Lack of Evidence for Perinatal Transmission of Canine Granulocytic Anaplasmosis From a Bitch to Her Offspring

Introduction

Recently, infection with Anaplasma (A.) phagocytophilum, a tick-borne rickettsial agent, has been documented in humans and dogs in the United States and other parts of the world.^1–6 The resulting clinical syndromes were first referred to as human and canine granulocytic ehrlichiosis. Initially, the cause was attributed to uncharacterized agents closely related to organisms then known as Ehrlichia (E.) phagocytophila and E. equi, which caused granulocytic ehrlichiosis in ruminants and horses, respectively.^4,7,8 Based on molecular similarities, E. phagocytophila, E. equi, and the agents of human granulocytic ehrlichiosis and canine granulocytic ehrlichiosis are now considered strain variants of a single species, A. phagocytophilum.⁸ Ixodes sp. ticks serve as the vector for A. phagocytophilum, and several mammalian species may act as reservoir hosts.¹ In the United States, disease corresponds to the distribution of Ixodes ticks, occurring mainly in the Northeast, Upper Midwest, and the Pacific Coast.¹

The clinical, hematological, and pathological manifestations of granulocytic anaplasmosis in dogs and humans are similar and recently have been reviewed.^1,2 The clinical syndrome in humans is now referred to as human granulocytic anaplasmosis (HGA), and in dogs it is called canine granulocytic anaplasmosis (CGA). Doxycycline is the antimicrobial of choice for treating HGA in adults and children and CGA in dogs.^1,9 Importantly, dogs may serve as sentinels for this disease in humans.

Human granulocytic anaplasmosis may have more severe clinical manifestations in immunosuppressive states such as pregnancy.¹⁰ Relatively few cases have been described in pregnant women, and perinatal transmission has been reported only once.^10,11 However, transplacental transmission has been documented in cows, and the organism has been detected in leukocytes in the milk of experimentally infected cattle. ^12,13 Infection in pregnant dogs has not been reported. Diagnosis of infection during pregnancy poses therapeutic dilemmas, because doxycycline is teratogenic. When infection is diagnosed after parturition, knowledge of the risk of perinatal transmission to offspring is important, because prophylactic therapy in neonates is also not without risk. Monitoring newborns for the presence of HGA is currently recommended when the disease is suspected in pregnant women.¹⁰ Herein we describe relatively severe clinical manifestations of A. phagocytophilum in a pregnant bitch and failure of perinatal transmission to her puppies.

Case Report

A 9-year-old, intact female Samoyed was referred to the University of Wisconsin-Madison, Veterinary Medical Teaching Hospital (VMTH) for evaluation of fever of unknown origin. Six days prior to presentation at the VMTH, the dog gave birth to five healthy puppies. Two dead puppies were removed by cesarean section. The day of parturition is referred to as day 0 for this report. On day 5, the dog became anorexic, lethargic, and febrile (rectal temperature 106.5°F). Laboratory findings at that time consisted of moderate thrombocytopenia (76,000 platelets/μL), moderate anemia (packed cell volume [PCV] 26%), mild neutrophilic leukocytosis, and marked lymphopenia (17,600 white blood cells [WBCs]/μL; 17,248 neutrophils/μL; and 352 lymphocytes/μL). The dog had received amoxicillin (500 mg orally every 12 hours) for 5 days following surgery. Ceftiofur (50 mg intramuscularly), flunixin meglumine (25 mg intramuscularly), and aspirin (360 mg orally) also were given 12 hours prior to referral. The owner had reported an episode of small bowel diarrhea the morning of presentation. Six weeks prior to presentation, a Western Blot assay for antibodies directed against Borrelia (B.) burgdorferi was consistent with vaccination and natural exposure.

At the time of presentation (day 6), rectal temperature was 105.2°F. Bilaterally increased bronchovesicular sounds and a soft, productive cough were noted on thoracic auscultation. A grade 2/6 left apical systolic murmur was initially auscultated but could not be heard subsequently. An engorged Ixodes sp. tick was removed from the left side of the dog’s neck. The puppies appeared healthy and occasionally nursed during the examination.

