A Review of 18 Cases of Feline Colonic Adenocarcinoma Treated with Subtotal Colectomies and Adjuvant Carboplatin
Feline colonic adenocarcinoma is a locally invasive, highly metastatic tumor that is most often treated with wide surgical excision (subtotal colectomy) and systemic chemotherapy either with or without nonsteroidal anti-inflammatory medications. In this retrospective study, the outcome of subtotal colectomy and adjuvant carboplatin in 18 client-owned cats is described. The median carboplatin dose was 200 mg/m2 (range, 200–254 mg/m2) q 4 wk with a median of five doses/cat (range was two to seven doses/cat). Limited toxicities were noted. Positive prognostic factors for the disease-free interval included cats that had weight loss as a presenting sign (P < 0.036) and negative prognostic factors for median survival included nodal and distant metastasis (178 versus 328 days and 200 versus 340 days, respectively). The median disease-free interval was 251 days (range, 37–528 days) and the median survival time was 269 days (range, 40–533 days). Subtotal colectomy and adjuvant carboplatin is a safe and potentially effective treatment for cats with colonic adenocarcinoma.
Introduction
Feline gastrointestinal adenocarcinoma is the second most common gastrointestinal tumor in the cat and the most common tumor in the colon of cats.1,2 Cats with colonic adenocarcinoma reportedly have locally invasive disease as well as advanced metastatic disease at the time of presentation.1 In one report of 21 cats, 76% of cats had distant metastasis at the time of presentation.1 In that same report, cats that had aggressive local surgery (subtotal colectomies) followed by doxorubicin had survival times of 280 days versus only 56 days in cats without chemotherapy.1 In a more recent report of two cats treated with both palliative stenting and a nonsteroidal anti-inflammatory medication, survival times were 274 and 54 days.3 The two cats in that study eventually died due to metastatic disease, with the colon remaining patent postsurgery or stenting.3 That finding indicates that with good local control, cats can have an improved survival time; however, they will eventually succumb to metastasis, warranting adjuvant treatment with chemotherapy. Carboplatina is a platinum agent in the family of alkylators shown to have efficacy in human sarcomas and carcinomas.4–7 Carboplatin has also shown efficacy in dogs with sarcomas and carcinomas.8–10 In cats, there are isolated case reports of treatment with carboplatin, a report with carboplatin in combination with gemcitabine, and two phase 1 studies reporting mild to moderate efficacy for carboplatin in cats with carcinomas.11–14 The efficacy of carboplatin against carcinomas in humans and dogs are limited, but the apparent safety of carboplatin in cats may warrant adjunctive use of carboplatin for feline colonic adenocarcinomas.4–14
The purpose of this retrospective study was to determine the signalment, diagnostic findings, disease-free interval, survival time, and prognostic factors for cats with colonic adenocarcinoma treated with subtotal colectomies and adjuvant carboplatin. Secondly, any toxicity and chemotherapy treatment-related toxicoses due to carboplatin treatment were reported.
Material and Methods
Medical records of all cats diagnosed with colonic adenocarcinoma at the The Animal Medical Center and Long Island Veterinary Specialists between 2003 and 2009 were reviewed. Cats that were treated with subtotal colectomies and carboplatin that had completed medical records with follow-up information were included. Eighteen cats met the inclusion criteria (14 cats from The Animal Medical Center and 4 cats from Long Island Veterinary Specialists).
Data abstracted from the medical records included signalment, clinical signs, duration of clinical signs, physical exam reports, results of thoracic radiographs, results of abdominal ultrasound reports, site of neoplasia, histopathologic diagnosis, completeness of surgical margins, and whether there was evidence of metastasis at surgery. Metastasis was determined by histologic diagnosis, except in the case of lung nodules on thoracic radiographs, which were assumed to be the spread of primary colonic adenocarcinoma.
The dose of carboplatin, interval, number of doses, and toxicities attributed to carboplatin were recorded. Toxicities were graded according to the published Veterinary Co-operative Oncology Group (VCOG) criteria for adverse events.15 Hematology and serum biochemistry were performed on the day of treatment. On days 14 and 21, hematologic parameters were repeated.
