Abstract
Objectives
The aim of this study was to evaluate the benefit of intracavitary carboplatin chemotherapy in cats with malignant pleural effusion of epithelial origin.
Methods
The medical records of cats with a cytological diagnosis of neoplastic pleural effusion of epithelial origin were reviewed at three referral institutions between January 2013 and June 2018. Only cats treated with intracavitary carboplatin chemotherapy were enrolled. Data collection included signalment, medical history, clinical signs, pleural effusion analysis, diagnostic imaging findings, intracavitary carboplatin chemotherapy protocol, adverse events, response to chemotherapy, outcome and underlying primary tumour, if possible.
Results
Eight cats met the inclusion criteria. Three cats had previous surgical removal of a tumour, including a poorly differentiated primary lung carcinoma, a uterine adenocarcinoma and a benign mammary tumour. The main clinical signs were tachypnoea and/or dyspnoea, inappetence and weight loss. Thoracic radiographs revealed marked bilateral pleural effusion in all cats. Pleural fluid analysis was consistent with a modified transudate, with malignant epithelial cells on cytology, leading to a diagnosis of pleural carcinomatosis. All cats received only one cycle of intracavitary carboplatin chemotherapy at a dose of 200–240 mg/m2. Recurrence of pleural effusion was reported in 7/8 cats within 4–15 days of chemotherapy, and death was recorded in all cats within 5–16 days, owing to recurrent pleural effusion or poor general condition. The primary cancer was suspected to be of pulmonary, mammary and pancreatic origin in four cats, two cats and one cat, respectively, and of unknown origin in the remaining cat.
Conclusions and relevance
In this study, intracavitary carboplatin chemotherapy seems ineffective in managing neoplastic pleural effusion of epithelial origin in cats with pleural carcinomatosis. Other cytotoxic drugs and/or techniques should be investigated in the future to improve the quality of life and survival of cats with pleural carcinomatosis.
Keywords
Introduction
Pleural effusion is frequently encountered in veterinary medicine and can be challenging to manage, depending on the underlying cause. In cats, pleural effusion secondary to neoplasia is a common finding, representing nearly 26% of all pleural effusions, 1 similar to its human counterpart (15–35%). 2 Lymphoma and thymoma are among the most typical tumours associated with pleural effusion in cats. In recent reports, effusion was present in 51% and 36% of cats with mediastinal lymphoma and thymoma, respectively.3,4 In these conditions, pleural effusion usually resolves with adequate treatment and does not require any additional intervention.
In some instances, neoplastic pleural effusion may result from mesothelioma, carcinomatosis or sarcomatosis. Carcinomatosis refers to the widespread metastatic dissemination of malignant epithelial tumours, with metastasis to the visceral and/or parietal pleura. 5 Metastatic disease will result in disruption of capillary integrity and obstruction of lymphatic vessels, which subsequently rupture and allow inflammatory reactions and exfoliation of tumour cells into the pleural fluid.5,6 Direct seeding of body cavities and surfaces is another way carcinomatosis may develop, with penetration of a serosal surface by tumour cells, and subsequent desquamation.6,7 In human oncology, primary lung, breast and ovarian cancers are mainly related to the development of malignant pleural effusion.5,6,8 In such cases, the removal of the underlying tumour is not always possible, and palliative therapy should be considered.
Multiple techniques have been described to treat neoplastic pleural effusion and, consequently, to improve quality of life and extend survival time of patients with cancer. To date, no consensus has been reached regarding the best treatment of pleural carcinomatosis in veterinary medicine. Repeated thoracocentesis and thoracostomy tubes, 9 talc pleurodesis using video-thoracoscopy, 10 pleuroperitoneal or pleurovenous shunts,10,11 thoracic omentalisation12,13 and the use of a modified vascular access port (PleuralPort; Norfolk Vet Products) have all been anecdotally described.9,14
Intracavitary chemotherapy (IC) has also been reported for the treatment of pleural and/or peritoneal carcinomatosis, sarcomatosis or mesothelioma in veterinary medicine.7,15–18 Two cats were treated with intracavitary carboplatin chemotherapy for mesothelioma (one pleural and one peritoneal) with survival times of 4 months and 6 months, respectively.16,17 Another publication discusses the use of IC with cyclophosphamide (1–4 cycles at 1 week intervals, dose of 250 mg/m2) in six cats with pleural carcinomatosis secondary to mammary tumours. 18 None of the treated cats responded to IC, with persistent pleural effusion in all cats, and a median time to progression and survival time of 0.6 months.
The objective of this study was to investigate the benefit of IC with carboplatin in cats with a diagnosis of pleural carcinomatosis. We hypothesised that IC may allow partial-to-complete resolution of neoplastic effusion, with improved quality of life and extended survival time.
