Abstract
Case series summary
A retrospective multicentre case series of feline primary erythrocytosis (PE) was evaluated. The aim was to gain better understanding of disease presentation and progression to guide management and prognostication. Case records were assessed for evidence of increased packed cell volume (PCV; >48%), sufficient investigation to rule out relative and secondary erythrocytosis, and follow-up data for at least 12 months or until death. Eighteen cats were included in the case series. No significant trends in signalment were noted. Seizures and mentation changes were the most common presenting signs (both n = 10). Median PCV was 70% (median total protein concentration of 76 g/l) with no other consistent haematological changes. Sixteen cats survived to discharge. Phlebotomy was performed initially in 15/16 surviving animals and performed after discharge in 10/16. Hydroxyurea was the most common adjunctive therapy, used in 10/16 cats. Of the 16 patients surviving to discharge, 14 patients were still alive at the conclusion of the study (survival time >17 months post-discharge), with the two non-survivors having lived for 5 years or more after diagnosis. PCV, when stabilised, did not correlate with resolution of clinical signs.
Relevance and novel information
In contrast to perceptions, feline PE was generally well managed via a combination of phlebotomy and medical therapy, with evidence of prolonged survival times. The use of hydroxyurea enabled cessation or repeat phlebotomies.
Introduction
Feline primary erythrocytosis (PE), also known as polycythaemia, is a rare myeloproliferative disorder resulting in increased red blood cell mass. 1 It is caused by an erythropoietin-independent clonal expansion of a single haematopoietic stem cell 1 and in humans is considered a form of myeloproliferative neoplasm. 2
Excessive erythropoiesis is suspected to be caused by mutations in the erythropoietic signalling cascade, resulting in constitutive downstream activation. 3 In humans, the main mutation responsible for primary polycythaemia (PP) is well characterised. Ninety-eight percent of human PP is linked to mutations in Janus Kinase 2 (JAK2), a tyrosine kinase, 4 of which 95% are a valine to phenylalanine substitution 5 in exon 14 of the JAK2V617F gene.6,7,8 An identical mutation has also been identified in a dog with PE. 9
Affected animals may present acutely with severe signs, often including seizure activity,1,10 and may require intensive initial support. Clinicians therefore need to provide rapid information to owners regarding the morbidity and mortality of this condition. Current knowledge of feline PE is based on isolated case reports, and while prognosis is given as ‘guarded’, survival times of up to 6 years have been reported. 1
The aim of this case series was to collate the largest set of data on feline PE so far, and thereby provide a comprehensive description of the disease course and possible treatment options to guide practitioners.
Case series summary
Patient records of client-owned cats diagnosed with feline PE from January 2006 to October 2015 were requested from veterinary hospitals across the UK. Five centres provided a total of 18 cases: Royal Veterinary College, London (10 cases); Willows, Solihull (four cases); Animal Health Trust, Newmarket (two cases); Vets Now, Glasgow (one case); Goddard’s Veterinary Group, London (one case).
Diagnosis of PE was based on marked increase in packed cell volume (PCV; >48%) 1 , with exclusion of relative erythrocytosis or other secondary causes such as underlying diseases resulting in hypoxia (eg, cardiac disease) or the potential for inappropriate erythropoietin production (eg, productive renal tumour). 11
Histories, initial clinical examination, and laboratory results were reviewed. Treatment protocols and further clinicopathological results were assessed during the period of hospitalisation. Follow-up information was obtained via email or telephone requests, including details of further treatment, ongoing clinical signs, changes in haematological values, and survival at time of data collection.
Signalment
Eight cats were female and 10 were male, all were neutered. The median age was 5 years and 7 months (range 35–126 months). Median body weight was 4.63 kg (range 2.6–6.4 kg). The majority of patients were domestic shorthair (DSH) cats (15/18). One cat had been successfully treated for immune-mediated haemolytic anaemia (IMHA, suspected primary) 1 year prior to presentation, and one cat had been treated for eosinophilic granuloma complex many years previously: neither were receiving medication at time of diagnosis. The remainder were reported as having unremarkable previous medical histories. A summary of the cases can be found in the online supplementary material (Appendix 1).
