Introduction

Plasmapheresis, the therapeutic removal of plasma in exchange for replacement solutions, is an extracorporeal blood purification technique that has become increasingly available to veterinarians in recent years. Also known as therapeutic plasma exchange (TPE), its novel therapeutic advantages provide a treatment option for patients with advanced disease that otherwise may not survive. Immune-mediated hemolytic anemia (IMHA), a widespread disease resulting in the destruction of red blood cells, is one of such conditions. Traditional management for IMHA includes the exclusion of associative causes (e.g., infectious disease, neoplasia), immunosuppression with corticosteroids, consideration of a secondary immunomodulatory drug, and supportive care, including thromboprophylaxis and the maintenance of adequate arterial oxygen delivery through the provision of red blood cells.¹ However, TPE can rapidly remove anti-erythrocyte autoantibodies, circulating complement components, and immune complexes, serving as an adjunct treatment option with unique benefits not provided by traditional management.^2,3

A major complication of severe IMHA is the development of hyperbilirubinemia. This bilirubin elevation has the potential to result in a rapid onset of severe progressive neurological signs, often termed acute bilirubin encephalopathy (ABE), or kernicterus in its chronic form.⁴ This condition can result in a comatose state, refractory seizures, and cardiopulmonary arrest.⁵ To date, there is a single case report of a dog with severe ABE secondary to IMHA that was successfully managed with the provision of once daily membrane-based TPE treatments over 3 days.⁶ The dog was clinically improved following the third treatment and was discharged after a total of six packed red blood cell (pRBC) transfusions and 11 days of hospitalization.⁶

This report describes a critically ill dog with severe ABE and IMHA who failed to adequately respond to a previously reported protocol of three daily treatments with TPE.⁶ However, continued intensive care was provided, and disease stabilization was successfully achieved following a fourth plasmapheresis treatment. This is the first report of positive outcome in a dog requiring more than three TPE treatments, with or without the complication of ABE. Informed consent was obtained from the owner, and the patient was managed according to contemporary standards of care. The diagnosis, treatment, and ultimate recovery are described.

Case Report

An 8 mo old spayed female mixed-breed dog was referred to the Mathew J. Ryan Veterinary Hospital, University of Pennsylvania for plasmapheresis and intravascular IMHA. The dog had been adopted 3 mo before from Anguilla, an island in the eastern Caribbean Sea. One month before presentation, a SNAP 4DX^a test was positive for Anaplasma, and a 4 wk course of doxycycline^b (6 mg/kg per os [PO] q 12 hr) was prescribed. Two days following an ovariohysterectomy, pale mucous membranes were noted, and she was taken to a referral hospital (day 1). A complete blood count revealed a moderate anemia (packed cell volume [PCV], 22%; reference interval [RI] 36%–55%) with marked spherocytosis and a mild thrombocytopenia (platelet count, 120 × 10³/μL; 148–484 × 10³/μL). Saline agglutination was positive. Serum biochemistry revealed mild hyperbilirubinemia (total bilirubin [TBILI] 1.9 mg/dL; RI 0.0–0.9 mg/dL). Thoracic radiographs were unremarkable, an abdominal ultrasound identified splenomegaly, and a vector-borne pathogen panel^c was submitted. The dog was hospitalized and started on dexamethasone sodium phosphate^d (0.2 mg/kg IV q 24 hr), doxycycline (6 mg/kg PO q 12 hr), pantoprazole^e (1 mg/kg IV q 12 hr), maropitant^f (2 mg/kg PO q 24 hr), ondansetron^g (0.3 mg/kg IV q 8 hr), gabapentin^h (8 mg/kg PO q 12 hr), and clopidogrelⁱ (1.6 mg/kg PO q 24 hr). An 11 mL/kg pRBC transfusion was administered. The following morning (day 2), the dog’s TBILI increased to 17.5 mg/dL, and her PCV was 24%. The dog developed an abnormal mentation and hemoglobinuria. An infusion of IV immune globulin (IVIG)^j (0.5 g/kg IV once) was administered. Her neurological status continued to deteriorate, and a recheck TBILI was 48.1 mg/dL 10 hr later. The dog was subsequently referred to the Mathew J. Ryan Veterinary Hospital, University of Pennsylvania.