Materials and methods
The medical records from three referral institutions (Oncovet, Villeneuve d’Ascq, France; Alliance, Bordeaux, France; Advetia, Vélizy-Villacoublay, France) were searched for cats with a cytological diagnosis of neoplastic pleural effusion of epithelial origin between January 2013 and June 2018. To be included in the study, cats had to be treated with carboplatin IC. Patients with incomplete records and pleural carcinomatosis that was not treated or treated with another cytotoxic drug were excluded.
Each record was examined for information pertaining to signalment; concurrent illness; clinical signs; physical examination; haematological and biochemical results; pleural effusion analysis; diagnostic imaging findings; dose of carboplatin administered; number of IC cycles; side effects; suspected primary tumour if possible; response to IC; follow-up; and outcome. Adverse events were recorded according to the Veterinary Cooperative Oncology Group guidelines. 19
For each cat, carboplatin (10 mg/ml; Hospira or Accord) was diluted in 200–250 ml/m2 sterile saline or 5% dextrose at the discretion of the clinician, and administered over 10 mins through chest tubes surgically placed under general anaesthesia or using ultrasound guidance. Before IC administration, malignant pleural effusion was completely removed to achieve the highest concentration of carboplatin. Chest tubes were withdrawn immediately after the procedure (when used for chemotherapy instillation), and cats were gently rolled to distribute the carboplatin solution throughout the pleural space. Cats were hospitalised and monitored for 24 h, as it is a regulatory requirement in France. Treatments were scheduled every 3 weeks.
Results of the descriptive analysis are expressed as mean ± SD.
Results
Signalment
Eight cats met the inclusion criteria: five cats from Oncovet, two cats from the Veterinary Clinic Alliance and one cat from the Veterinary Hospital Advetia. All cats were domestic shorthairs. There were three neutered males, four spayed females and an intact female. Age at diagnosis was 12 ± 4.4 years (range 2.2–18.9 years). Weight was 3.7 ± 1.2 kg (range 2.6–6.0 kg).
History and clinical presentation
Previous or concurrent illness was reported in 6/8 cats, including untreated hyperthyroidism (n = 1), restrictive cardiomyopathy and recently removed poorly differentiated primary lung carcinoma (n = 1), chronic kidney disease (n = 2), and previous surgical resection of a benign mammary tumour (n = 1) and of a uterine adenocarcinoma (n = 1, histological report unavailable).
Common presenting signs were tachypnoea and/or dyspnoea (n = 8), anorexia or dysorexia (n = 6), weight loss (n = 4), vomiting (n = 2), coughing (n = 1), hypodipsia (n = 1), syncope (n = 1) and fever (n = 1). Mean duration of clinical signs before presentation was known for 6/8 cats, and ranged from 3 days to 1 month. Further clinical signs observed during clinical examination were left parasternal systolic heart murmur (n = 2), muffled heart or lung sounds (n = 2), presence of mammary nodules (n = 2) with cutaneous ulceration in one of the cats, and presence of a thyroid nodule in the cervical region (n = 1).
Haematology and biochemistry
Complete blood count was performed in all cats and the most common abnormalities included mild-to-moderate neutrophilic leukocytosis (n = 7), mild thrombocytopenia (n = 1) and moderate thrombocytosis (n = 1). Serum biochemistry was performed in 8/8 cats and the main abnormalities consisted of moderate hyperglobulinaemia (n = 2), mild increase in uraemia (n = 2) and creatininaemia (n = 1) and marked increase in gamma-glutamyltransferase activity (n = 1). Calcaemia was investigated in 2/8 cats, and was within the normal reference interval. Feline immunodeficiency virus/feline leukaemia virus status was tested in 3/8 cats, and was negative in all three cats evaluated.
Thoracic radiographs and pleural effusion
Thoracic radiographs were performed in all cats as part of the initial investigations, and revealed moderate-to-severe bilateral pleural effusion in all cats with subsequent lung collapse. Additionally, multiple pulmonary nodules were observed in one cat, and a primary pulmonary mass was suspected in three other cats, as shown in Figure 1.

Thoracic radiograph (lateral view) of a cat with moderate pleural effusion and suspected lung mass in one of the caudal lobes
Thoracocentesis was performed in all cats as a diagnostic and therapeutic procedure. Pleural effusion was serosanguineous on gross appearance, and pleural fluid analysis was consistent with a modified transudate in all cats, with a total protein level ranging from 4.3 g/dl to 6.0 g/dl, and a cell count ranging from 780 cells/μl to 12,320 cells/μl. Cytological assessment of pleural effusion allowed a diagnosis of malignant effusion of epithelial origin in all cats, based on the presence of numerous clusters of cohesive cells demonstrating clear criteria of malignancy and moderate-to-marked anisocytosis and anisokaryosis, as shown in Figure 2. Macrophages, neutrophils and mature lymphocytes were also present in different proportions.

