cats had no measurable primary disease, but in cat 6 the ipsilateral retropharyngeal node was enlarged on CT scan and FNA had confirmed metastasis.
Four cats (1, 2, 3 and 6) were treated with the 4 × 8 Gy or 9 Gy (one fraction per week for four consecutive weeks) protocol, and two (cats 4 and 5) with the 12 × 4 Gy (on a Monday–Wednesday–Friday basis for four consecutive weeks) protocol. Palliative (4 × 8 Gy) treatment was recommended for cat 1 with a very large volume tumour because of the advanced nature of the disease, aiming to avoid significant acute side effects that could impact on the cat negatively when expected lifespan was short. In the other three cases, the palliative protocol was chosen rather than the definitive by the owner (for logistic or financial reasons). Cats 1, 2 and 6, treated with the palliative protocol, received 32 Gy in four fractions, whereas cat 3 received 36 Gy: this was because of a change in institutional policy for the palliative protocol during the study period.
Cats 1 and 5 were treated with computer-generated plans (one 4 ×8 Gy and one 12 × 4 Gy) using coplanar beams at 6 megavolts (three beams in one case and two in the other). The remaining cats were treated using electrons, all using single fields to include the primary site and ipsilateral submandibular nodes, with manual planning to include the ipsilateral retropharyngeal lymph node in the 90% isodose for two. Electron energy was 6 mega electron volts (MeV) in cat 3, and 12 MeV in cats 2, 4 and 6: in cat 5, energy was reduced to 9 MeV half way through treatment, to spare deep tissues.
RT side effects were very mild: all cats had VRTOG grade 1 toxicity at the end of therapy, although in cats 1 and 6 self-trauma (thought to be due to irritation as a result of the tumour itself, rather than the RT) was thought the likely cause of mild erythema. In cats 2 and 3, there was alopecia only, and in the other four there was grade 1 erythema of skin or mucosa, and/or alopecia. Four cats (2, 4, 5 and 6) were prescribed prednisolone at between 0.2 mg/kg and 1 mg/kg at the end of RT, tapering off after 10 days, to treat suspected adverse effects (cutaneous erythema, mucositis). Cat 1 was prescribed prednisolone for a 6 week course as part of the management of aural inflammation associated with tumour mass effects (1 mg/kg). Cat 5 had topical treatment with silver sulfadiazine cream (Flamazine, 1%w/w cream; B&S Healthcare). At the first check-up, no additional toxicity was noted and previous skin/ear canal toxicity had reduced or resolved in all cats: no further medications were prescribed. The exception to this was cat 1 with large tumour volume, where ongoing otitis as a result of the obstructive effects of the mass was evident.
Best response to RT was recorded: four cats (3, 4, 5 and 6) achieved or remained in complete remission. For cats 1 and 2, although there was a subjective reduction in tumour volume, measurements confirmed stable disease based on RECIST criteria (in cat 2, the cat with the two small nodules postoperatively, stable disease was only confirmed at first follow-up and progression was suspected soon after: the cat was not seen again but was euthanased after 55 days).
Five cats received medical therapies during or following RT, and one received no additional therapy (cat 3). Cats 1 and 5 received carboplatin (cat 1 starting at the first RT treatment and cat 5 at the last). Cat 1, the cat with the largest volume disease, tolerated carboplatin well and completed four cycles as planned, starting at a dose of 180 mg/m2 (hospital standard dose of 200 mg/m2 – 10% owing to concerns about myelosuppression due to concurrent RT and chemotherapy). Cat 5 initially also received 180 mg/m2. The dose was escalated to 240 mg/m2 and the cat then developed neutropenia. At the time of the planned third dose, the cat had a VCOG grade 1 neutropenia and, after discussion with the owner, treatment was delayed for a week and dose reduced to 200 mg/m2. At the time of the next planned treatment, grade 2 neutropenia was found and the owners elected to discontinue chemotherapy. The cat then started toceranib therapy 117 days after the first carboplatin (145 days after starting RT).
