Dermatitis herpetiformis

An outbreak of a blistering disease was reported in a military unit training in the Arizona desert during heavy rain and flooding. In a unit of 249 personnel, 33 presented with dermatologic complaints, and 4 met stringent diagnostic criteria for dermatitis linearis. A fifth patient presented with symptoms and signs of "Nairobi eye." Staphylinid (rove) beetles related to the paederids, which have been responsible for vesicular dermatitis outbreaks in other parts of the world but not previously in the United States, were collected at the site. Reports in the antique scientific literature document paederids in the area after periodic floods. These findings suggest that rove beetle dermatitis should be added to the differential diagnosis of vesicular dermatitis in western North America.

Introduction

September 1997 brought unseasonable heavy rain to desert areas of the American southwest. On the night of September 24, during the second of two drenchings, Yuma, Arizona, received more than double its normal annual rainfall. Flooding in the adjacent mountains and desert included a remote training site where a unit of 249 Marines was subjected to heavy rain and corresponding runoff. In the days that followed, an acute blistering skin disorder prompted Marines to seek medical attention. One service member with an 8-cm flaccid bleb was evacuated for evaluation by a dermatologist. Another presented with unilateral conjunctivitis accompanied by gross palpebral swelling. A multidisciplinary team representing epidemiology, infectious disease, and entomology expertise deployed to the site to investigate.

Methods

Before the investigating team's arrival, one service member was transferred to the Naval Medical Center, San Diego, for dermatology evaluation. Biopsy was performed, and clinical specimens were referred to the Walter Reed Army Institute of Research and Aberdeen Proving Grounds to test for evidence of chemical warfare agents. This patient was nominally labeled case 1.

In the field, questionnaires were completed by all service members who had received medical evaluation for dermatologic complaints during the previous 2 weeks. A "possible" case was considered to be a service member with blistered skin. An epidemic curve was plotted, and history and exposure factors were recorded. Subsequently, a stringent case definition was established for heightened specificity: physician-characterized bullae of inapparent cause ("confirmed" case).

While this investigation was ongoing, another service member was referred for similar urgent dermatologic evaluation from a training area representing the same desert ecosystem but a separate military installation (Marine Corps Base Twentynine Palms) approximately 100 miles north of the original site. Exposure factors, history, symptoms, and physical findings were consistent with our case definition. Biopsy was performed.

Insect collections were conducted using light and pitfall traps. The latter were of the type described by Southwood' and were filled with approximately 200 ml of radiator antifreeze. Initial light trapping on October 9 was performed with CDC traps commonly used for trapping mosquitoes. Subsequent trapping on October 16 used a white bed sheet erected in a tent-like shape with several lights inside, including ultraviolet, incandescent, and fluorescent bulbs. Beetles that were attracted to the tent surface were collected in 4-dram vials and preserved in 70% isopropyl alcohol. Light tents were monitored from 30 minutes before dusk until approximately 3 hours after sundown. Staphylinid beetles were referred by team entomologists for precise identification by a coleopterist (J.H. Frank, University of Florida, Gainesville).

Results

All case presentations appeared to represent primary cutaneous disease, not systemic disease with cutaneous manifestation. No exposure factor or links between the cases could be identified other than an apparent association with heavy rainfall (Fig. 1). The four service members who met the stringent case definition demonstrated remarkable features in common: (a) a history of awakening and finding large, burning" painful bleb(s) with no recollection of trauma, bite, sting, or chemical, plant, or beetle contact; (b) absence of any pulmonary or systemic symptoms; and (c) absence of fever, lymphangitis, or lymphadenopathy. Blebs were flaccid and asymmetric, linear or curvilinear, resembling accidental droppings of a hot or caustic liquid on the skin of the neck or an exposed extremity. These four cases were diagnosed as dermatitis linearis. In two individuals, secondary blisters appeared in the drainage areas of bleb contents; -kissing" lesions were evident in the popliteal fossa of one individual (Fig. 2). In the absence of instrumentation, these lesions resolved spontaneously over a 5- to 7-day period (Fig. 3). One service member presented in the same time frame with abruptonset unilateral conjunctivitis and palpebral swelling. Treated empirically in the field, he experienced rapid resolution of symptoms with no sequelae. Twenty-eight other service members with reported dermatologic findings failed to meet the stringent case definition. Their symptoms may have been the result of mundane injuries.