Cytological examination of the pleural fluid of a cat with pleural carcinomatosis and an associated primary lung mass, showing clusters of neoplastic epithelial cells with strong criteria of malignancy: large cohesive cells with a high nucleo-cytoplasmic ratio, round-to-oval nuclei with multiple prominent nucleoli, moderate-to-marked anisocytosis and anisokaryosis, and deeply basophilic cytoplasm with punctate vacuolation (May-Grünwald Giemsa staining, × 100)
Additional diagnostic imaging and suspected primary tumour
Echocardiography was performed in 6/8 cats, and confirmed restrictive cardiomyopathy in one cat as previously reported. Abdominal ultrasound was performed in 1/8 cat and was unremarkable. Thoracic and abdominal contrast-enhanced CT was performed in 4/8 cats and identified a lung mass in two cats (which was suspected on thoracic radiographs in both cases), with small pulmonary nodules consistent with metastatic disease, and a pancreatic mass in one cat with multiple pulmonary nodules. In this latter, ultrasound-guided aspiration of pancreatic and pulmonary lesions confirmed the presence of an exocrine pancreatic adenocarcinoma with lung metastases.
Based on history, physical examination and diagnostic imaging results, pleural carcinomatosis was assumed to be the consequence of a primary lung carcinoma in four cats, a mammary tumour in two cats, an exocrine pancreatic adenocarcinoma in one cat, and of unknown primary in the one remaining cat.
IC and outcome
Only one cycle of IC was given to each cat, in association with analgesics when required. Carboplatin was used at a dose of 200 mg/m2 in four cats and 240 mg/m2 in the four remaining cats. Half of the carboplatin dose was infused into each hemi-thorax. Chemotherapy instillation was performed through chest tubes in 5/8 cats and with ultrasound-guidance under general anaesthesia in 3/8 cats.
As anorexia was reported in 6/8 cats at presentation, supportive treatment was instituted in most of the cats after IC, with the use of maropitant citrate (n = 6), low-dose prednisolone (n = 6), mirtazapine (n = 2) or cyproheptadine (n = 1). Three to 8 days after IC, anorexia was persistent in four cats, despite supportive treatment. The owners did not consider feeding tube placement in any of these cats. Despite IC, tachypnoea and/or dyspnoea was reported in 6/8 cats, and progression and/or recurrence of pleural effusion was radiographically confirmed in 7/8 cats, within 4–15 days after chemotherapy (7.7 ± 3.3 days).
In all cats, death was recorded 5–16 days after the first IC (9.6 ± 3.3 days). Cause of death was euthanasia in six cats, with recurrent pleural effusion on thoracic radiographs in all of them and persistent anorexia in 2/6 cats. Death was attributed to persistent anorexia and pleural effusion in cat 5, and to persistent anorexia and weakness in the cat with a pancreatic carcinoma (cat 8). Data are summarised in Table 1.
Description of signalment, concurrent illness, diagnostic imaging findings, intracavitary carboplatin chemotherapy protocol, suspected primary cancer and survival time for each cat
IC = intracavitary chemotherapy; CKD = chronic kidney disease; MN = male neutered; FN = female neutered; F = female intact
Adverse events were difficult to assess during this short follow-up period. Complete blood count was repeated 7–10 days after IC in only two cats, and one of them developed grade 4 neutropenia (cat 8).
