Clinical Characteristics of Hepatocellular Carcinoma in 19 cats from a Single Institution (1980–2013)

Introduction

With an estimated prevalence ranging from 1–3% of all feline neoplasms, primary hepatic tumors are uncommon in the domestic cat. Compared to the dog, the cat has a lower risk for HCA, as well as for metastatic hepatic neoplasia.^1–5 Even though HCA is considered the second most common primary liver tumor in cats, there has been limited clinical characterization of this neoplasm.^1–7 Despite an association between viral infections and HCA in other species (e.g., hepatitis B and C in humans and woodchuck hepatitis virus in woodchucks), there is no evidence for a viral association in cats.^1–4

Clinical signs of feline hepatobiliary tumors are often non-specific. For example, one study of hepatic neoplasia in 21 cats (including two with HCA and 15 with bile duct adenocarcinoma or adenoma) reported common findings of anorexia, lethargy, and weakness.³ HCA has received far more attention in the dog where it is frequently affiliated with increased liver enzyme activities (particularly serum alanine aminotransferase [ALT] and alkaline phosphatase [ALP]) and an aspartate aminotransferase (AST)-to-ALT ratio <1.^8–10 Previous studies of malignant primary hepatic neoplasia in cats show a similar association with increased hepatic transaminase activities; however, cats are more likely to demonstrate hyperbilirubinemia because biliary carcinomas commonly cause cholestatic jaundice.^1,3

Given the dearth of information regarding feline HCA, there are few details regarding patient prognosis or survival. For dogs, positive prognostic factors for HCA include surgical resection and left-sided liver lobe involvement.¹⁰ Negative prognostic factors for canine HCA include increased ALT and AST activities, and increased ALP-to-AST and ALT-to-AST ratios.¹⁰ Neither completeness of surgical resection nor tumor size appears to influence survival in dogs, with the exception of massive HCA where liver lobectomy substantially improves survival (i.e., median survival for non-surgical management of 270 days versus survival after liver lobectomy of >1,460 days).¹⁰ Survival statistics for cats undergoing resection of primary malignant liver tumors are dismal, with 60% of cats dying during the immediate post-operative period.¹ However, in the small number of case series reported, most cats suffered from tumor-impaired bile flow (i.e., biliary tree obstruction) thought to impose a substantial negative impact on peri-operative survival because of intractable hypotension.¹¹

The goals of this retrospective study were to describe clinical features, clinicopathologic parameters, imaging characteristics, hepatic mass lesion size and lobe distribution, concurrent health disorders, and survival in cats with HCA.

Materials and Methods

Using a series of search strings, electronic databases of the New York State College of Veterinary Medicine at Cornell University were explored for feline diagnoses coded as HCA, hepatic adenoma, and hepatocellular atypia or dysplasia from January 1, 1980, to January 1, 2013. Study inclusion required availability of hepatic histologic sections to allow confirmation of microscopic features consistent with a diagnosis of HCA as well as access to the patient's signalment, clinicopathologic data, imaging features, medical and surgical treatments, and survival duration. Neoplasms characterized as HCA were invasive, demonstrated loss of lobular architecture, had variable cellular atypia, and frequently contained foci of necrosis and hemorrhage. Hepatocellular adenomas, characterized as expansile non-invasive neoplasms, causing distortion of lobular architecture but with minimal cellular atypia, were excluded. Likewise, lesions interpreted as hepatic nodular hyperplasia, typically <1 cm in dimension, composed of focal or multifocal nodules of proliferating hepatocytes with occasional architectural dysplasia (e.g., tortuous hepatic cords), but that retained lobular architecture, were also excluded. All sections of interest identified by the search strings (51 liver biopsies) were independently reviewed by two board certified anatomic pathologists (B.P.B., S.P.M), ultimately confirming 19 cats with HCA.