Presenting signs and clinical examination
Presenting clinical signs are detailed in Table 1. Neurological signs were the most common presenting complaint, affecting 16/18 patients, followed by gastrointestinal signs (6/18). The median average duration of clinical signs prior to presentation was 2 days (range 1–167 days). Physical examination findings are listed in Table 2, with the most notable finding being congested mucous membranes.
Presenting clinical signs in feline primary erythrocytosis cases
Physical examination findings on presentation in feline primary erythrocytosis cases
Clinicopathological data (at presentation)
Median PCV at presentation was 70% (range 64–83%), with a total protein (TP) concentration of 76 g/l (range 51–90 g/l). All cases with an available complete blood cell count demonstrated increased red blood cell concentration, haemoglobin concentration and haematocrit. Low mean cell haemoglobin (9/15) and lymphopenia (10/15) were the other frequent changes seen on haemogram (for more information, see Appendix 2 in the supplementary material). Evidence of increased erythrocyte production (eg, anisocytosis, polychromasia and rubricytosis) was frequently seen on blood smear analysis.
Serum biochemistry results were predominantly within the reference limits of the Diagnostic Laboratory Services at the Royal Veterinary College (10/17). Increased creatinine kinase (CK) activity was present in eight animals (range 525–103351 IU/l, reference interval [RI]: 52–506) and three cases had increased urea concentration (range 12.7–25.0 mmol/l, RI: 6.1–12.0). Four cats (unknown number tested) were reported to have evidence of hypoglycaemia on patient-side testing: unfortunately, numerical values were not recorded. Arterial blood gas analysis was only performed in two patients, neither of which revealed hypoxaemia.
Further laboratory testing included erythropoietin (EPO) assay (n = 11), retroviral testing (n = 8), thyroxine assay (n = 1), urinalysis (n = 8) and urine culture (n = 2). EPO levels were within given laboratory RIs in 7/11 cats, low in two cats and not reported in two patients. Seven out of eight cats were negative on in-house lateral flow immunochromatography retroviral testing (result not reported in one patient), and T4 level was within reference limits for the single cat tested. The only abnormalities documented on urinalysis were microscopic haematuria in four cats and growth of Escherichia coli from the urine of one cat. Non-invasive blood pressure was measured in six cats: all were normotensive (range 95–142 mmHg).
Diagnostic imaging
A summary of diagnostic imaging modalities and findings is presented in Table 3. Two cats were euthanased prior to full diagnostic imaging but underwent full post-mortem examination. No abnormal findings on imaging or post-mortem were considered sufficient to cause secondary PE via increased EPO production (appropriate or inappropriate). For example, renal lesions documented on ultrasound were not considered consistent with pathology likely to induce excessive EPO production. Similarly, degrees of cardiac functional impairment were considered minimal, and certainly insufficient to cause hypoxaemia, in all cats. No thoracic abnormalities were documented on imaging in any cats.
Imaging modalities and findings in feline primary erythrocytosis patients
Initial management and outcome
One cat was euthanased on the day of presentation due to severity of clinical signs (collapse and respiratory distress) and one cat was treated purely for seizures, with erythrocytosis only diagnosed retrospectively. Stabilisation of the erythrocytosis was therefore attempted in 16 cats.
All 16 cases underwent phlebotomy under sedation within 24 h of presentation. Median volume of blood removed was 12.5 ml/kg (range 4.5–17.9 ml/kg). Eight cats required a second phlebotomy during hospitalisation to achieve clinical stabilisation and sufficient PCV reduction (aiming for PCV <50–55%). Median total volume of blood removed was therefore 19.5 ml/kg (range 4.6–27.5 ml/kg). No adverse events were reported.
Three cats commenced treatment with hydroxyurea (HU) during the initial hospitalisation period. Dosages varied from 32 mg/kg PO q24h to 100 mg/kg PO every five days. Antiepileptic medication was started on presentation in five cats: levetiracetam (n = 3), phenobarbital (n = 1), or both drugs (n = 1). One cat underwent splenectomy.
Sixteen of 17 cats survived to discharge. The remaining cat was euthanased after 2 days of hospitalisation due to owner concerns over reported poor prognosis and quality of life with ongoing treatment. The average time of hospitalisation was 6 days (range: 1–16 days).
