Case Report

Dermatitis was noted in a group of three captive Nile hippos at the Milwaukee County Zoo. The group consisted of a 28-yr-old male, a 27-yr-old female, and their 21-yr-old female offspring. The hippos were separated by gender, but the two groups utilized several of the same areas at different times of the day. The housing for these animals consisted of an indoor pool, two indoor concrete floored holding areas, and an outdoor yard which has a traffic bond substrate and a concrete pool. Both pools are cleaned by hosing daily after draining, with a subsequent refill using city tap water. The indoor holding areas are hosed daily and are cleaned with a quaternary ammonium disinfectant. The hippos are fed high fiber herbivore cubes and timothy/grass hay. No change in housing, feeding, or cleaning programs occurred prior to the development of dermatitis.

In February of 1997, the older female hippo developed numerous linear cracks in its skin, especially over the bridge of the nose and bilaterally along the shoulders and flanks. A 2-cm circular wound was present ventrally at the angle of the right mandible. Initial treatment included topical application of dilute chlorhexidine solution by sprayer to the entire body once daily which later changed to 0.1% gentamicin in glycerin. Over the next month, the animal was lethargic intermittently, would dog-sit, and appeared to be losing weight. Actual weights could not be obtained. The hippo was treated empirically with trimethoprim-sulfamethoxazole 45 g p.o., s.i.d. for 2 wk. Attempts to obtain skin samples via biopsy darts were unsuccessful. After oral treatment, the animal was less lethargic, but continued to lose weight.

Milder lesions, consisting of skin cracks over the face, limbs, and trunk, were present in the other two hippos. Over the next 6 wk, all hippos developed bilateral draining wounds under the mandible. A quick scalpel biopsy of the margin of a submandibular lesion in the older female was obtained through a partially opened stall door. The 1×2 cm flap of skin was submitted for histopathology, which revealed suppurative, multifocal epidermitis teeming with numerous dense colonies of gram-positive cocci, as well as severe hyperkeratosis and multifocal parakeratosis. Cultures of two different skin wounds were obtained in each of the two older animals. All four wounds grew a heavy growth of group G beta-hemolytic Streptococcus. Moderate numbers of Morganella morganii and Citrobacter freundii also grew in one of the two cultures from each animal. The isolated Streptococcus was susceptible to all drugs tested, including trimethoprim-sulfa combinations, but had only intermediate susceptibility to gentamicin.

In April of 1997, the male was treated with oral trimethoprim-sulfamethoxazole 60 g p.o., s.i.d. for 2 wk. The mild lesions in the male hippo and the younger female began to resolve. Weight loss became marked in the older female hippo, and serosanguinous fluid began to ooze from the skin lesions. The older female was retreated with a higher dose of trimethoprim-sulfamethoxazole, 75 g p.o., s.i.d. for 2 wk, without any signs of improvement. Weight loss continued, and a large draining tract formed on the lateral aspect of the right carpal region. Feces of all hippos were negative for parasites and enteric pathogens including Salmonella spp., Campylobacter spp., and Mycobacterium paratuberculosis. The hippos were switched to a Mazuri herbivore diet fortified with natural vitamin E, beta-carotene, essential fatty acids, and organic zinc.

Immobilization of the older female for blood work and biopsies was planned. A medical infectious disease specialist recommended treatment with an oral penicillin, as streptococcal infections can show in vivo resistance to trimethoprim-sulfa drugs even when in vitro susceptibility is reported (G. Dorff, personal communication). Due to the severity of the illness in this animal, treatment with amoxicillin 30 g p.o., s.i.d. for 6 wk began in mid-May, ten days prior to the immobilization and workup. Attempts at intramuscular treatment by pole syringe were completely unsuccessful due to the thickness of the hippo skin and its aggressive reaction to the attempted treatment.

