Acute Granulocytic Ehrlichiosis in a Rottweiler

Introduction

Natural infection with granulocytic ehrlichiosis has been reported in dogs, horses, cows, humans, and, less commonly, cats.^1–3 The blacklegged tick (Ixodes scapularis) transmits the parasite in the northeast and upper midwestern United States, whereas the western black-legged tick (Ixodes pacificus) is the vector in northern California.⁴ The agents responsible for granulocytic ehrlichiosis have been reclassified into the A. phagocytophilum genus group (formerly the human granulocytic ehrlichiosis agent, E. phagocytophilia, and E. equi). They are obligate intracellular bacteria that primarily infect granulocytes. These pathogens are very similar or likely identical to each other, exhibiting serological cross-reactivity and cross-protective immunity.^2,5–7 In addition, isolates obtained from humans will infect dogs, cats, sheep, and horses.¹ These findings suggest that granulocytic ehrlichiosis may be an important, emerging zoonotic disease. Infections in people have now been reported from over 25 states in the United States (incidence of one per 10 to 100 million population), including a rare case in Illinois.⁴

Experimental studies of canine granulocytic ehrlichiosis have shown an incubation period of approximately 4 to 11 days.⁸ In both naturally and experimentally infected dogs, early clinical signs are often nonspecific and limited to fever and depression.^3,8,9 The most consistent laboratory findings in naturally infected dogs include thrombocytopenia, lymphopenia, increased serum alkaline phosphatase and amylase activities, and proteinuria.^3,9

Laboratory confirmation of ehrlichiosis requires the use of microscopic, serologic, molecular, or culture-based methods. Definitive diagnosis requires culture isolation, although this is the most difficult and time-consuming approach.⁵ In acutely infected dogs, Ehrlichia inclusions may be detected in granulocytes on a peripheral blood smear. However, parasitemia is often low or transient and is easily overlooked. In experimentally infected dogs, antibody titers may be detected as early as 10 days after inoculation and within a few days after observing Ehrlichia inclusions.⁸ A fourfold rise or fall in the antibody titer with a peak titer >1:40 is considered diagnostic in dogs with compatible clinical signs.⁸ Molecular methods have been applied to the diagnosis of granulocytic ehrlichiosis, on the basis of deoxyribonucleic acid (DNA) sequencing of the 16S ribosomal ribonucleic acid (rRNA) gene.⁵ Detection of DNA from granulocytic Ehrlichia species for up to 5.5 months after experimental inoculation of dogs suggests persistence of infection.¹⁰

The purpose of this paper is to describe a dog that was presented for possible pancreatitis and was concurrently diagnosed with acute granulocytic ehrlichiosis.

Case Report

A 6-year-old, 58-kg, spayed female rottweiler was referred to the University of Illinois Veterinary Teaching Hospital for evaluation of possible pancreatitis. The dog was presented to the referring veterinarian for lethargy and lameness 30 days prior to referral. Physical and neurological examinations at that time were unremarkable. The referring veterinarian vaccinated the dog for Lyme disease. No other therapy was instituted. On the day of referral, the referring veterinarian examined the dog for a 2-day history of depression, anorexia, weakness, soft stool, and vomiting. Rectal temperature was 105.4°F. Laboratory abnormalities included a mild, nor-mocytic, normochromic anemia (35.1%; reference range 37% to 55%); mild thrombocytopenia (155 × 10⁹/L; reference range 175 to 500 × 10⁹/L); increased alkaline phosphatase (1568 U/L; reference range 23 to 212 U/L); and increased amylase (>2500 U/L; reference range 500 to 1500 U/L). The dog received a single dose of amoxicillin (10 mg/kg per os [PO]) and was referred to the teaching hospital’s Acute Illness Clinic at the University of Illinois.

At presentation, the dog was listless with a rectal temperature of 103°F. Auscultation of the lungs was suboptimal, owing to obesity and heavy panting. The remainder of the physical examination revealed only mild popliteal lymphadenopathy. An engorged female Ixodes scapularis tick was identified in the dog’s cage the day following presentation. Complete blood count (CBC) at admission revealed a normal total white cell count (WBC) of 12,800/μL (reference range 6000 to 17,000/μL); however, the presence of Dohle bodies and cytoplasmic basophilia in the neutrophils suggested an inflammatory response. The WBC differential included 85% neutrophils, 1% bands, 5% lymphocytes (mildly decreased), 8% monocytes, and 1% eosinophils. A peripheral blood smear revealed that approximately 10% of the neutrophils along the feathered edge contained intracytoplasmic inclusions consistent with Ehrlichia spp. morulae [see Figure]. A mild, normocytic, normochromic anemia (34.5%; reference range 35% to 52%) and severe thrombocytopenia (20.7 × 10⁹/L; reference range 200 to 900 × 10⁹/L) were present. Abnormalities identified on the chemistry panel included elevations in serum alkaline phosphatase (1087 U/L; reference range 12 to 110 U/L); corticosteroid-induced alkaline phosphatase (452 U/L; reference range 0 to 40 U/L); amylase (5789 U/L; reference range 190 to 1357 U/L); and lipase (4389 U/L; reference range 25 to 534 U/L). The marked increase in the pancreatic enzymes, in the absence of renal disease, was consistent with pancreatitis. A coagulation panel (including prothrombin time, partial thromboplastin time, and fibrin degradation products) was normal, which ruled out fulminate disseminated intravascular coagulation.