Ongoing treatment
A variety of treatments were used to reduce PCV and manage clinical signs of PE long term, as shown in Table 4 (a more detailed outline of treatment modalities is provided in Appendix 3 in the supplementary material). Treatment was usually deemed successful by the attending veterinarian when no clinical signs were seen with a PCV <60%.
Long-term follow-up data of feline primary erythrocytosis cases. Cases 4 and 5 did not survive to initial discharge and are therefore not included
DSH = domestic shorthair; DLH = domestic longhair; BSH = British Shorthair; MN = male neutered; FN = female neutered; HU = hydroxyurea; PCV = packed cell volume; PTS = euthanased; N/A = not applicable
Phlebotomy post-discharge was performed in 10/16 cats over a median time frame of 16.4 months (range 3–44 months), although one cat (case 2) only had a single phlebotomy performed 44 months post-discharge. The median number of phlebotomies performed post-discharge was seven (range 1–37), with an average interval of 6.5 weeks (range 1.5–16 weeks). One cat (case 11) underwent 37 phlebotomies over a period of 18 months, gradually increasing in interval from weekly (1.5–5 ml/kg) up to monthly (3–6 ml/kg), and was able to undergo conscious draining on a number of occasions owing to amenable temperament. Six cats eventually maintained a stable PCV without further phlebotomies after a variable period of time (range 3–24 months). Mild side effects were intermittently reported, including difficulty in extracting blood due to increased blood viscosity and progressive sclerosis of jugular veins.
Ten cats received treatment with HU, most commencing in the first month after diagnosis. Initial doses ranged from 9 mg/kg PO q24h to 113 mg/kg PO every five days, with titration dependent on sequential PCV and haematology results. Reformulated hydroxyurea capsules and a liquid formulation are available, which can facilitate administration of smaller doses. The median stable dose was 22 mg/kg PO q48h (range 10 mg/kg q24h to 45 mg/kg q48h). All cats receiving HU were able to stop phlebotomy treatment, either immediately or within 5 months of starting therapy.
Side effects of HU were reported in 6/10 treated cases. Methaemoglobinaemia was reported in three cats, with clinical signs including dyspnoea, tachypnoea, and cyanosis. These cats had received doses at the upper end of the dose range (>100 mg/kg per dose), with signs seen after first dose administration. Clinical signs responded to oxygen supplementation and N-acetylcysteine administration. Dose reduction and S-adenosylmethionine were effective in preventing recurrence. Evidence of oxidative damage to haemoglobin was documented in three cats (Heinz bodies ranged 30–48%). Four cats also demonstrated evidence of myelosuppression, with neutropenia in three cats and thrombocytopenia in one case. Neutropenia was mild in two cats (2.46 × 109/l and 2.21 × 109/l, RI 2.50–12.50) and moderate in one cat (1.78 × 109/l). The cats with mild neutropenia were receiving 35–45 mg/kg on alternate days, while the cat demonstrating moderate neutropenia was receiving 12.5 mg/kg daily. Thrombocytopenia was moderate in the affected cat (56 × 109/l, RI: 200–600), who was receiving 20 mg/kg HU daily. In all cases, concentrations normalised after reduction of HU dose (either by 50% or by moving to alternate day dosing).
Hirudotherapy (leech treatment) was attempted in three cats. In one patient (case 3), treatment appeared to cause pain and triggered seizure activity. In the other patient (case 7), hirudotherapy was performed twice with no obvious side effects, but provided insufficient reduction in PCV. Six cases required antiepileptic therapy, which was continued long term in three cats due to ongoing seizure activity. Phenobarbital was used in two cases, levetiracetam in three cases (one discontinued after 2 months) and both drugs in case 12: this cat was receiving no additional treatment for erythrocytosis. One patient (case 3) was prescribed aspirin (0.5 mg/kg PO q24h) due to development of thrombocytosis.
Ongoing monitoring
All cats surviving to discharge had at least 17 months of follow-up data. Fifteen patients demonstrated reduction of PCV from presentation over the course of treatment (mean reduction 16%, range 3–32%). Details of PCV changes for case 9 were not available. There was no standardised frequency of monitoring.