At presentation, the dog was stuporous to comatose with opisthotonos and forelimb extension. She was nonambulatory tetraparetic and had an absent menace response. The PCV was 17%, and the ammonia concentration was normal (61 μg/dL; RI 0–98 μg/dL). A dose of mannitol^k (0.5 g/kg IV once) was administered, though no improvement in mentation was noted. A second 11 mL/kg pRBC transfusion was started.

The dog was premedicated with midazolam^l (0.25 mg/kg IV once) and fentanyl^m (5 mcg/kg IV once), and general anesthesia was induced with propofolⁿ (3 mg/kg IV once), followed by endotracheal intubation and maintenance of anesthesia via inhalational isoflurane^o titrated to effect (0.5%–1.5%). An 11.5 French 24 cm hemodialysis catheter^p was placed, and the dog was subsequently extubated before the treatment. No further sedation or anesthesia was required for the TPE treatments. The Spectra Optia^q platform (priming volume of 141–185 mL) was used to perform a centrifugal TPE treatment exchanging two plasma volumes over 3 hr. The plasma volume was calculated according to the following formula: plasma volume (mL) = total blood volume (90 mL/kg × body weight) × (1 − PCV%). An exchange of two plasma volumes were chosen based on an anticipated removal of 86% of intravascular autoantibodies from circulation; a plasma exchange of one to two total plasma volumes is the traditional target for each TPE treatment.^7,8 Treatment started 7 hr following its most recent TBILI measurement. Because of the dog’s anemia and small body size (body weight 12.4 kg), it was elected to prime the extracorporeal circuit with a combination of 70 mL pRBCs, 60 mL fresh frozen plasma (FFP), and 70 mL 0.9% NaCl^r to improve hemodynamic stability during the treatment. Regional citrate anticoagulation (anticoagulant citrate dextrose, solution A [ACDA])^s was used in accordance with the Spectra Optia operating system at a starting rate of 1.0 mL/min/L total blood volume (TBV). A continuous rate infusion (CRI) of calcium gluconate^t (50 mg/kg/hr) was started to maintain serum ionized calcium >0.8 mmol/L and was adjusted as needed. The replacement volume consisted of 25% FFP, 25% frozen plasma (FP, FFP stored >12 mo), and 50% Plasmalyte^u. A single dose of 5 mL/kg 7.2% hypertonic saline^v was administered at the start of treatment given the concern for intracranial hypertension. A bolus of unfractionated heparin^w (25 U/kg IV once) and a CRI (25 U/kg/hr) was started following treatment initiation because blood clotting was noted in the circuit chamber. Two hours into treatment, the dog was hypertensive (Doppler^x blood pressure, 186 mm Hg) and bradycardic (70 beats/min), and a second mannitol bolus (0.5 g/kg IV once) was administered. Moderate hypokalemia was noted (2.9 mmol/L; RI 4.0 − 5.2 mmol/L) and a KCl^y CRI was started at 0.25 mEq/kg/hr. The posttreatment PCV was 16%, and the saline agglutination was negative (Figure 1). A third 11 mL/kg pRBC transfusion was administered. The dog remained stuporous to comatose with opisthotonos and forelimb extension. An unfractionated heparin CRI (37.5 U/kg/hr) was continued in addition to dexamethasone sodium phosphate (0.2 mg/kg IV q 24 hr) and gastrointestinal support. Levetiracetam^z (30 mg/kg IV q 8 hr) was started for seizure prophylaxis, and mycophenolate^aa (10 mg/kg IV q 12) was added as a secondary immunosuppressant agent. Rhythmic head bobbing was noted after recovery, and a dose of midazolam (0.25 mg/kg IV once) was administered.