Fine-needle aspiration cytology of the prominent popliteal lymph node revealed a reactive node, characterized by a mild plasma cell hyperplasia. Thoracic radiographs were within normal limits. An abdominal ultrasound indicated a diffusely hypoechoic liver and a heterogenous spleen with one hypoechoic nodule. Although no ultrasonographic abnormalities were noted in the region of the pancreas, the dog consistently exhibited pain in the right cranial quadrant when pressure was applied with the ultrasound probe. Fine-needle aspirates of the liver and spleen revealed extramedullary hematopoiesis. Additionally, rare neutrophils and immature granulocytes from both sites contained intracytoplasmic inclusions of Ehrlichia organisms.

Ophthalmic examination revealed miosis and aqueous flare bilaterally and decreased intraocular pressure (7 mm Hg; normal 15 to 25 mm Hg) in the left eye, consistent with anterior uveitis. No active retinal lesions were noted. Topical dexamethasone 0.1% was initiated in both eyes q 8 hours for the anterior uveitis.

Upon completion of the initial workup, the working diagnoses were acute granulocytic ehrlichiosis and pancreatitis. It was not known if these two conditions were related or if they simply occurred coincidentally. During hospitalization, the dog received sodium chloride (NaCl; 0.9% intravenously [IV] at a daily maintenance rate) and ampicillin (1 gram IV q 8 hours). The dog received nothing PO for the first 24 hours. Fresh-frozen plasma was given as a source of alpha macroglobulin and preventatively for disseminated intravascular coagulation. A 3-week course of doxycycline (5 mg/kg PO q 12 hours) was initiated for the apparent Ehrlichia infection.

Forty-eight hours after therapy was started, the platelet count increased to 53.6 × 10⁹/L. The total alkaline phosphatase activity decreased to 948 U/L; however, the corticosteroid-induced alkaline phosphatase remained unchanged. Amylase and lipase activities were markedly decreased from pretreatment values (1804 U/L and 699 U/L, respectively). Clinical improvement was evident by an increase in appetite, cessation of vomiting, and normal stool consistency. Because of the rapid response to therapy and an increasingly aggressive temperament, the dog was discharged 48 hours after admission.

Two days following discharge, the platelet count was normal (278 × 10⁹/L). A CBC performed 2.5 weeks later was also completely normal. The chemistry panel had a persistent mild elevation in alkaline phosphatase activity (531 U/L; reference range 23 to 212 U/L) and a normal amylase value (886 U/L; reference range 500 to 1500 U/L).

Pretreatment serum from the dog was frozen and sent to a reference laboratory^a for serological evaluation of tick-borne diseases. Serology was performed for Babesia canis, Ehrlichia (E.) canis, E. platys, E. equi (Anaplasma [A.] phagocytophilia), Lyme disease, and Rocky Mountain spotted fever. All titers were negative with the exception of Lyme disease, which was positive at ≥1:60. Because the dog had previously been vaccinated for Lyme disease, it could not be determined if this titer represented a vaccine-induced response or a natural infection. Western blot analysis for Lyme disease submitted to a reference laboratory^b was consistent with a vaccine-induced titer. A convalescent serum sample obtained from the dog 3 weeks posttreatment was sent to the original reference laboratory^a for serological evaluation. All titers were negative at this time, including the Lyme titer. Aliquots of the acute serum sample and a convalescent serum sample obtained 6 weeks posttreatment were sent to another research laboratory^c to confirm titer results. In contrast to previous results, both the acute and convalescent serum samples were found to contain immunoglobulin G (IgG) antibody that reacted with E. equi (A. phagocytophilia) at extremely high titers (>1:8,000).

Blood was also submitted to a research laboratory^d for molecular identification of the Ehrlichia organism. A set of universal primers was used in a polymerase chain reaction (PCR) to amplify the 16S rRNA gene of the organism. Gel-purified 16S rRNA products were sequenced in both the sense and antisense directions by a dideoxy terminator method. Analysis of the sequence data revealed a remarkable similarity to the A. phagocytophilia genus group (including E. phagocytophilia, E. equi, and human granulocytic ehrlichiosis). The sequence was 100% homologous to previously reported isolates of granulocytic ehrlichiosis identified in Swedish dogs and horses, as well as the causative agent of human granulocytic ehrlichiosis identified by Chen in Galveston, Texas.¹¹ Using primers specific for the A. phagocytophilia genus group, a second research laboratory^c was also able to amplify the 16S rRNA gene from cell cultures inoculated with peripheral blood from the infected dog. However, attempts to isolate the Ehrlichia organism from the tissue culture failed.