Clinical data were collated for each case, including signalment, co-existent health problems, clinical signs, method of definitive diagnosis (antemortem biopsy or postmortem examination), presenting clinical features, clinicopathological parameters (complete blood cell count, serum biochemical profile), coagulation tests (prothrombin time [PT], activated partial thromboplastin time [APTT], fibrinogen concentration), FeLV and FIV status, imaging findings (thoracic and abdominal radiographs and abdominal ultrasonography [performed by experienced board certified radiologists]), presence or absence of abdominal effusion, interpretation of mass aspirate or imprint cytology, HCA liver lobe distribution, largest measured mass dimension (cm), surgical procedures (resection or non-resection biopsy), and survival from date of definitive diagnoses. Contemporary standard of care was offered for each animal, although some pet caretakers elected euthanasia for their cat.

Results

Of 19 cats with HCA, median age was 14 (2–21) yr with breeds distributed as 12 domestic shorthair, 3 Persian, 3 domestic longhair, and 1 Siamese. Thirteen cats (68%) were >10 yr of age. Gender distribution included 11 neutered females and 8 males (1 intact, 7 castrated); there was no significant gender predisposition.

Most cats (16/19) had one or more co-existent or pre-existent health disorders. These disorders included chronic lymphocytic rhinitis and hyperthyroidism treated with radioiodine; hyperthyroidism treated with methimazole; hyperthyroidism treated with radioiodine; chronic rhinitis; malabsorptive intestinal disease; FeLV-associated thymic lymphoma; metastatic pulmonary adenocarcinoma and severe non-suppurative cholangiohepatitis; purulent bronchiolitis; hyperthyroidism treated by thyroidectomy, chronic renal disease, and cardiac failure; end-stage glomerulonephritis, seizures, and blindness; renal disease, hypertrophic cardiomyopathy, and thyroid adenoma; squamous cell carcinoma of the mandible and thyroid adenoma; primary copper-associated hepatopathy; non-suppurative cholangiohepatitis associated with pathologic hepatic copper accumulation; chronic inflammatory bowel disease and chronic renal disease; and metastatic functional thyroid carcinoma.

One of the most common co-existent or antecedent diseases was hyperthyroidism. Of 13 cats with measured total T4 concentrations, 2 were hyperthyroid at the time of diagnosis of HCA and 4 had been treated for hyperthyroidism (radioiodine therapy [2], thyroidectomy [1], methimazole [1]) at 0.4, 2, 3, and 4 yr before HCA diagnosis. Additionally, one cat had a functional metastatic thyroid carcinoma confirmed by necropsy examination, and two cats had thyroid adenomas identified on necropsy examinations without antemortem thyroid hormone measurements. Two cats had hepatic copper concentrations measured by atomic absorption spectroscopy after hepatic copper accumulation was confirmed by microscopic examination of tissue sections stained with rhodanine. Liver copper concentrations in non-neoplastic tissue exceeded 1,000 μg/gm dry weight basis (reference range <180 μg/gm dry weight basis) in these cats.^12–14 One cat had primary copper-associated hepatopathy with panlobular copper distribution in hepatocytes including neoplastic tissue. The other cat had secondary hepatic copper accumulation associated with severe non-suppurative cholangiohepatitis where copper had accrued in hepatocytes abutting portal inflammation.

Eight cats had antemortem diagnosis of HCA with tissue collected during surgical excision (n=6) or biopsy (n=2), whereas 11 cats had HCA confirmed on postmortem examination. Unsuspected liver tumors were identified at necropsy in eight cats that were either euthanized or died from some other disease process. Of cats with HCA confirmed by postmortem examination, death was directly attributable to hepatic neoplasia in three cats that had mass lesions recognized antemortem, and due to concurrent illnesses in the remainder. Metastatic HCA was not recognized in any cat. Cats diagnosed with HCA antemortem were evaluated for regional lymphadenopathy, mass lesions in multiple liver lobes, and pulmonary nodules suggesting metastatic or primary neoplasia using radiographic and ultrasonographic imaging. In cats undergoing hepatic mass excision, visual inspection of abdominal structures failed to disclose lesions suggestive of metastatic neoplasia. Antemortem biopsy of regional hepatic lymph nodes in two cats failed to disclose metastasis. Postmortem examinations also consistently failed to reveal gross or microscopic metastases. In one cat, recurrent hepatic masses were discovered by ultrasonographic imaging 3 yr after HCA resection. Despite formation of large nodules (2.6 × 3.6 × 3.4 cm in central liver, 2.7 × 3.5 × 4.2 cm in right caudodorsal liver), this cat survived for an additional 3 yr without clinical illness. Unfortunately, histologic characterization of these recurrent masses was unconfirmed.