A complete blood count (CBC) confirmed severe thrombocytopenia (37,000 platelets/μL, reference interval 200,000 to 500,000/μL); moderate normocytic, normochromic anemia (PCV 28%, reference interval 37% to 55%) with four nucleated red blood cells (RBCs) per 100 WBCs; moderate poikilocytosis; and slight schistocytosis. A neutrophilic leukocytosis, a marked left shift, and neutrophil toxicity (18,100 WBCs/μL, reference interval 5400 to 15,300/μL; 12,851 mature neutrophils/μL, reference interval 2750 to 12,850/μL; 4163 band neutrophils/μL, reference interval 0 to 150/μL; and 543 metamyelocytes/μL) were present. The dog was markedly lymphopenic (181 lymphocytes/μL, reference interval 1000 to 4800/μL). Many of the circulating neutrophils contained one to six morulae identified as A. phagocytophilum.

Serum biochemical abnormalities included mild hypoalbuminemia (2.2 gm/dL, reference interval 2.5 to 4.0 gm/dL) and moderately increased alkaline phosphatase (1798 U/L, reference interval ≤ 166 U/L). Total carbon dioxide was decreased (12 mmol/L, reference interval 17 to 26 mmol/L). Hypercholesterolemia was mild (363 mg/dL, reference interval 111 to 290 mg/dL). Urinalysis revealed proteinuria (3+) and glucosuria (1+) with a specific gravity of 1.030. Moderate hematuria with 25 to 35 RBCs per high-power field were noted. Coagulation parameters included a greatly prolonged buccal mucosal bleeding time of 10 minutes (normal <5 minutes); prothrombin and partial thromboplastin times within the reference intervals; hyperfibrinogenemia (591 mg/dL); elevated fibrin degradation products (FDPs) (10 to 40 μg/mL); and decreased antithrombin concentration (66%, reference interval 80% to 120%).

Thoracic radiographs demonstrated diffuse interstitial lung disease, with a focal area of alveolar infiltrate in the ventral right middle lung lobe and sternal lymphadenopathy. Abdominal radiographs showed a small amount of free abdominal gas and a poorly mineralized fetal skeleton in the caudoventral abdomen cranial to the urinary bladder. Abdominal ultrasound showed a small fetus with minimal mineralization in the left horn of the uterus and echogenic luminal contents.

Anaplasma phagocytophilum infection, a retained and mummified fetus, pyometra, and early disseminated intravascular coagulation (DIC) were suspected. Because of the possible pyometra, an exploratory laparotomy and ovariohysterectomy were performed. The uterine tissues were submitted for aerobic and anaerobic culture and histopathology. A tracheal wash, performed at induction, revealed suppurative inflammation containing morulaeladen neutrophils. Upon recovery, cefazolin (20 mg/kg) followed by doxycycline (5 mg/kg) were given by slow intravenous infusion. After surgery, the dog’s PCV was 21%. The dog was crossmatched and transfused with fresh whole blood.

Cultures of the uterine tissues, urine, and tracheal wash fluid did not result in bacterial growth. Histopathological examination of uterine tissues was consistent with normal postpartum inflammation. Evidence of fetal resorption and a mummified fetus were seen.

The dog recovered well from surgery, but the following day the serum albumin decreased to 1.8 gm/dL. Repeat urinalysis revealed a urine specific gravity of 1.030, persistent proteinuria (2+) and glucosuria (trace), and resolution of the hematuria. The urine protein:creatinine ratio was 3.6. The dog received doxycycline (5 mg/kg orally every 12 hours for 21 days). Amoxicillin-clavulanic acid was added (15 mg/kg orally every 12 hours) to provide more broad-spectrum antimicrobial coverage until after the uterine, tracheal wash, and urine cultures were found to be negative. Enalapril (0.25 mg/kg orally once daily) was administered until the urinary protein loss resolved on day 18. Because of the possibility of perinatal transmission of A. phagocytophilum, the puppies were weaned at the time of the dam’s diagnosis, and they were subsequently monitored for evidence of infection. All puppies were normal on physical examination. The dam and the puppies were discharged 4 days after presentation.