The older female hippo was moved into an elephant chute for containment immediately prior to immobilization. At its smallest setting, the chute measured 2×4.5 m, which still allowed the hippo to move forward and backward, as well as turn around inside the chute. While the dimensions of the chute were too large to safely manipulate an awake hippo, the containment it afforded proved very helpful during the immobilization. Ultrapotent narcotics were not used due to concern regarding collapse and death during anesthesia in this debilitated animal. The hippo was estimated to weigh 950 kg and was immobilized with 40 mg of detomidine and 60 mg of butorphanol delivered by a Cap-chur dart into the muscles of the upper neck region. The animal became recumbent at 7 min, but maintained limited movement of its head and limbs throughout the procedure. Respirations varied between 1–4 breaths/min. Measurements of oxygen saturation could not be obtained. Supplemental oxygen was supplied intranasally. Heart rate was 36 beats/min. Multiple skin biopsies were taken from the flank region and submitted for histopathologic examination and culture. The hippo was skin tested for tuberculosis using intradermal injections of mammalian isolates and PPD-bovis in the thinner skin adjacent to the ears. Blood was drawn from the ventral tail vein using an 18-ga needle attached to an extension set. The animal stood with little warning at 65 min after the initial darting, prior to the administration of 250 mg of yohimbine and 500 mg of naltrexone, which were subsequently given intramuscularly.

The skin test was negative for reaction to tuberculin at 72 hr by manual palpation. Evaluation of the hippo’s blood work proved problematic, as normal values could not be obtained. The following values were possibly elevated: blood urea nitrogen 48 mg/dl, creatinine 1.8 mg/dl, globulin 4.0 g/dl, lactate dehydrogenase 1,529 U/L, aspartate transaminase 251 U/L, and alanine transaminase 68 U/L.

Cultures of the biopsy sites revealed numerous organisms, but no Streptococcus spp. Serratia liquefaciens, Proteus vulgaris, Klebsiella pneumoniae, Escherichia coli, and moderate diphtheroids cultured from various sites were considered to be contaminants. Blood cultures were also contaminated, growing numerous gram-negative organisms, including Klebsiella pneumoniae, Escherichia coli, Proteus vulgaris, and Enterococcus faecalis. Fungal cultures of the skin were negative for growth. No serum antibodies to Blastomyces sp. were detected by radial immunodiffusion.

Histopathology of the skin taken ten days after beginning amoxicillin therapy revealed mild to moderate epidermitis with moderate numbers of unidentified fungal hyphae and spores in the keratin layers, along with occasional scattered cocci. Hyperkeratosis was present, but was much less severe than in the previous biopsy.

The skin of the older female hippo began healing within 1 wk after beginning amoxicillin therapy, and visible weight gain was noted by 3 wk. Scars remain present on this animal’s skin, but no new lesions have been observed over the last year.

Discussion

Group G streptococci are normal flora of the skin and mucosa of animals.^1,6 However, these organisms can act as rapid opportunistic pathogens. Group G streptococcal infections tend to be severe and prolonged, and resemble those caused by Groups A and B. Skin infections and cellulitis in humans caused by Group G streptococci may be associated with abnormal lymphatic drainage or underlying malignancies.^1,6 Bacteremia, endocarditis, and glomerulonephritis are possible, although uncommon, sequelae to infection with Group G streptococci. Suggested antibiotic therapies for humans with streptococcal skin infections include penicillin or a first-generation cephalosporin, with surgical debridement if indicated.^1,6 Poor clinical response is often seen with erythromycin or chloramphenicol. Group G beta-hemolytic Streptococcus spp. was cultured at necropsy from an infected metacarpophalangeal joint in an adult male hippo with a chronic draining toe lesion (R. Junge and R.E. Miller, unpublished data).

The dermatitis in these cases is likely the result of opportunistic infection of skin lesions by normal skin flora. The inciting factor for the development of skin infections in these hippos remains unclear. Topical or intramuscular treatment is recommended in zoo medicine texts,^2,5 but topical treatment was ineffective in these animals. Intramuscular treatment was spectacularly unsuccessful. The mobility of the animal in the chute made neck injections unsafe, and penetration of the truncal skin could not be achieved with a pole syringe. Repeated darting was not used for antibiotic administration due to the need for multiple large darts, and concern that the hippo would become reluctant to enter the chute, making follow-up observations and manipulations more difficult. Hippos have a complex stomach with four sections,^2,5 and although they are not true ruminants, efficacy of oral treatment with antibiotics was questioned. However, dramatic improvements in skin condition was noted immediately after starting oral amoxicillin. The fungal infection noted in the second biopsy may represent a secondary invasion of the wounds, which resolved spontaneously upon treatment of the primary skin disease.