Common presenting clinical features associated with feline HCA included weight loss (95%), hyporexia (53%), lethargy (32%), historically increased liver enzyme activities (26%), diarrhea (21%), vomiting (21%), and palpable abdominal mass (21%). Comorbid conditions common in elderly cats often obfuscated characterization of clinical features that could be attributed to hepatic neoplasia. In 4 out of 19 (21%) cats, it was difficult to determine if hepatic neoplasia, concurrent disease, or both generated clinical signs. Most cats diagnosed with hyperthyroidism had been successfully managed with anti-thyroid drugs, thyroidectomy, or radioiodine before discovery of the hepatic neoplasm, negating the impact of this endocrinopathy on clinical features.

Common clinicopathologic parameters included increased activities of liver enzymes (Table 1). The most commonly increased liver enzyme was AST in >90% of cats. Increased activities of ALT and ALP were seen in >71%, and hyperbilirubinemia in >38 % of cats with HCA. Of 12 cats assessed for coagulation status, prolonged PT and APTT were found in 3 out of 8 and 3 out of 8 cats, respectively. Marked increases in both PT (>90 s; reference range, 15–20 s) and APTT (> 90 s; reference range, 14.5–19.0 s) were encountered in two cats with large (>7.2 cm largest dimension) heteroechoic hepatic masses that contained hypoechoic regions histologically characterized as blood-filled necrotic foci. Mildly increased PT (21.1 s; reference range, 15–20 s) was identified in one cat, and mildly increased APTT (30 s; reference range, 14.5–19.0 s) was seen in another cat. One of these cats also had a large (>4.8 cm) heteroechoic HCA containing many hypoechoic, blood-filled necrotic foci. One of four cats tested had increased fibrinogen concentration, whereas the remainder had fibrinogen concentrations within reference limits. One cat with the markedly increased PT and APTT underwent successful tumor excision following peri-operative treatment with vitamin K (1.5 mg/kg intramuscular q 12 hr for 3 treatments) and fresh frozen plasma. The coagulation times normalized following surgery. Because comorbid conditions frequently complicated the clinical status of cats with HCA, clinicopathologic abnormalities likely reflected both the co-existent health conditions and HCA. Only 1 out of 19 and 0 of 16 cats tested for FeLV and FIV, respectively, were positive. The FeLV-positive cat was 2 yr old and had well-differentiated HCA and thymic lymphoma.

Table 1 Hematologic and Serum Biochemical Parameters

View Fullscreen

Thoracic radiographs (that included the region of the entire liver) and/or abdominal radiographs allowing assessment of hepatic margins were available from 14 cats. Among these, a hepatic mass was suspected in only one cat. Abdominal ultrasonography completed in 14 cats disclosed localized hepatic mass lesions characterized as having heterogeneous echogenicity and hypoechoic (possibly fluid-filled) nodules in nine cats and isoechoic masses in two cats. Hepatic masses were not detected in three (21%) cats. Three cats had abdominal effusions that were characterized in one cat each as a pure transudate, modified transudate, or exudate. Concurrent disorders in these cats included primary copper-associated hepatopathy, cholangiohepatitis with secondary copper accumulation, and chronic lymphoplasmacytic rhinitis, respectively.