View Figure

• Download as Powerpoint
• Download Figure

The following morning (day 3), continued head bobbing was appreciated, and a CRI of midazolam was started at 0.25 mg/kg/hr. The dog remained stuporous and opisthotonic with forelimb extension. Anisocoria and a vertical nystagmus were noted. An episode of systemic hypertension occurred (Doppler blood pressure 190 mmHg) and mannitol (0.5 g/kg IV once) was administered. A complete blood count showed a moderate thrombocytopenia (62 × 10³/μL; RI 186–545 × 10³/μL), regenerative anemia (PCV 30%, absolute reticulocytes 150.3 × 10³/μL; RI 11–92 × 10³/μL), and an inflammatory leukogram (WBC 25.3 × 10³/μL, RI 5.7–14.2 × 10³/μL; neutrophils 20.75 × 10³/μL, RI 2.7–9.4 × 10³/μL; band neutrophils 0.25 × 10³/μL, RI 0.0–0.1 × 10³/μL). Serum biochemistry showed mild liver enzyme elevations (aspartate aminotransferase 280 U/L, RI 23–65 U/L; alkaline phosphatase 221 U/L, RI 20–155 U/L; gamma-glutamyl transferase 40 U/L, RI 7–24 U/L) and a TBILI of 41 mg/dL (RI 0.1–0.5 mg/dL). Saline agglutination was once again positive. A second TPE treatment was started exchanging two plasma volumes following similar regional citrate anticoagulation, calcium CRI, and replacement volume protocol. The pretreatment PCV was 28%; therefore, the extracorporeal circuit was only primed with 0.9% NaCl and did not include pRBCs or plasma, given the anticipated improved hemodynamic stability and tolerance of the treatment. Continued head bobbing was noted, and a midazolam bolus (0.25 mg/kg IV once) was administered. The posttreatment PCV was 18%, TBILI was 28.3 mg/dL, and saline agglutination was negative. The unfractionated heparin CRI was discontinued following treatment given the concern for bleeding from the catheter site. The dog’s mentation improved from stuporous to obtunded.

The next morning (day 4), the dog remained obtunded but was aware and no longer opisthotonic. The PCV was 12%, and a fourth pRBC transfusion (8 mL/kg) was administered. TBILI showed mild improvement at 23.7 mg/dL, and blood ammonia was normal (15 μmol/L, RI 11–35 μmol/L), though a low total protein (4.3 g/dL, RI 5.4–7.1 g/dL) and albumin (2.0 g/dL, RI 2.5–3.7 g/dL) had developed. Enteral nutrition was started in addition to Vitamin B Complex^bb supplementation. Empirical treatment with azithromycin^cc (10 mg/kg IV q 24 hr) and atovaquone^dd (13.3 mg/kg PO q 8 hr) was started to cover a Babesia sp infection because the vector-borne pathogen panel results were still not anticipated for multiple days and Babesia infections are relatively common in Anguilla. Saline agglutination was once again positive, and a third TPE treatment was pursued with a target of 1.5 plasma volumes exchanged. The target plasma volume exchanged was reduced from 2 to 1.5 at this time given the patient’s improved stability and increased cost and treatment duration associated with a higher exchange rate. The replacement volume consisted of 54% FFP, 26% FP, and 20% Plasmalyte. A greater proportion of plasma was administered for colloidal support given the development of hypoproteinemia. The pretreatment PCV was 16%, and a fifth pRBC transfusion was started alongside a blood prime (12 mL/kg pRBC total). An unfractionated heparin CRI was restarted at 10 U/kg/hr. A high-pressure alarm was noted early into treatment, and blood clots were identified in the return chamber of the extracorporeal circuit. The heparin CRI and ACDA infusion were increased to 20 U/kg/hr and 1.2 mL/min/L TBV, respectively. A single bolus of calcium gluconate (50 mg/kg IV once) was administered during the procedure for ionized hypocalcemia. The final reported plasma volumes exchanged was 1.3. The posttreatment PCV was 14%, and a sixth pRBC transfusion (12 mL/kg) was administered. Posttreatment saline agglutination was negative.

The dog’s neurological status remained static the next morning (day 5). The posttransfusion PCV was 20%, and saline agglutination was once again positive. The TBILI increased again to 44.5 mg/dL. Intermittent alertness was noted, and clopidogrel (3 mg/kg PO q 24 hr) was started. The PCV decreased to 15%, and a seventh pRBC transfusion was administered. Given the previous report of three consecutive daily TPE treatments for IMHA and ABE⁶ and a previously reported protocol describing every other day treatments for immune-mediated hematologic disorders following the first two treatments,² no TPE treatment was performed on day 5.