Four cats received toceranib therapy: cat 2 started at the end of RT (as recommended by the clinician); cat 6, 70 days post-RT (treatment had been recommended during RT but the owner initially declined); cat 4, 80 days post-RT (owing to suspected recurrence); and cat 5, 145 days post-RT after stopping carboplatin as described above. Cats received 2.3 or 2.6 mg/kg on a Monday–Wednesday–Friday basis, or 2 mg/kg on an every other day basis. Cat 2 was lost to follow-up after the first revisit, when stable disease was noted, but was euthanased after 55 days: the primary care veterinarian could not provide any information on the reason for euthanasia, and our ethics approval did not allow us to contact owners, but progressive disease (PD) is a likely cause. Cat 6 developed PD after another 50 days, at which point metronomic chemotherapy (cyclophosphamide 5 mg q24h, meloxicam dose for 5 kg q24h) was started. At the owners’ request, cyclophosphamide was discontinued after another 42 days: the owner did not want to restage to re-evaluate disease response/progression and felt monitoring visits were having a negative impact on the cat’s quality of life. The cat was alive at the time of writing (206 days post-RT), receiving meloxicam only. Cat 5 was last seen at the hospital at 164 days after starting RT, with no evidence of tumour recurrence on clinical examination, and then seen by the referring veterinarian after 202 days (for another condition) with no signs of recurrence, again based on clinical examination. The cat was not re-examined, but the owner reported it was alive with recurrent disease at 549 days post-RT. An accurate time to recurrence could not be determined.
Time to recurrence/progression-free interval from the start of RT in the two cats where this was properly documented (cats 4 and 6) was 144 days (possibly suspected at 80 days but not confirmed) and 120 days. Cat 3 was signed off at 766 days with no evidence of recurrence. Cat 5 was apparently in remission at 202 days (when last seen by the referring practice) but subsequently developed recurrence (documented by the referring veterinarian and owner, but date uncertain).
Survival time was known for three cats (55 days [cat 2], 258 days [cat 1] and 570 days [cat 4] from initiation of RT). In cats 1 and 4, PD was confirmed, whereas in cat 2 it was presumed as the cause of death. Cat 3 was lost to follow-up after sign off at 766 days post-RT. Cats 5 and 6, known to have developed PD, were alive at the time of writing (at 206 and 549 days, respectively).
Discussion
This case series documents variable biological behaviour and outcomes in a small number of cats treated with RT for salivary gland carcinoma. Rapid recurrence after surgery occurred in two cases prior to RT, and recurrence or progression after RT was documented in 4/6 cases and considered highly likely to have occurred in one other. Lymph node metastasis was common (documented in the 3/6 cases presented here), but distant metastasis was not confirmed in any cat at diagnosis, and no cat was known to have died or been euthanased owing to distant metastatic disease: in the one case known to have been euthanased but where the reason for euthanasia was not definitely available, PD was strongly suspected owing to lack of response post-RT. The literature also supports that postoperative recurrence and regional lymph node metastasis are common.2,3,8,9,13,18,19 The most important aspect of the management of salivary gland carcinomas therefore appears to be management of locoregional disease, which means surgery and RT are likely to be the most valuable treatment modalities. However, the role of RT is unclear and outcomes in the current study were variable.
As in cats, salivary gland carcinomas in humans have a very low incidence. Overall, benign tumours are more common than malignant ones, 2 but most tumours affecting minor salivary glands are malignant. 20 As in cats, there is no optimised standard of care for malignant tumours. 20 The role of RT remains controversial. It seems relatively clear that for surgical candidates (even with large tumours), surgery and RT are superior to RT alone. 20 However, the benefits of postoperative RT vs surgery alone are less clear.
One large analysis (8243 patients) compared survival outcomes between surgery alone or with adjuvant radiation after resection of major salivary gland tumours, and found that adjuvant radiation for high-risk major salivary gland cancers (patients were considered high risk if they had extracapsular extension and/or positive margin after resection) appeared to offer a survival benefit, but could not demonstrate this for low- and intermediate-risk salivary gland cancers. 21 Other studies have also suggested a clear benefit, with postoperative RT improving local control, reducing the risk of local recurrence and extending survival,22,23 with the benefits most consistent for patients with higher stage and grade tumours, and risk factors such as incomplete excision. Importantly, for early-stage disease, there is no clear benefit. Unlike other (more common) human tumours, there is a lack of randomised prospective trials, and selection bias may affect the results of the available studies. However, overall, postoperative radiation therapy is recommended for patients with adverse prognostic factors: large and invasive (T3 or T4) primary tumours, close or incomplete margins, high histopathological grade, histological evidence of neural or vascular invasion, or positive lymph nodes. 20
All cats in the current study would meet one or more of these criteria. Interestingly, histological features may be of prognostic relevance in cats, and reporting the histological grade of feline tumours and systematically classifying these tumours according to the WHO criteria may allow future studies to better assess this; certainly, this will be impossible if more complete and consistent histological descriptions are not available.