Cytologic interpretation of aspirates (5) or impression smears (1) from hepatic mass lesions in six cats was diagnostic for HCA in only two (33%, one aspirate and the impression smear). Hepatic distribution of HCA was specified in 15 of 19 cases; seven involved right-sided lobes, six involved left-sided lobes, and two involved multiple mass lesions in several liver lobes. Dimensions of hepatic mass size were reported for 17 of 19 cats. The largest dimensional measurement for cats ranged from 1–9 cm with 47% exceeding 4 cm.

Median survival of eight of ten cats diagnosed antemortem and not immediately euthanized was 1.4 (0.6 to 6.5) yr. Six cats underwent complete HCA resection as judged from surgeon opinion and microscopic demonstration of tumor-free surgical margins (0.5 cm minimum). Median survival in these cats was 2.4 (range, 1.0 to 6.5) yr. All of these cats had clinical signs attributed to hepatic neoplasia with each demonstrating improved general health and clinicopathologic parameters post-operatively. The longest surviving cat had its HCA resected at 12 yr of age. Although recurrent hepatic masses were identified ultrasonographically at 15.5 yr of age, that cat survived free of clinical signs until 18.5 yr old. Only one cat undergoing HCA excision received chemotherapy (90 mg/m² carboplatin IV q 30 days for 5 treatments during the initial post-operative interval). During the following 4.5 yr, no hepatic mass lesions recurred, but the cat developed chronic renal insufficiency associated with calcium oxalate ureterolithiasis. Of the remaining four cats with complete HCA excision, one was lost to long-term follow-up, one died of primary copper-associated hepatopathy, and two cats remained alive at time of manuscript submission. Signs of illness and increased liver enzyme activity in the cat with primary copper-associated hepatopathy responded to chelation therapy with d-penicillamine (10 mg/kg per os q 12 hr for 3 mo), vitamin E (10 U/kg per os q 24 hr) supplementation as an antioxidant, and feeding of a commercial feline diet for hepatic insufficiency. However, on withdrawal of d-penicillamine, rapid worsening of inappetence, lethargy, weight loss, and weakness led to euthanasia. It remains unclear if cessation of d-penicillamine therapy was responsible for this cat's deterioration.

One cat with a large HCA (mass dimensions 6.8 × 4.7 × 6.4 cm) diagnosed by surgical biopsy survived for 6 mo before being euthanized because of anorexia, weight loss, lethargy, and weakness. The other cat with HCA diagnosed by surgical biopsy also had non-suppurative cholangiohepatitis associated with severe hepatic copper accumulation and survived for 1 yr.

At the time of manuscript submission, two cats remained alive at 15.5 and 16.5 yr of age with survivals to date of 1.9 and 4.9 yr after HCA resection; 16 cats have been euthanized; and one cat was lost to follow-up after a documented 4 yr post-operative survival.

Discussion

This study characterizes the largest data set of domestic cats with histologically confirmed HCA published to date. During the 33 yr time span of this retrospective study at our institution, HCA was histologically diagnosed in 446 dogs, but in only 19 cats, substantiating that feline HCA is comparatively rare. However, the prevalence of this neoplasm in the feline population studied remains undetermined because not all cats with ultrasonographically detected hepatic mass lesions underwent definitive diagnosis and necropsies were not completed on all deceased cats. The latter point is especially important because many of the cases reported herein were incidentally recognized during necropsy.

We found no gender or breed predilection for feline HCA. However, cats with HCA had a median age of 14 yr with 68% of cats older than 10 yr. Historical and presenting clinical signs were non-specific including weight loss, hyporexia, lethargy, diarrhea, and vomiting. In many cats, these clinical signs may have reflected concurrent health problems. Despite finding that 47% of cats had at least one HCA dimension exceeding >4 cm, an anterior abdominal mass was palpated in only 21% of all cats.

Increased serum enzyme activities of ALT, AST, ALP, and gamma glutamyltransferase were common with increased activity of AST in >90%, increased ALT and ALP in > 71%, and hyperbilirubinemia in 38%. Marked increases in AST and ALT activities in some cats likely reflected necrotic foci histologically recognized within neoplastic tissue. The small study size and lack of an age-matched population of cats with and without other liver disorders preclude assessment of test diagnostic utility. Furthermore, concurrent disorders undoubtedly influenced many clinicopathologic parameters in the cats studied.