The dog’s mentation was slightly improved the next day, although saline agglutination remained positive (day 6). Her PCV was 22% with a TBILI of 39.2 mg/dL. Additional IVIG was discussed, though it was elected to proceed with a fourth TPE treatment with a target of 1.5 plasma volumes exchanged. A blood prime was performed (eighth pRBC transfusion), and the replacement volume consisted of 25% FFP, 25% FP, and 50% Plasmalyte. Clotting was again noted in the return chamber with an ACDA infusion rate of 1.2 mL/min/L TBV and a blood inlet to anticoagulant ratio of 8:1. A 25 U/kg heparin bolus was administered, and the heparin CRI was gradually increased to 70 U/kg/hr. Mild hypocalcemia occurred (1.01 mmol/L, RI 1.21–1.51 mmol/L), and the calcium gluconate CRI was adjusted accordingly. The posttreatment PCV was 28%, TBILI was 16.9 mg/dL, and saline agglutination was negative. The dog began making attempts to stand immediately following treatment with an improved mentation.

The next morning (day 7), the dog was alert and attempting to ambulate. Her PCV was static at 28%, TBILI was 16.9 mg/dL, and saline agglutination remained negative. Transition to oral medications and physical rehabilitation was initiated. She was started on rivaroxaban (0.5 mg/kg PO q 12 hr) in addition to clopidogrel (3 mg/kg PO q 24 hr). The dog remained hospitalized for 5 additional days and was discharged from the hospital on day 12 with a PCV of 32% and a TBII of 2.7 mg/dL. The vector-borne pathogen panel was negative. The dog was discharged with prednisone^ee (2 mg/kg PO q 24 hr), mycophenolate^ff (11 mg/kg PO q 12), levetiracetam extended release^gg (43 mg/kg PO q 12 hr), rivaroxaban (0.9 mg/kg PO q 12 hr), clopidogrel (3 mg/kg PO q 24 hr), azithromycin (11 mg/kg PO q 24), atovaquone (14 mg/kg PO q 8 hr), and doxycycline (6 mg/kg PO q 12 hr). Recheck examination 5 days later showed a static PCV of 32%, ongoing regenerative response (absolute reticulocytes 111.2 × 10³/μL, RI 11–92 × 10³/μL), mild spherocytes, thrombocytosis (686 × 10³/μL, RI 186–545 × 10³/μL), and a TBILI of 1.1 mg/dL (Figure 2). Eleven months later at the time of writing, the dog is neurologically normal and continues to do well at home on monotherapy with mycophenolate (10 mg/kg PO q 24 hr).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

This report describes the successful use of four centrifugal TPE treatments in a critically ill dog with severe ABE secondary to IMHA. Despite failing to demonstrate a sufficient clinical response to three daily TPE treatments as previously reported,⁶ the ongoing support of plasmapheresis allowed for continued neurological improvement, cessation of autoagglutination, and successful discharge from the hospital with no permanent neurological damage. Prior publications on the use of TPE for severe IMHA have reported successful outcomes following two to three treatments; however, no dog with IMHA has been reported to survive if three treatments were deemed inadequate, with or without the presence of ABE^2,8,9. This report highlights the utility of providing ongoing TPE treatments in the face of severe persistent disease.

TPE can be achieved manually via the removal of aliquots of whole blood from a central line, centrifugation, and return of cellular components to the patient or via extracorporeal platforms.³ Membrane-based TPE relies on a filtration platform in which plasma proteins are filtered from cellular components via a nonselective membrane, whereas centrifuge-based TPE relies on centrifugal force to isolate blood components within an extracorporeal circuit based on their individual densities.³ All three of these modalities have been reported for the successful management of IMHA in dogs.^2,8–11 Plasmapheresis is unique in its ability to not only reduce concentrations of circulating autoantibodies against red blood cells but also decrease bilirubin levels.³ Individual reports of manual¹² and membrane-based TPE⁶ exist among the veterinary literature for the treatment of ABE in dogs with IMHA; however, only membrane-based TPE resulted in survival. Importantly, this latter report⁶ played a major role in both owner and clinician decision-making when opting for continued plasmapheresis in this case. Although a report of one dog receiving four centrifuge-based TPE treatments (approximately 6.5 times total plasma volume exchanged) exists in the literature, the dog did not respond despite its fourth treatment and was euthanized.⁸ Furthermore, although a protocol of two treatments on consecutive days followed by additional treatments every other day is described for immune-mediated hematologic disorders in dogs, no dog was reported to have required more than three treatments.² The case described in this report highlights the utility of continued TPE treatments if three treatments are inadequate because positive outcomes are still achievable. In the face of ongoing autoantibody generation and TBILI accumulation secondary to continued hemolysis, each TPE treatment only provides a temporary resolution. Additional time beyond 3 to 4 days of extracorporeal therapy may be required for immunosuppressive medications to take effect, and the number of treatments and target plasma volumes should be guided based on disease severity and response to medical management.