The blood parasites in this dog were also visualized by transmission electron microscopy. Blood was processed for electron microscopy as previously described.⁵ Typical morulae of Ehrlichia were detected in the cytoplasm of the neutrophils. These organisms were pleomorphic and surrounded by a distinct double membrane.

Discussion

To the authors’ knowledge, this is the first documentation of granulocytic ehrlichiosis in a dog in Illinois. A presumptive diagnosis of granulocytic ehrlichiosis in this dog was determined after identification of characteristic inclusion bodies within the cytoplasm of neutrophils [see Figure]. Confirmation of the diagnosis was made by PCR and electron microscopy. Polymerase chain reaction is an extremely sensitive assay, which can be a useful diagnostic tool during the acute stage and may yield positive results prior to the animal mounting an antibody response and before parasites are identified within granulocytes.⁸ Sequencing of the amplified DNA is necessary for strain differentiation, as PCR cannot specifically distinguish between strains that differ by only a few DNA base pairs.^5,6,12 Although not a useful tool for routine diagnosis, the presence of morulae within neutrophils, as detected by electron microscopy, added further support for the diagnosis in this dog.

There are several unique and interesting aspects to this case. First, the case highlights the inherent problems that may be encountered when serological techniques are used as the primary means to diagnose an Ehrlichia infection. Serology is currently the most common diagnostic test used to detect Ehrlichia diseases. Interestingly, neither the acute nor convalescent sera initially submitted revealed any cross-reactivity to E. equi (A. phagocytophilia), despite the positive visual and PCR identification of the organism. Previously reported studies of PCR-confirmed cases in people infected with A. phagocytophilia and horses infected with E. equi (A. phagocytophilia) suggest that serology may not be sensitive in identifying acute infections.^12,13 Both in humans and horses, it has been postulated that with prompt detection and early treatment, IgG seroconversion may not occur until later in the disease course, if at all.^13,14 If the immune system can be circumvented in this manner, the possibility for reinfection following a subsequent exposure is an issue that requires further investigation. Although failure to mount an IgG response was considered as an explanation for the negative serological results in this dog, serology repeated by another laboratory revealed titers consistent with infection. Thus, the most likely explanation for the negative convalescent result originally reported is a pro-zone phenomenon. This is thought to occur when serum containing high concentrations of antibody thwarts the normal antigen-antibody complex formation, giving a false-negative test result. A positive titer was obtained on the same sample once the serum was sufficiently diluted to eliminate the antibody excess.

An additional diagnostic and clinical consideration in this case was the possibility of a concurrent infection with Borrelia burgdorferi, the agent of Lyme disease. The Western blot for Lyme disease confirmed the vaccinated status of the dog; however, a natural infection could not be definitively ruled out. Serum antibodies to both granulocytic ehrlichiosis and Borrelia burgdorferi have been isolated in dogs with clinical disease.^15–17 Both agents are transmitted by exposure to infective ticks, so simultaneous transmission and coinfection are possible.^15–17 Ticks that carry both A. phagocytophilia and Borrelia burgdorferi are reportedly capable of transmitting both agents with equal efficiency.¹⁸ The prevalence of coinfection is currently unknown.

To the authors’ knowledge, this is the first dog to be reported with concurrent granulocytic ehrlichiosis and pancreatitis. Although increased amylase and lipase activities (>4 and >8 times the normal range, respectively, in this dog) are not definitive for a diagnosis of pancreatitis, these elevations taken together with the clinical evidence of abdominal pain in the pancreatic region were supportive of pancreatitis. Elevations of amylase in animals with granulocytic ehrlichiosis have been previously reported; however, amylase is a nonspecific enzyme with pancreatic, renal, gastrointestinal, and hepatobiliary sources.^3,9,19 Increased lipase activity with concurrent Ehrlichia infection has not been previously reported. It is not known if the pancreatitis was a coincidental finding or if it was directly related to the Ehrlichia infection. Further studies are needed to determine if a causal relationship exists between these two diseases.

Conclusion

Infection with granulocytic ehrlichiosis should be suspected based on clinical signs and the possibility of exposure to infective ticks. Although serology is the most commonly used diagnostic test for Ehrlichia infection, it may be ineffective in identifying acutely infected cases. Characteristic Ehrlichia inclusion bodies within the cytoplasm of neutrophils may be visualized in acute cases; however, the parasitemia is often low or transient. During the acute stage of infection, PCR is useful and may yield positive results prior to detection of an antibody response and/or parasitemia on a peripheral blood smear. Although the dog in this report was diagnosed with concurrent pancreatitis, any definitive association between these two diseases remains unknown.

Acknowledgment

The authors thank Dr. J. Stephen Dumler from the Department of Pathology, Johns Hopkins Hospital, for attempting to cultivate this organism, for confirming the presence of a granulocytic Ehrlichia organism by PCR of cultured material, and for serological analysis.