Limited coagulation testing in cats of this study restricts conclusions regarding tumor-associated coagulopathy. However, finding profoundly prolonged clotting times in two cats with large tumor burdens suggests that HCA may lead to coagulation disorders. It is clinically relevant that one of these cats underwent successful tumor excision while receiving peri-operative vitamin K and fresh frozen plasma. Infectious disorders (i.e., FeLV, FIV) were not linked with HCA in the population studied.

Thoracic and abdominal radiographs were rarely informative regarding HCA. However, abdominal ultrasonography typically disclosed hepatic parenchymal abnormalities ranging from mixed parenchymal echogenicity to hypoechoic nodules that reconciled histologically with blood-filled necrotic foci within neoplastic tissue. Ultrasonographic imaging by an experienced operator failed to identify hepatic masses in 21% of cats. The inability to detect these liver tumors likely reflected absence of a distinct echogenic interface between normal and neoplastic tissue, absence of neoplasia-associated necrotic or neo-vascularized foci large enough for ultrasonographic detection, or variables compromising thorough interrogation of each liver lobe (e.g., poor animal cooperation, overlying enteric gas, interfering gastric ingesta).

Cytologic evaluation of aspirate or imprint samples from HCA often failed to definitively diagnose neoplasia. Thus, tissue biopsy remains the gold standard for definitive diagnosis of feline HCA. Preferential location of feline HCA within the liver was not apparent; two cats had multiple HCA distributed among several liver lobes, as has been described in dogs.^4,9 Although dimensions of HCA were broadly variable with some tumors ranging up to 9 cm in largest dimension, massive HCA as described in dogs (i.e., expansive masses often invading and fusing adjacent liver lobes) was uncommon.¹⁰ However, similar to dogs, feline HCA does not appear to rapidly metastasize to local lymph nodes or other organs but rather expands locally by invading adjacent hepatic parenchyma.^4,10

Many antecedent or concurrent disorders common to elderly cats confounded interpretation of clinical and clinicopathologic findings. In this cohort, the most commonly observed concurrent disorders were hyperthyroidism, inflammatory bowel disease, and cholangiohepatitis. Although hyperthyroidism was a common antecedent or comorbid disease process, most cats had their endocrinopathy successfully managed with anti-thyroid drugs, thyroidectomy, or radioiodine before HCA was discovered. Nevertheless, it is interesting to speculate that increased metabolic activity and oxidative stress linked with hyperthyroidism might have provoked formation of dysplastic hepatocellular foci. A more expansive study to estimate relative risk associated with hyperthyroidism would require study of a larger case population including age-matched controls. The comorbid association between inflammatory bowel disease or cholangiohepatitis and HCA in some cats might reflect the influence of inflammatory cytokines and hepatocellular injury, as incriminated in humans.¹⁵ Discovery of copper-associated hepatopathy in two cats with HCA was unprecedented, as copper-associated liver injury is rare in feline patients.^12–14 It is possible that the chronic necroinflammatory hepatitis associated with hepatic copper accumulation might contribute to neoplastic transformation, as has been suggested in human beings with Wilson's disease and animal models of that disorder.^15–17

Conclusion

In summary, this study suggests that HCA is a rare feline neoplasm affecting elderly cats (median age 14 yr) with clinical signs and clinicopathologic features sometimes obscured by comorbid conditions. As in other species, definitive diagnosis cannot be reliably deduced from clinical signs, clinicopathologic parameters, cytology, or imaging studies. Feline HCA appears slow to metastasize, expands locally into hepatic parenchyma, and was effectively palliated by surgical resection in six cats in this study. Surgical resection resulted in improved well-being and in survival times ranging from 1 to 6.5 yr despite advanced age at initial diagnosis.