The owner observed all dogs daily for appetite, activity level, attitude, and growth (in puppies). Rectal temperatures were monitored for the initial 2 weeks. The owner routinely inspected all dogs for ticks. The dam was also treated with a once-monthly topical acaricide to control additional tick infestations, and the puppies were treated after 2 to 3 months of age.

Ethylenediaminetetraacetic acid (EDTA) whole blood and serum samples were collected for CBC, buffy coat smear analysis, polymerase chain reaction (PCR), and A. phagocytophilum serological analysis as indicated in Tables 1 and 2. After day 8, the samples were collected at the owner’s local veterinary clinic and shipped overnight on ice. The serum and remaining EDTA specimens were then frozen at − 70°C within 4 hours of arrival.

Complete blood counts were performed on the dam, but because of the puppies’ small size and thus the limitations for the amount of blood withdrawn, complete hematological analyses were not done. For the puppies, the PCV and total plasma protein were measured, and two peripheral blood smears were examined. Additionally, for the dam and the puppies, peripheral blood smears and two to four buffy coat smears were examined at 200 × for Anaplasma sp. morulae within granulocytes.

The initial EDTA specimen (day 6) from the dam was assayed for 16S recombinant deoxyribonucleic acid (rDNA) of A. phagocytophilum, E. canis, E. ewingii, and E. chaffeensis as described previously.¹⁴ Based on the dam’s PCR results, the initial specimens (day 8) from the puppies were assayed only for A. phagocytophilum DNA. Antibody titers to A. phagocytophilum, E. canis, and E. chaffeensis were measured in all serum samples from the dam and the puppies as described.¹⁴ An immunofluorescence assay (IFA) performed on the dam’s serum (day 6) to detect antibodies to Rickettsia (R.) rickettsii was negative.^a

The clinical abnormalities found in the dam rapidly resolved. Anaplasma sp. morulae were no longer visible on peripheral blood smears 1 day after doxycycline therapy began, and they were not seen on any subsequent buffy coat smears. Thrombocytopenia resolved within 4 days. The WBC count normalized by day 19. The FDP levels were normal, and plasma antithrombin concentration had increased to 75% at the time of discharge. The hematocrit, thoracic radiographs, and urine protein:creatinine ratios were normal by day 18, as were all serum biochemical parameters with the exception of alkaline phosphatase, which remained elevated at 800 U/L. Serum biochemical analysis was repeated 7 months after presentation, at which point all parameters were normal.

As expected, PCR confirmed the presence of A. phagocytophilum DNA in the initial EDTA sample (day 6). Additional PCR assays for E. canis, E. chaffeensis, and E. ewingii DNA were negative. Anaplasma phagocytophilum 16S rDNA was not detected in subsequent EDTA samples [Table 1]. The dam had a moderate antibody titer to A. phagocytophilum (1:320) on day 6. The titer peaked on day 20 (1:2560) and thereafter gradually declined to 1:320 by day 105. Additionally, the dam had low to moderate serological cross-reactivity to E. canis and E. chaffeensis antigens during the first 2 months of the study [Table 1].

The puppies appeared healthy throughout the initial 4- month monitoring period. The owner consistently reported that all puppies had good appetites, high levels of energy, and were gaining weight at an appropriate rate. Platelet, erythrocyte, and leukocyte numbers appeared adequate on all peripheral blood smears. No morulae were seen at any time on any peripheral blood or buffy coat smears. Similarly, PCR on day 8 and all serological tests during this 4-month period were negative [Table 1].