Disease-free interval was assessed from the date of histopathologic diagnosis to the diagnosis of recurrence or metastasis. Survival time was calculated from the date of diagnosis to either the date of death from local disease or metastasis from colonic adenocarcinoma. If thoracic nodules were seen on thoracic radiographs, they were considered metastasis from colonic adenocarcinoma. On abdominal ultrasound, if lesions were seen that were not on the staging ultrasound, those masses were considered metastasis from colonic adenocarcinoma. The thoracic radiographs and abdominal ultrasounds were routinely performed 6 mo after the last carboplatin treatment. Additional restaging was performed earlier if owners reported recurrence of clinical signs. Recurrence and/or metastasis were assessed by abdominal ultrasound and/or thoracic radiographs in all cases. Necropsies were performed on 13 of the cats in the study.
Statistical Analysis
Disease-free interval and median survival time were assessed using the Kaplan-Meier method. Comparisons between populations were made using log-rank tests. Variables tested included signalment, clinical signs, duration of clinical signs, hematologic and biochemical parameters, histopathologic type of colonic adenocarcinoma, site of colonic adenocarcinoma, nodal metastasis, distant metastasis at surgery, the dose of carboplatin given, and toxicities from carboplatin administration. A Cox proportional hazard (PH) regression model was also used. Significant prognostic factors were then assessed using multivariate analysis. A P value of ≤ 0.05 was considered significant. Cats that were either lost to follow-up or that died because of reasons other than colonic adenocarcinoma were censored.
Results
Eighteen cats met the criteria for inclusion in this study. Median age was 11 yr (range, 4.6–19 yr). Eleven cats were castrated males and seven were spayed females. There were 13 domestic shorthairs, 2 Siamese, 2 domestic longhairs, and 1 Burmese. The median weight was 4.8 kg (range, 2.9–7.4 kg).
Clinical signs on presentation included one or more of the following: weight loss (n = 14), inappetence or anorexia (n = 12), vomiting (n = 8), diarrhea (n = 13), and tenesmus (n = 13). Median duration of clinical signs was 18.9 days (range, 5–28 days). On physical exam, 11 of 18 (61%) cats had palpable abdominal masses. All 18 cats had hematology and serum biochemistry performed. Characteristic hematologic or serum biochemical abnormalities were not noted.
Thoracic radiographs were performed on all 18 cats. One cat had one nodule in the right cranial lung lobe and one cat had a solitary nodule in the right caudal lung lobe. Both were confirmed as metastatic lesions on necropsy (i.e., 11% had lung metastasis). Abdominal ultrasonography revealed a mass in the abdomen in 14 of 18 (88%) cats. In all but one of the cats identified with abdominal masses, a mass was isolated in the intestinal tract by ultrasound (94%). In 7 of 13 (54%) cats with masses identified in the intestine, the ultrasonographer was able to localize the mass to the large intestines. Six of the cats underwent helical computed axial tomography performed, which revealed a large intestinal mass in each of the cats with a median colonic mass diameter of 3.2 cm (range, 2.5–7 cm).
Subtotal colectomies were performed in all 18 cats. Locations of the primary mass in the cats included the following: cecal (n = 2), ileocecal (n = 7), and colonic (n = 9). Histologic type of colonic adenocarcinoma included the following: tubular (n = 10), mucinous (n = 4), and undifferentiated (n = 4). Surgical margins were complete in all 18 cases. Following surgery, histologic evidence of local metastasis was confirmed in eight cats (44%), distant metastasis in three cats (17%), and seven cats (39%) had no confirmed metastasis. Nodal metastasis was found in six cats (33%). Metastatic sites included colonic lymph nodes (n = 4), mesenteric lymph nodes (n = 2), liver (n = 2), mesentery (n = 1), mesocolon (n = 1), and urinary bladder (n = 1). The two cats with pulmonary nodules were among those that had histopathologically confirmed distant intra-abdominal metastasis.
The median carboplatin dose was 200 mg/m2 (range, 200–254 mg/m2) q 4 wk. Dosing was at the discretion of the doctor treating the case. A median of five doses was administered to each cat (range, 2–7). One cat’s treatment was discontinued after the fifth dose due to reported azotemia (blood urea nitrogen was 173 g/dL [range, 6–31 mg/dL] and creatinine was 5.1 mg/dL [range, 0.5–1.6 mg/dL]), hyposthenuria (1.008), and mild neutropenia (2,400/μL). That cat reportedly had a normal blood urea nitrogen (22 g/dL), creatinine (1.6 mg/dL), and specific gravity (1.032) before treatment. There were neither dose reductions nor delays secondary to either myelosuppression or gastrointestinal toxicity in any of the cats. After the cumulative 84 doses of carboplatin administered, there were three grade 1 neutropenias, two grade 1 thrombocytopenias, and one grade 2 thrombocytopenia. There was also one grade 1 gastrointestinal toxicity (Table 1).