In dogs with severe or immediately life-threatening disease, the 2019 American College of Veterinary Internal Medicine consensus statement on the treatment of IMHA recommends the use of corticosteroids and a second immunosuppressive drug from the outset of treatment, continued supportive care for approximately 7 days, followed by IVIG if the disease remains uncontrolled.¹ If control has yet to be attained 7 days following IVIG, the addition of a third immunomodulatory drug or a splenectomy is recommended.¹ However, not all patients can survive this waiting period in the face of aggressive critical illness. Although the integration of plasmapheresis into these recommendations is limited by a paucity of clinical studies, the use of TPE is discussed as an additional treatment option in severely affected cases that can lead to a rapid reduction in autoagglutination, reduce the need for transfusions, and may lower mortality (Supplementary Appendix 4).¹ Most recently, a 2022 American College of Veterinary Internal Medicine abstract reported a significantly greater survival for dogs receiving at least three plasmapheresis treatments (92.9%) compared to those that received traditional medical management (61.8%) for IMHA.¹³

The application of TPE in the treatment of various immune-mediated conditions is widespread in humans.¹⁴ The 2023 Guidelines on the Use of Therapeutic Apheresis in Clinical Practice classify severe autoimmune hemolytic anemia as a Category III condition in humans, in which the optimum role of apheresis therapy is not established and individualized decision-making is recommended (weak recommendation, low quality or very low-quality evidence).¹⁴ However, TPE is considered a first-line therapy for numerous immune-mediated conditions diseases such as acute polyradiculoneuropathy, anti-glomerular basement membrane diseases (e.g., diffuse alveolar hemorrhage), demyelinating polyneuropathies, myasthenia gravis, and antibody-mediated transplantation rejection.¹⁴ Furthermore, plasmapheresis has been reported to be a safe and effective treatment option in humans with life-threatening hyperbilirubinemia.¹⁵ Its application has continued to evolve alongside medical advancements, as exemplified by the successful management of severe hyperbilirubinemia in a premature newborn supported by veno-arterial extracorporeal membrane oxygenation.¹⁶

There are several limitations of this case report. Given the critical nature of the patient and concern for cardiopulmonary arrest, no repeat TBILI concentration was obtained before commencing its first TPE treatment. However, it is strongly suspected that its concentration had continued to increase beyond the 48.1 mg/dL reported 7 hr before initiating the first TPE treatment. Additionally, other causes of severe neurological signs cannot be definitively ruled out given the lack of advanced imaging such as brain MRI. However, this diagnostic was pursued in the prior report and did not reveal any structural abnormalities.⁶ Despite no difference in change in hematocrit following membrane-based TPE compared to centrifuge-based TPE in humans with nonhemolytic indications,¹⁷ it has been anecdotally proposed that centrifugal TPE may result in increased hemolysis in dogs with IMHA given increased red cell fragility.¹⁸ It is possible that this may have contributed to the dog’s persistent dependence on pRBC transfusions. Additionally, care was taken to avoid administration of highly protein-bound drugs with a low volume of distribution (e.g., pantoprazole) before TPE treatments given their anticipated removal from circulation.¹⁹ Finally, IgG and IgM concentrations were not measured for kinetic modeling.

Conclusion

Plasmapheresis should be offered as a treatment option for severe cases of IMHA, particularly in the face of ABE. Despite previously reported protocols requiring three TPE treatments, an inadequate response to three treatments does not rule out a positive outcome. Continued treatments may still allow for complete resolution and control of both immune-mediated disease and ABE without residual neurological damage. Increasing awareness among the veterinary community of these successful outcomes associated with plasmapheresis has the potential to save numerous lives where continued care may otherwise appear futile.

The authors thank Julie Randall, PharmD, DICVP, for her contribution to pharmaceutical data collection.