During the fall, the owner noticed and removed a number of ticks from her dogs (the dam and puppy nos. 4 and 5). In December, 7 months after initial presentation, puppy no. 5 died suddenly. The puppy was sent to the VMTH for a necropsy. Histopathological abnormalities were seen only in the lung and thymus, and they consisted of acute pulmonary congestion with intraalveolar hemorrhage and secondary histiocytic infiltrate and thymic hemorrhage. Examination of other organs was unremarkable with the exception of autolytic change.

Shortly after puppy no. 5 died, titers were drawn from puppy no. 4, which shared the same environment. The A. phagocytophilum titer for puppy no. 4 was positive. The titer decreased from 1:1024 on day 235 to 1:320 on day 252. This puppy was treated prophylactically with doxycycline. Additional paired EDTA whole blood and serum samples were collected from the dam and puppy nos. 1 and 3 [Table 2]. No additional samples were available for puppy no. 2, which was lost to follow-up.

Eight months after initial presentation, the dam’s A. phagocytophilum titer decreased to 1:160 (day 252), and no Ehrlichia or Anaplasma sp. DNA was detected by PCR [Table 2]. Amplification of PCR from EDTA whole blood samples from puppy nos. 1, 3, and 4 did not detect Ehrlichia or Anaplasma sp. DNA. Antibody titers to A. phagocytophilum, E. canis, and E. chaffeensis remained at ≤ 1:20 for puppy nos. 1 and 3.

Discussion

In this report, we found no evidence of perinatal transmission of A. phagocytophilum from the bitch to her live offspring. Lack of evidence of perinatal transmission has also been documented in pregnant women.¹⁰ In case reports involving transplacental infection in a human baby and a calf, both neonates became seroreactive, and organisms or their DNA could be demonstrated in peripheral blood samples within several days of parturition.^11,13 Studies have shown that the majority of naturally and experimentally infected dogs became seroreactive.^4,15,16 Taken together, the lack of clinical illness, lack of PCR detection of the organism, failure to detect morulae, and lack of seroconversion after exposure suggest that transplacental and transmammary transmission did not occur in the surviving puppies. This is an important finding, because prophylactic therapy in neonates is not without risk. Doxycycline is the treatment of choice for HGA and CGA.^1,9 A theoretical risk of tooth discoloration exists following doxycycline use in neonates, although this side effect has not been reported to our knowledge. Other side effects, including esophagitis and gastrointestinal upset, are also potential complications. ¹⁷ Chloramphenicol has been suggested as an alternative treatment for A. phagocytophilum infection; however, in vitro data suggest it may not be as effective as doxycycline. ¹ A risk of bone marrow suppression is also associated with its use.

Due to the nature of the study, the precise risk of perinatal transmission of A. phagocytophilum cannot be determined. However, the lack of evidence for infection in any puppy during the perinatal period suggests that monitoring neonates for infection rather than treating prophylactically is a reasonable option. Serological evidence in this case suggests the bitch was infected shortly before parturition. However, as an engorged Ixodes tick was found at presentation, the dog possibly was infected after parturition due to continued exposure.

Anaplasma phagocytophilum morulae can be readily demonstrated in circulating neutrophils in experimentally infected dogs transiently, approximately 4 to 14 days after infection.^16,18 The DNA can be detected even earlier after inoculation.¹⁹ However, experimentally infected dogs initially become seropositive (defined as a titer >1:40) around day 9 postinfection, and seroconversion occurs shortly thereafter.¹⁹ The fact that the dog in this report was seropositive at the time of presentation and subsequently seroconverted suggests that infection occurred shortly before parturition. Still, the timing of the infection offers circumstantial evidence that A. phagocytophilum could have contributed to the death of the fetuses; thus, transplacental transmission cannot be completely ruled out in this case.