The survival times in the current study are apparently shorter than those reported by Hammer et al in a larger series of surgically treated dogs and cats. 13 It is not possible to calculate a median survival time for the current cohort, but given that we have survival times of 55, 258 days and 570 days, with two cats alive at 206 and 549 days and one signed off at 766 days, the median would be greater than 403 days. In the study by Hammer et al, only 5/30 cats had RT and it may be that the quoted median survival time of 561 days was not representative of this subgroup.
In the current study, one cat had unresectable disease, two had recurrent disease at the time of RT, and one cat had had two surgeries prior to RT. In general terms, for other locally invasive feline tumours, such as injection site sarcoma, 24 irradiation of gross disease and multiple surgeries are associated with a poorer prognosis. 25 These factors may therefore have impacted on survival times in the current population. In any case, the survival times for the current, small cohort are more in keeping with those reported as case reports. 11
It is likely that there was some bias to longer survival in the study by Hammer et al, 13 as all cats were surgical candidates. However, there is no description of the surgery performed. Nonetheless, there was a high rate of local lymph node metastases in the study by Hammer et al, 13 suggesting that surgery can achieve good results, though many cases will require lymphadenectomy. Therefore, thorough staging is recommended. Finally, as survival of our cohort was calculated from the date of starting RT, survival times would have been reported as longer if they had been calculated from surgery (median time to starting RT was 31 days after the most recent surgery; range 29–52 days).
Although the survival benefit achieved is not clear, all cats tolerated the RT treatment well. Acute radiation toxicity was minimal in the current study, with cats only experiencing grade 1 toxicity. This is to be expected from palliative weekly protocols, and side effects are often minimal in cats treated with definitive protocols: feline skin, in particular, is apparently more resistant to radiation-induced side effects than canine or human, and acute effects are typically limited to epilation/alopecia, erythema and dry desquamation. 26
The role of medical therapies in the management of salivary gland carcinomas is unclear in all species. In the current study, two cats received carboplatin. Protocols containing platinum compounds are most commonly used in aggressive salivary carcinomas in humans:20,27,28 as cisplatin is toxic in cats, carboplatin was considered the most appropriate choice. However, no chemotherapy regime is proven to improve either disease-free or overall survival, although in a palliative care setting, both platinum-based and anthracycline-based chemotherapy can slightly increase survival in humans. 29 In the veterinary literature, one cat is reported to have experienced partial remission of a gross salivary gland carcinoma in response to doxorubicin and cyclophosphamide, and survived for 9 months. 30 There are no other reports of a measurable response to chemotherapy. In addition, combining chemotherapy with RT for humans with high-risk salivary gland carcinomas (ie, those with large and invasive primary tumours, close or incomplete margins, high histopathological grade, histological evidence of neural or vascular invasion, or positive lymph nodes) does not offer survival benefit but does increase morbidity, so is not recommended. 31
Toceranib phosphate has been used in cats with a variety of malignancies, including carcinomas and adenocarcinomas, and is generally well tolerated,32–34 but data on response are limited. The short median durations of treatment (40, 77 and 100 days, respectively)32–34 in the toxicity studies suggest that there is often relatively rapid disease progression. However, there has been no specific evaluation of toceranib in salivary gland carcinomas: the one cat included in the toxicity review from our own hospital is included in the current study. Currently, there is no clear evidence to suggest toceranib improves survival. Interestingly, many human salivary gland adenocarcinomas express high levels of c-KIT, but responses to imatinib have been poor.35–37 The KIT status of feline salivary gland carcinomas is unknown, and the rationale for use of toceranib was mainly for its anti-angiogenic properties. Similarly, metronomic chemotherapy is generally well tolerated in cats, 38 but there is, as yet, no evidence to suggest efficacy in salivary gland carcinoma.
Conclusions
The optimal treatment for feline salivary gland carcinomas is unknown. Based on the current series and the available literature, good outcomes rely on good locoregional control and cats that are surgical candidates should be treated surgically. Extrapolating from the human literature, many cats would fall into the ‘high-risk’ groups (ie, those with large and invasive primary tumours, close or incomplete margins, high histopathological grade, histological evidence of neural or vascular invasion, or positive lymph nodes), and for these cats, RT may offer a survival benefit. Responses to RT may occur in a gross disease setting but may be short lived, as in the current study. Distant metastasis appears an infrequent cause of death. Very long-term control (>2 years) was only achieved in one cat in the current series, treated postoperatively in a minimal residual disease setting with no documented metastatic disease at presentation; ie, in a minimal residual disease setting. Survival in cats treated with large-volume primary disease is likely to be short.
Footnotes
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