Death during immobilization of hippos with ultrapotent narcotics has been reported.^4,5 Juvenile hippos have been immobilized at the San Diego Zoo with detomidine and butorphanol.³ The low dose detomidine and butorphanol administered to this adult animal provided good immobilization for the procedures performed. However, the amount of head and limb movement retained, along with a surprise spontaneous recovery, made the use of a chute or containment device important for personnel safety. Pulse oximetry and electrocardiogram could not be obtained on this animal with our current probes. Anesthetic monitoring was limited to visual assessment of respiration rate and palpation of heart rate over the thorax. Blood collection is reported to be difficult in hippos.^4,5 The ease of collection of large amounts of blood from the ventral tail vein in this animal may have been due to a light level of anesthesia, as the blood was collected in the 10 min prior to spontaneous recovery. The lack of accurate weights and reference ranges for standard hematologic and serum chemistry values hampered assessment of the animal’s condition during the course of the disease.

Conclusions

Opportunistic infection with group G beta-hemolytic streptococci should be included in the differential diagnosis of skin disease in hippos. Treatment of infected hippos with amoxicillin at 15 mg/kg p.o., s.i.d. appears to be effective. Detomidine and butorphanol can be used for immobilization of adult hippos, although containment of the animal within a chute or cage is recommended for personnel safety. Blood should be collected from normal hippos when possible, to aid in the development of reference ranges for this species.

Acknowledgments

The successful resolution of this case would not have been possible without the hard work, persistence, and ingenuity of the Pachyderm staff. The Milwaukee County Zoo is also indebted to Dr. Gerald Dorff, MD for his advice and assistance with this and numerous other cases. As always, veterinarians across the country helped with advice and comments. Special thanks to the veterinarians of the San Diego Zoo, the St. Louis Zoo, and the Metro Washington Park Zoo for sharing anesthetic regimes, previous cases, and comparison tissues. Drs. Howard Steinberg and Annette Gendron-Fitzpatrick reviewed the histology slides. Dr. Bill Sadler and Karen Wright from Purina Mills, Inc. assisted in formulation and production of the fortified diet used during the treatment and recovery from dermatitis in these hippos.

Literature Cited

1.  Gallis, A. 1992. Viridans and β-hemolytic (non-group A, B, and D) streptococci. In: Principals and Practice of Infectious Diseases, 3rd ed. G.L. Mandell, R.G. Douglas, Jr., and J.E. Bennett (eds.). Churchill Livingstone, New York, New York, 1563–1572.

2.  Jarofke, D. 1993. Hippopotamidae (Hippopotamus). In: Zoo and Wild Animal Medicine, Current Therapy 3. M.E. Fowler (ed.). W.B. Saunders Co., Philadelphia, Pennsylvania, 522–525

3.  Morris PJ. 1996. Recent developments in anesthesia of exotic ungulates. Proceedings of the North American Veterinary Conference, 901–902.

4.  Ramsay, E.C., M.R. Loomis, K.G. Mehren, W.S.J. Boardman, J. Jensen, and D. Geiser. 1998. Chemical restraint of the Nile hippopotamus (Hippopotamus amphibius) in captivity. J. Zoo. Wildl. Med. In press.

5.  Taylor, D. and A. Greenwood.1986. Hippopotamidae (Hippopotamus). In: Zoo and Wild Animal Medicine, 2nd ed. M.E. Fowler (ed.). W. B. Saunders Co., Philadelphia, Pennsylvania, 967–969

6.  Vartian, C., P.I. Lerner, D.M. Shlaes, and K.V. Gopalakrishna. 1985. Infections due to Lancefield Group G streptococci. Medicine. 64: 75–88.