In subsequent restaging abdominal ultrasounds, distant metastasis was eventually found in all 18 cats. Clinical signs related to metastasis included inappetence/anorexia, lethargy, weight loss, vomiting, and shallow, quick breathing. Restaging was performed 6 mo after carboplatin treatment unless clinical signs occurred before that time. Overall, 7 of 18 cats had abdominal ultrasound examinations before the scheduled 6 mo examination due to reports of recurrence of clinical signs. Of those seven cats, six had metastatic lesions. One cat did not have metastasis and instead proved to be constipated, which resolved with lactulose. Thoracic radiographs performed at either recurrence of clinical signs or at the 6 mo restaging found that seven cats had pulmonary metastasis. One cat with a pulmonary nodule on presentation had resolution of the nodule after the second carboplatin treatment. That same cat eventually developed another colonic mass and abdominal carcinomatosis with no evidence of pulmonary metastasis at 251 days. Thirteen of the cats were necropsied and all cats had metastatic adenocarcinomas. Metastasis after treatment occurred in the lymph nodes (n = 13), liver (n = 11), peritoneum (n = 11), spleen (n = 8), omentum (n = 8), colon (n = 5), lungs (n = 5), urinary bladder (n = 3), thoracic nodes (n = 5), and skin (n = 1).
Survival was censored for one cat that had metastatic disease; however, that cat died due to congestive heart failure. As shown in Figure 1, the median disease-free interval was 251 days (range, 37–528 days). A positive significant predictor of disease-free interval was weight loss (290 days versus 75 days, P = 0.0189). Significant negative predictors included cats who presented with nodal metastasis (85 days versus 325 days, P = 0.0349), and distant metastasis (87 days versus 330 days, P = 0.0202). A positive predictor of significance was weight loss (PH, 4.118; 95% confidence interval [CI], 1.146–14.8; P = 0.031). The only negative predictor was gastrointestinal toxicity (PH, 16.93; 95% CI, 1.031–250; P = 0.0475). Predictive for disease-free interval with multivariable analysis was weight loss (PH, 16.904; 95% CI, 1.303–219.8; P = 0.0475).
Citation: Journal of the American Animal Hospital Association 48, 6; 10.5326/JAAHA-MS-5807
As shown in Figure 2, the Kaplan-Meier overall median survival time was 269 days (range, 40–533 days). With univariate analysis, a negative predictor for survival was presence of nodal metastasis at presentation (328 days for cats without nodal metastasis [range, 92–475 days] versus 178 days for cats with nodal metastasis [range, 40–533 days]; P = 0.0373). Also significant on univariate analysis was distant metastasis. Cats without distant metastasis survived 340 days (range, 168–475) and cats with distant metastasis lived 200 days (range, 40–533 days; P = 0.029). There were no statistically significant prognostic factors influencing survival on multivariate analysis.
Citation: Journal of the American Animal Hospital Association 48, 6; 10.5326/JAAHA-MS-5807
Discussion
Signalment, clinical signs, durations of clinical signs, and physical exam findings were similar to data in previous studies.1,16 Hematology and serum biochemistry were not noted to be consistently abnormal either before or after surgery or following treatment with chemotherapy. Abdominal ultrasonogaphy was able to target a mass in the gastrointestinal system in 14 of 18 cats, which is also similar to previous reports.1 Six of the cats had helical computed axial tomography performed, and the mass and size of the mass were confirmed at the time of surgery. There do not appear to be any studies comparing abdominal ultrasounds to abdominal computed tomography for accuracy of intestinal masses in cats. The accuracy of detecting colonic masses in this sample of six cats suggests such a study is warranted. Histologic subtypes of colonic adenocarcinoma in this study population did not have a significant correlation with prognosis; however, this may be due to poor statistical power from too few cases.
Interestingly, cats that had weight loss as a presenting sign had a longer disease-free interval on multivariate analysis. Weight loss was the most commonly reported clinical sign, with 14 of 18 (78%) of the cats having weight loss. Three of the four cats that did not have weight loss did have distant metastasis, which was also found to be prognostic for shortened disease-free interval and survival time. The four cats that did not have weight loss presented mainly for acute anorexia of a few days duration, which might not have been enough time for significant weight loss. Because 3/4 of the anorexic cats had distant metastasis, more aggressive disease might explain their shortened survival time compared with cats with a more chronic duration of clinical signs.