Thrombocytosis (platelet count >800 × 109/l) was reported in three cases. One cat exhibited intermittently increased platelet concentration from diagnosis onwards for the two and a half years of follow-up data available, up to a maximum of 1176 × 109/l. Two cats developed intermittent thrombocytosis 5 months post-discharge, rising to a maximum of 1360 × 109/l.
Leukocytosis occurred in two of the 16 surviving cases. One case (case 17) developed a transient mild lymphocytosis (8.4 × 109/l, RI: 0.2–5.4) 2 years after diagnosis. This patient was found to have concurrent Toxoplasma gondii infection. Case 1 developed marked, inappropriate rubricytosis with the presence of circulating erythroblasts and occasional abnormal myeloid precursors 60 months post-discharge, raising suspicions of a leukaemic transformation. The patient developed marked weight loss and polyphagia, and was euthanased due to concerns regarding prognosis and quality of life.
Long-term survival
One case (case 1) was euthanased 5 years after diagnosis due to potential leukaemia, and a further cat (case 2) died 4.75 years post-diagnosis, with comorbidities including chronic kidney disease and senile behavioural changes. Of the 14 remaining cases, all were alive at the time of data collection. Survival time post-discharge is shown in Table 4.
Continued clinical signs were reported in five cases: seizures (n = 3), periodic mentation changes (n = 1), and development of hind limb ataxia (n = 1). The ataxic patient was additionally diagnosed with toxoplasmosis, and so it is possible that residual clinical signs were not related to PE. A total of seven cats achieved a PCV below 50%. Of these, two experienced continued seizures. Of the nine cats with a stable PCV >50%, three experienced ongoing clinical signs, while the remaining six cats were asymptomatic. The mean PCV of cats with long-term clinical signs was 46%, compared with 62% for those with no long-term signs. All cats with ongoing clinical signs were perceived to still have a good quality of life by their owners.
Discussion
Neurological signs have previously been reported as a common presenting complaint in erythrocytotic cats.1,10,11 Incidence in this case series was 14/16 cats, slightly higher than previous estimations of 50–87%.1,10 Erythrocytosis is therefore a possible, if rare, differential for acute onset seizures in the cat. The suspected aetiology is cerebral hypoglycaemia 12 or hypoxia with possible ischaemic episodes, as cerebral blood flow is known to be reduced in humans once PCV exceeds 46–52%. 13
Imaging and laboratory findings were similar to those reported in the literature.1,11 A number of the most common changes included increased CK activity, hypoglycaemia and left ventricular hypertrophy. Increased CK activity may be caused by ischaemic myopathy or thrombotic events. Hypoglycaemia is known to occur with increased PCV due to decreased plasma volume for glucose transport, 12 and post-sampling due to increased glucose metabolism in vitro by erythrocytes. 14 Point-of-care glucometers have been documented to give lower results than reference analysers; however, these differences tend to be of minimal clinical significance, and therefore may not account for the hypoglycaemia documented here. 15 Low fructosamine values are likely to be more reflective of true chronic hypoglycaemia than direct glucose measurements in PE patients. There is known to be a significant overlap in EPO values between normal cats and those with both primary and secondary erythrocytosis. 16 EPO assay is therefore deemed to be of limited use in diagnosing PE. Mild ventricular hypertrophy was present in four cats (31% of those undergoing echocardiography), and may be incidental, or could reflect compensatory hypertrophy due to increased cardiac workload due to increased blood viscosity.
Phlebotomy has always been the key treatment of an acutely presenting erythrocytotic patient, 13 rapidly improving clinical signs. The average amount of blood removed to achieve in this case series was 19.5 ml/kg, equating to a reduction in PCV of approximately 15%. 17 Intermittent phlebotomy can be continued long term, with frequency determined by regular monitoring of PCV and clinical signs. Potential side effects of phlebotomy include hypoferritaemia, hypoproteinaemia, sedation risks, venous thrombosis 3 and patient stress, in addition to financial and time investment by the owner.