The death of the puppy with acute respiratory distress syndrome (ARDS) at 7 months of age was not likely due to A. phagocytophilum infection, because the puppy was active and playful up until the moment of death and showed no premonitory signs of illness. However, ARDS and a macrophage activation syndrome have been described in humans infected with the organism;²⁰ thus, A. phagocytophilum infection cannot be ruled out as a cause of death. Chronic infection occurs in sheep and has been reported in one study in which dogs were experimentally infected.^19,21 Chronic infection with A. phagocytophilum in naturally infected dogs has never been documented. Importantly, both the puppy that died of ARDS and puppy no. 4 (a healthy puppy in which a titer of 1:1024 was detected 7 months after parturition) continued to reside in the northern Wisconsin habitat where the original exposure occurred, and they continued to suffer from tick infestation. The complete lack of evidence of acute infection in the weeks following parturition and the continued exposure to A. phagocytophilum infected ticks suggest that a later exposure may explain the clinical findings in these puppies.

The clinical findings of fever, anemia, neutrophilia, lymphopenia, thrombocytopenia, elevated alkaline phosphatase, hypoalbuminemia, and transient proteinuria in the dam are consistent with other reports of A. phagocytophilum infection in dogs.^2,4,15,19 Interestingly, glucosuria has been described in only one dog with CGA.¹⁵ Cephalosporin administration or tubular damage from transient renal ischemia associated with nonsteroidal antiinflammatory medication (i.e., flunixin meglumine) could also explain the transient glucosuria without hyperglycemia in this case.

Although coughing has been described in dogs during both natural and experimental infections with CGA, this is the first report to actually document pneumonia.^6,16 Interestingly, atypical pneumonitis has been described in human patients.²² The presence of interstitial and alveolar pulmonary disease, the presence of morulae within neutrophils obtained from this dog’s airways, and the lack of growth of other organisms from the fluid obtained by tracheal wash strongly suggest that the pulmonary disease in this dog was attributable to A. phagocytophilum infection.

Disseminated intravascular coagulation was suspected based on the combination of thrombocytopenia, schistocytosis, elevated FDP levels, and decreased antithrombin. Although a rare complication, DIC has been documented in a dog infected with A. phagocytophilum.²³ Therefore, infection with A. phagocytophilum may have contributed to the development of DIC in this case. Immunosuppression is believed to be a risk factor for the development of more severe disease syndromes in humans infected with A. phagocytophilum.¹⁰ Immunosuppression associated with pregnancy may have contributed to the severity of disease in this dog.

Coinfection with tickborne agents can result in increased severity of clinical signs. Coinfection with R. rickettsii was unlikely in this dog based on geographic locale and infecting tick species. However, because PCR testing and convalescent titers were not obtained, coinfection with R. rickettsii cannot be completely ruled out. Acute infection with E. canis was considered unlikely based on the negative PCR at presentation, the lack of seroconversion, and because Ixodes scapularis is not a known vector for E. canis. However, chronic infection with E. canis cannot be completely ruled out, as all samples were not tested using PCR.

Interestingly, evidence of exposure to B. burgdorferi in this dog suggests that coinfection with B. burgdorferi could have contributed to the severity of clinical signs in this case. Coinfection with A. phagocytophilum and B. burgdorferi has been described in humans and dogs, as these organisms share the same vector.^1,24 Furthermore, a recent report provides evidence that coinfection with B. burgdorferi results in more severe clinical disease in dogs infected with A. phagocytophilum.²⁴ However, active infection with B. burgdorferi cannot be differentiated from previous exposure using serology. Because culture and PCR for B. burgdorferi are insensitive and nonspecific, they were not performed.²⁵ Therefore, the hypothesis that coinfection with B. burgdorferi contributed to the severity of clinical signs in this case cannot be definitively proven.

Conclusion

Anaplasmosis caused by A. phagocytophilum is an emerging tick-borne disease in dogs and humans. Determining the risk of perinatal transmission from infected individuals to offspring has important therapeutic implications. In the case described here, we did not find evidence of perinatal transmission of A. phagocytophilum from a bitch to her puppies. Immunosuppression associated with pregnancy or coinfection with B. burgdorferi could have contributed to the severe clinical manifestations reported in this dog. Further studies will be necessary to definitively determine the risk of perinatal transmission of A. phagocytophilum in dogs and the impact of pregnancy on the clinical signs of disease.