Cats with colonic adenocarcinoma treated with subtotal colectomies and adjuvant carboplatin had mild toxicity with a median survival of 269 days. It was shown in the study by Slawienski et al. (1997) that cats with colonic adenocarcinoma should receive subtotal colectomies to assure clean surgical margins and extended survival.1 In that same study, cats receiving chemotherapy had longer survival times than cats not receiving chemotherapy; however, the sample size included only four cats, and the group of cats receiving chemotherapy was not stratified regarding the presence of local or distant metastatic disease nor was that a randomized trial to delineate true significance of chemotherapy benefit. The presence of metastatic disease in the current study does indicate a shorter disease-free interval and survival time; however, the cat with the longest survival time in the current study had distant metastasis on presentation (survival time was 533 days). In the study by Hume et al. (2006), one cat survived 274 days after a colonic stent was placed.3 That cat was treated adjunctively with nonsteroidal dugs; however, it eventually succumbed to its disease due to multiorgan metastasis.3 Both cats in the study by Hume et al. (2006) that received palliative colonic stenting retained the ability to defecate and were euthanized due to metastatic disease.3
The chemotherapeutic agent used in the study by Slawienski et al. (1997) study was doxorubicin. Cats receiving surgery and chemotherapy survived 280 days, whereas cats with just surgery survived only 56 days.1 In the current study, with subtotal colectomies and carboplatin, a median survival time of 268 days is comparable to the cats treated with doxorubicin in the former study. These data argue that prospective studies with a group of control cats are warranted to evaluate the impact of the various treatments. As the disease-free interval and survival time is decreased in cats with metastatic disease, it is arguable that cats that have metastasis should not have such extensive surgery as subtotal colectomies and perhaps should undergo palliative procedures such as colonic stenting instead. It is important to note that the cat with the longest survival time in this current study did have distant metastasis at presentation, and that cat remained stable for 528 days.
The dose and dose interval of carboplatin administered in this study was at the discretion of the oncologist treating that particular case. Most of the included cats were treated prior to the publication of the VCOG phase 1 study of carboplatin in cats and the glomerular filtration rate clearance studies conducted by Bailey et al. (2009).16,17 Three cats in that study were treated at 240 mg/m2 q 4 wk as suggested by the VCOG study, and one cat at 250 mg/m2 as per the individual prescription dose equation suggested in Bailey et al. (2009).16,17 Twelve of the cats were treated at 200 mg/m2 q 4 wk, one cat was treated at 210 mg/m2 and one was treated at 220 mg/m2. The sample size is too small to compare survival times in the context of the dose; however, it should be noted that there were no toxicities noted in the medical records at either the VCOG suggested dose or individual dose equation of carboplatin compared with the lower 200 mg/m2 dose that most of the cats were treated at. In a future prospective study, either the more recent suggested doses by VCOG or the individualized dose prescription model derived by Bailey et al. (2009) should be administered.16,17
Only one cat in this study had reported renal toxicity, this was at the 200 mg/m2 dose after the fifth dose. This cat’s azotemia resolved within 2 days with IV fluids, enrofloxacin, and famotidine. Unfortunately a urinalysis, urine culture, and abdominal ultrasound were not performed making it difficult to know if that cat actually had renal toxicity from carboplatin administration, a toxin, pyelonephritis, acute dehydration, or some other etiology. No further treatment was given to the cat, and subsequent biochemistry assays revealed normal renal values.
Because of the retrospective nature of this study and small sample size, there are several inherent limitations to this study. The most obvious limitation is there is no comparison of cats with aggressive surgery alone versus cats with surgery and adjunctive carboplatin. As this is a such an aggressive local and distant disease, all clinicians in this study offered adjunctive therapy after subtotal colectomies, and only one client was found in the medical records that opted for surgery only.
Conclusion
Feline colonic adenocarcinoma is a locally invasive and highly metastatic cancer. Cats often have metastasis at presentation; however, the cats often present for signs secondary to obstruction, warranting wide surgical excision. Cats remained patent after sub-total colectomy, colonic stenting and are euthanized due to metastasis, warranting early adjunctive treatment. Carboplatin, in this study, had minimal toxicity and could be a viable adjunctive treatment of this disease.
Kaplan-Meier disease-free analysis of cats treated with subtotal colectomies and adjuvant carboplatin.
Kaplan-Meier overall survival analysis for cats treated with subtotal colectomies and adjuvant carboplatin. The asterisk (*) indicates a censored cat.
Contributor Notes