If phlebotomy is required at an unacceptable frequency or side effects are noted, medical therapy in the form of HU should be considered. HU is a ribonucleotide reductase inhibitor, reducing cellular proliferation by interfering with the de novo synthesis of deoxyribonucleotides. 18 HU also inhibits recycling of methaemoglobin back to haemoglobin, and can result in a significant increase in methaemoglobin levels, usually maintained at 1%. 19 Dyspnoea occurs when this level exceeds 25%. 19 Additionally, inhibition of DNA synthesis can cause reversible bone marrow suppression. 11 The data from this case series suggests that lower doses at a more regular frequency are associated with reduced incidence of symptomatic methaemoglobinaemia. Starting cats on daily or alternate day dosing and maintaining long-term dose at less than 20–30 mg/kg HU PO per day (or alternate day dosing) is therefore advisable.
Previous literature has advised to aim for a stable PCV of less than 50%. 1 This was only achieved in 7/16 of patients in this study; however, 6/9 remaining cats were asymptomatic long term. The average stable PCV was in fact higher in cats without ongoing clinicial signs than those with continued signs. It is therefore not possible to suggest a new therapeutic target from this data, however it would appear that PCV <50% is an unnecessarily low goal.
Only two cats appeared to achieve remission, defined in human medicine as clinical and haematological resolution without intervention for at least 3 months. 20 At present, a clear dialogue should be maintained with the owners to discuss the difference between achieving clinical and haematological control of the disease when deciding thresholds for intervention.
This study suggests that feline PE has a good prognosis for survival, with survival times of up to, and possibly exceeding, 5 years. Cats also maintain a good quality of life, with the patients in this case series either becoming asymptomatic or experiencing readily controlled clinical signs (eg, seizures managed with anti-epileptic medication).
Human literature reports 3–19% incidence of progression of PP to acute leukaemia, and 5–14% incidence of transformation to myelofibrosis. 21 One cat demonstrated possible malignant transformation; however, this was not confirmed with bone marrow examination. Longer-term follow-up data would be required to determine the risk of malignant transformation.
A notable difference between feline PE and the most common form of human PP is the absence of consistent trilineage hypercellularity in cats. Thrombocytosis and leukocytosis were documented in this case series in only a handful of patients, with neither abnormality occurring concurrently. This casts doubt on the likelihood of the prominent human JAK2V617F mutation as the underlying genetic cause. However, erythrocytosis is known to occur as the only haematological abnormality in a subset of gain-of-function JAK2 mutations affecting exon 12 as opposed to exon 14.21,22 This is believed to be the genetic cause of PP in approximately 3% of human patients. 23 Congenital erythrocytosis, suspected to be caused by a truncated EPO receptor,4,24 also causes pure erythrocytosis; however, the signalment of patients in this study is not supportive of a congenital basis for feline PE. JAK2 is now a target for novel medical treatments for polycythaemia. Ruxolitinib, a JAK1 and JAK2 inhibitor, appears to be well tolerated, and a study comparing efficacy in humans with HU therapy has recently been completed (the RELIEF trial). 13
Conclusions
This study provides information to aid practitioners in diagnosing and managing cases of feline erythrocytosis. Importantly, it should be noted that the historic term ‘polycythaemia’ is a misnomer: ‘erythrocytosis’ is a more accurate name, reflecting the lack of multilineage hypercellularity in this species.
In contrast to previous literature, this case series suggests a good outcome for treated cats in the medium to long term. Phlebotomy is required for immediate cytoreduction and can be combined with, or supplanted by, medical therapy. Elucidation of the underlying cause of feline erythrocytosis would likely result in novel targets for treatment, possibly with improved reduction in PCV and reduced adverse effects. This would require genetic comparison between affected and unaffected animals. Extrapolation from human medicine suggests that JAK2, with a possible focus on exon 12, could prove fruitful ground for further investigation.
Supplemental Material
Supplemental Material
Appendix 1
Supplemental Material
Supplemental Material
Appendix 2
Supplemental Material
Supplemental Material
Appendix 3
Footnotes
Acknowledgements
The staff, clients and patients at the following institutions are gratefully acknowledged: QMHA (Royal Veterinary College), Willows Veterinary Centre, Animal Health Trust, Goddard Veterinary Group and Vets Now (Glasgow).
Supplementary material
The following files are available:
Appendix 1: summary of feline PE cases (n = 18).
Appendix 2: common haematological changes in feline PE cases.
Appendix 3: long-term treatment and survival data for feline PE cases.
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.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