Discussion
As in humans, pleural carcinomatosis has a guarded prognosis in dogs and cats. To date, the only published article in cats dealing with IC gave a median survival time of 15 days in six cats presented with a diagnosis of pleural carcinomatosis of mammary origin and treated with cyclophosphamide. 18 In this same report, only 20% of cats survived at least 1 month after treatment, and none survived as long as 1.5 months. 18 Consequently, the results of our study are similar to those previously reported, with a mean survival time of 9.6 days and a median survival time of 8 days, and 0% of cats were alive 1 month after initiation of treatment. In human oncology, presence of malignant effusion is usually associated with a median survival that does not exceed 6–12 months, despite multiple treatment options.20,21 Owing to this, less invasive approaches are now used on an outpatient basis in order to minimise discomfort and cost. 21
Several hypotheses may be considered to explain the poor prognosis of feline pleural carcinomatosis in our study. First, most of the cats were presented with systemic clinical signs and/or a poor general condition, such as anorexia, weakness or weight loss, with a late presentation for some of them (up to 1 month). Moreover, concurrent illness was reported in some cats, including chronic kidney disease in two cats. As such, IC tolerance may have been affected with possible greater side effects in these cats. Secondly, given the persistence of anorexia and weakness in 50% of cats after IC despite supportive treatment, some owners may have not considered any additional treatment, leading to premature euthanasia. Finally, suspected primary tumours were of pulmonary, mammary and pancreatic origin in our study. Recent retrospective studies have been published discussing the aggressive nature of these carcinomas in cats.22–25
Surgical resection of primary lung carcinoma was associated with a median survival time of 11 and 156 days in 20 and 28 cats, respectively.22,23 Presence of pleural effusion at diagnosis was considered a negative prognostic factor, with a median survival time of 2–2.5 days only.22,23 Similarly, median survival time of exocrine pancreatic carcinoma was 97.5 days in 34 cats, 24 and median survival time of high-grade mammary carcinoma was only 8 months with surgery. 25 Most of these epithelial tumours are highly metastatic, with lungs, intrathoracic lymph nodes and pleura frequently involved.26–29 In our study, at least three cats were diagnosed with a tumour of advanced clinical stage and lung metastases. However, complete clinical staging with thoracic and abdominal CT scan was performed in only 50% of cats. It is thus likely that distant metastasis has been underestimated, with a probable role in poor prognosis and progression of the underlying disease. Moreover, surgical resection of the primary tumour is rarely attempted in these circumstances, which contributes to the persistence of clinical signs and short survival times.
In our study, the diagnosis of pleural carcinomatosis relied on routine cytological examination of pleural fluid. According to the cytological reports, the findings were strongly suggestive of a malignant epithelial tumour, and mesothelioma was not considered a main differential diagnosis. Nevertheless, distinction between epithelial and mesothelial tumours may sometimes remain difficult to assess on routine cytology, and the application of immunocytochemistry or immunohistochemistry on cavitary effusions can assist in determining the cell lineage.30,31 In human oncology, the diagnosis of pleural carcinomatosis relies on histological examination of pleura biopsy tissue obtained through surgery or thoracostomy. Pleural biopsies were not performed in any cat of this study as most of these cats were debilitated at presentation and a prompt diagnosis was required. However, video-thoracoscopy is becoming more available in veterinary medicine as a non-invasive procedure, and provides an excellent exploration of intrathoracic structures and adequate biopsy specimen, with rare complications. 32
Reports of IC are sparse in the veterinary literature.7,15–18 The largest studies were first published in dogs with the use of cisplatin, carboplatin and/or mitoxantrone for cases of pleural or peritoneal carcinomatosis, sarcomatosis and mesothelioma.7,15 Encouraging results were reported with complete resolution of malignant effusion in most dogs after 1–2 IC cycles, allowing extended progression-free survival (129 to >807 days) and median survival (332 days) compared with untreated dogs, with minimal toxicity. In our study, carboplatin was administered at a dose ranging from 200 mg/m2 to 240 mg/m2, according to previous reports using intravenous and IC routes.16,17,33 It is unlikely that this dose discrepancy would have affected response to IC, but the low number of treated cats in each group was too small to draw any final conclusions. In cats, cisplatin should not be used owing to its fatal pulmonary toxicity. 34 As carboplatin gave frustrating results in our case series, mitoxantrone should be investigated, perhaps combined with the use of the PleuralPort device to reduce the stress and discomfort of treated cats. 35 Similarly, talc pleurodesis has not been attempted in cats with malignant pleural effusion, and may be considered using video-thoracoscopy. Recently, autologous blood-patch pleurodesis was described in dogs with persistent pneumothorax, and may represent a suitable, inexpensive and safe option for treatment of malignant pleural effusion in companion animals. 36
The limitations of this study are inherent in its retrospective nature, with a small sample size and a non-homogeneous group regarding type of primary tumour, presence of distant metastasis and/or bulky disease, clinical staging, technique for IC administration and delivered dose of carboplatin. Multiple clinicians and pathologists were also involved in the diagnosis and management of these cats given the multi-institutional recruitment. Finally, given the rarity of benign mammary tumours in cats, a possible misdiagnosis of the removed mammary tumour in cat 5 must be considered, but histopathological slides were not available for reassessment.
Conclusions
In this study, neither durable alleviation of clinical signs nor significant improvement in survival times were achieved with intracavitary carboplatin chemotherapy in cases of feline pleural carcinomatosis, but larger studies would be required to confirm these preliminary results. Other chemotherapeutic drugs and/or techniques should be investigated in the future, to provide palliation and increase quality of life. The use of the PleuralPort device in combination with mitoxantrone administration or the use of talc pleurodesis under video-thoracoscopy may be suitable alternative approaches.
Footnotes
Acknowledgements
We would particularly like to thank Dr Eve Ramery (VetCyt, France) for the courtesy of allowing us to use the cytological picture, and the boarded-certified radiologists who reviewed all the diagnostic imaging records.
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.
