Examination and treatment
On presentation, the dog was quiet and non-ambulatory. Physical exam indicated pallor with mild icterus and splenomegaly on abdominal palpation. The dog weighed 5.75 kilograms and was normothermic (38.2°C). He was tachycardic (150 beats/minute) and tachypneic (60 breaths/minute) with normal bronchovesicular and heart sounds on auscultation. Point-of-care bloodwork showed severe anemia (PCV 14%) and low refractometric serum total protein (TP, 5.4 g/dL). A venous blood gas/electrolyte panel (NOVA Stat Profile, Nova Biomedical, Waltham, Massachusetts, USA) was consistent with compensated metabolic acidosis with elevated lactate (5.2 mmol/L). Point-of-care ultrasound evaluation of the heart and lungs showed subjective volume contraction and no effusions or pulmonary B-lines were noted. Abdominal point-of-care ultrasound was unremarkable other than subjective splenomegaly. The dog was blood typed (DEA 1.1 positive) and major cross-matched, then admitted to the ICU. He was initially treated with 1 mg/kg IV maropitant (Cerenia, Zoetis, Kalamazoo, Michigan, USA), then was transfused with 14 mL/kg of pRBCs.
CBC and chemistry analysis were performed (see Tables 1 and 2), and cytologic evaluation of a blood smear by a boarded clinical pathologist showed intra-erythrocytic piroplasms morphologically consistent withBabesia canis . The dog was treated with 6.6 mg/kg IM imidocarb (Imizol, Merck, Kenilworth, New Jersey, USA) and was started on IV fluids (Lactated Ringer’s solution, Baxter, Deerfield, Illinois, USA) at 75 mL/kg/day. After completion of the blood transfusion, a PCV/TP showed no improvement of his anemia (PCV 13%, TP 5.0), and progressive hyperlactatemia (8.2 mmol/L). The dog was noted to be more alert, was ambulatory and eupneic (36 breaths/min) but febrile (39.4°C). He was given 2.2 mg/kg diphenhydramine IM (West-Ward, Eatontown, New Jersey, USA) due to concern for a transfusion reaction and an additional 13 mL/kg pRBC were administered over 6 hours. Fourteen hours after admission the dog became tachypneic (RR >100 breaths/min) and was transferred to an oxygen chamber to provide approximately 40% fractional inspired oxygen (FiO2). An hour later he developed diarrhea and hypersalivation, so he was given 0.02 mg/kg atropine IV due to concern for parasympathetic signs as an adverse effect of imidocarb. Over the next hour his tachypnea worsened, necessitating emergent endotracheal intubation using IV propofol (Abbot Animal Health, Chicago, Illinois, USA).
During intubation, a large volume of serosanguineous/icteric fluid was produced through the endotracheal tube. Initial manual ventilation using an anesthetic machine and 100% FiO2 yielded SpO2 values 88-92%. He was then switched to mechanical ventilation (MV) using a commercial ventilator (Puritan Bennett 840, Medtronix, Minneapolis, Minnesota, USA). Initial ventilator settings are listed in Table 3 (hour 0). Sedation was maintained with titrated continuous rate infusions (CRIs) of fentanyl (Hospira Inc., Lake Forest, Illinois, USA) at 5-10 mcg/kg/hr, midazolam (West-Ward) at 0.1-1 mg/kg/hr, and dexmedetomidine (Dexdomitor, Zoetis) at 1 mcg/kg/hr. An indwelling urinary catheter was placed, and urine output was adequate (2.3 mL/kg/hr over 12 hours).
Initially, the dog’s pulmonary static compliance was poor (0.7 mL/cm H2O/kg, normal range 1-1.6 mL/cm H2O/kg) and normoxemia measured by pulse oximetry (SpO2) was achieved only with aggressive ventilator settings (see Table 3) and high FiO2 (80%). Arterial catheterization/blood gas samples were not obtained due to the dog’s thrombocytopenia and coagulopathy noted on point-of-care coagulation parameters (prothrombin time 20 seconds, RI 12-17 seconds; activated partial thromboplastin time >300 seconds, RI 72-102 seconds). Blood titers were submitted to a reference lab to evaluate for co-infection with Borrelia , ehrlichiosis, anaplasmosis, and Rickettsia rickettsii . A 6 French, 90-cm nasogastric tube (Mila, Florence, Kentucky, USA) was placed and the dog was started on enteral doxycycline (10 mg/kg q 24 h; PuraCap, Laurelton, New York, USA). Repeat point-of-care lung ultrasound showed confluent B-lines bilaterally with no evidence of left atrial enlargement as determined by the LA:Ao ratio.
The dog’s PCV improved and remained stable throughout the day (23-24%), although he developed marked jaundice. He remained normothermic and normotensive but was dyssynchronous with the ventilator and was assessed to be in a relatively light plane of sedation. Additional sedation was added with a CRI of ketamine (Hospira Inc.) at 0.5 mg/kg/hr and propofol at 0.1-0.5 mg/kg/min titrated to effect. Adequate sedation/synchrony was achieved, and ketamine was discontinued within a few hours.
After 10 hours on MV, the dog developed progressive hypercapnia (ETCO2 55 mmHg). While troubleshooting, his endotracheal tube was sterilely suctioned and produced a large amount of icteric/serosanguineous fluid, which was cytologically consistent with a neutrophilic exudate and rare intracellular bacteria. Hand-held refractometry of the fluid showed a TP of 4.8 g/dL (peripheral TP was 5.6 g/dL). A sample of the fluid was submitted for aerobic culture and piperacillin/tazobactam (50 mg/kg IV q 6 h; Auromedics, East Windsor, New Jersey, USA) was started. The dog then developed hypotension (mean arterial pressure [MAP] 65 mmHg) which was initially responsive to a fluid challenge (10 mL/kg Lactated Ringer’s solution). At 12 hours on MV, hypotension recurred (MAP 44 mmHg) coupled with relative oliguria (urine output 1.5 mL/kg/hr). There was no improvement with another fluid challenge (5 mL/kg), so a norepinephrine (Claris Lifesciences, North Brunswick, New Jersey, USA) CRI was started at 0.2-0.6 mcg/kg/minute titrated to achieve a MAP of >70 mmHg and systolic blood pressure >100 mmHg.
At 13 hours on MV, the dog developed poor oxygen saturation (SpO2 89%, FiO2 60%) and frequent tachypnea despite adequate sedation (no response to physical/auditory stimuli). An arterial blood gas showed severe hypoxemia and hypercapnia (Table 3). Ventilator settings were adjusted using a modified version of the ARDSnet protocol (see Table 4) and were titrated to meet a minimum SpO2 >92%. He was also given 0.1 mg/kg dexamethasone (Dexaject SP, Henry Schein, Melville, New York, USA) IV once. The dog’s SpO2 mildly improved initially, but he continued to have significant tachypnea despite adjusting ventilator settings to address dyssynchrony. A propofol bolus IV was given to effect to induce complete apnea, after which his oxygen saturation substantially improved (SpO2 98%); the decision was made to induce neuromuscular blockade with atracurium (Sagent, Schaumberg, Illinois, USA) 0.3 mg/kg IV, followed by a CRI (6-9 mcg/kg/min) which rapidly improved his ventilator synchrony. Persistent hypercapnia occurred after neuromuscular blockade was started (ETCO2 60-65 mmHg) but maintained adequate oxygenation.
At 24 hours on MV, ventilator settings had been gradually weaned while maintaining SpO2 >93%. Portable thoracic radiographs (Figure 1) showed a bilateral mixed pulmonary interstitial and alveolar lung pattern which was worse in the left lung lobes (attributed to atelectasis per a board-certified veterinary radiologist). Diagnostic considerations included ARDS or transfusion-related acute lung injury (TRALI).
At 48 hours on MV, the dog was on minimal ventilator settings (Table 3). Atracurium was discontinued and norepinephrine was weaned/discontinued. Within hours, the other sedative infusions were also discontinued (fentanyl/midazolam/propofol), and his dexmedetomidine was lowered to 0.5 mcg/kg/hour. He was extubated and left to fully recover from sedation in an oxygen cage with 30% FiO2. Several hours later, oxygen supplementation was discontinued. Results of the pending infectious disease titers showed ehrlichiosis co-infection. His urinary catheter was removed, and he walked with mild support.