Clinical specimens referred to Walter Reed Army Institute of Research and Aberdeen Proving Grounds returned negative for the presence of thiodiglycol metabolites, which are the salient diagnostic feature for vesicant warfare agents.2 Serum porphyrin levels were also negative. Skin biopsy findings on two individuals revealed a process of direct toxic damage to the skin, acute epidermal necrosis, and no underlying dermal process. A diagnosis of acute vesicating dermatitis, consistent with dermatitis linearis, was established. All patients recovered within 10 days without specific treatment.

Initial entomologic collections did not produce insects with suspected vesicating abilities. No large beetles of the family Meloidae were observed. Similarly, a floral survey did not detect any apparent agents of plant-induced dermatitis. However, efforts 1 week later using a light tent produced 22 adult beetles from the family Staphylinidae, a taxonomic group containing the genus Paederus. Paederids are the known cause of dermatitis linearis outbreaks in other parts of the world.>6 Expert identification revealed that 12 of the beetles were of the same subfamily as the paederids (Paederinae) (Fig. 4), but none of the collected species was known to cause vesication and none was of the genus Paederus.

After the initial cluster of cases, troops were advised of possible insect involvement and counseled to wash the skin immediately with soap and water after any contact with insects and to sleep under mosquito netting. No other cases were reported.

Discussion

Diagnostic considerations in this outbreak included irritant or allergic contact dermatitis, thermal or chemical burns, impetigo, herpes simplex/zoster, porphyria, pemphigus, dermatitis herpetiformis, and cutaneous manifestations of coccidioidomycosis. Because of the military context and location, concerns also included environmental tricothecene mycotoxins and vesicant warfare agents. Some Marine officers surmised that their personnel had been exposed to a blistering agent such as mustard gas or lewisite, but the United States does not currently weaponize such agents and there are no indications that such agents have ever been investigated or stored in the Yuma area. These concerns were also rebutted by the total absence of reported pulmonary or systemic symptoms in the unit service members and by the simultaneous referral of a case from a geographically proximate, but militarily separate, training area (Twentynine Palms).

Insects, however, commonly use vesicant agents for predator deterrence.7 Beetles of the genus Paederus, family Staphylinidae (commonly known as rove beetles), contain vesicant amides that produce dermatitis and ophthalmic lesions in people and animals.353 Paederids measure an inconspicuous 6 to 8 mm long by 1 mm wide and weigh about 4 mg. Their natural habitat is among decaying vegetable matter, often in association with rodent burrows. They do not bite or sting, but they may be provoked by minor trauma to release hemolymph containing the vesicant agents paederin, pseudopaederin, or pederone. They characteristically emerge as adults in association with rains, at which time they are attracted to lights and may come into contact with people.6,7,9,10 Toxic effects of various species of Paederus have been described in Brazil, Paraguay, Australia, Thailand, Okinawa, Vietnam, Sri Lanka, Israel, Egypt, Uganda, Tanzania, and Kenya.4,7,9,11-15 The popular name for a Paedeniscaused dermatitis epidemic in Nigeria in 1990 was "wake and see," reflecting painless contact, the nocturnal habit of the insect, and the 12 to 24 hours taken for the vesicant fluid to act.3 Similar outbreaks have been described in Egypt in 19944 and Kenya in 1998 after sudden floods,15 in Brazil, Sri Lanka, and Australia during the rainy season,9 and throughout Africa at the end of the rainy season.3 Cases have been described in British Army troops deployed in Africa5 and in children in Italy.6 The conjunctivitis case in our series is consistent with descriptions of the condition known in other parts of the world as "Nairobi eye" or "Christmas eye."14,16 This is believed to be the ophthalmic analog of dermatitis linearis, caused by direct contact between beetle and eye or inadvertent rubbing of the eye with a vesicantcontaminated hand. Perplexingly, attempts to reduplicate skin or eye findings in volunteers using crushed beetles have met with only occasional success, owing perhaps to histocompatibility locus antigen type or other undefined individual determinants.4,17

Text discussions of vesicular dermatitis consider paederid vesication in tropical and warm temperate regions but specifically exclude the United States." However, antique sources dating from 1858 and 1876 report Paederus species endemic in southern Arizona and "lower" California. ls l9 Paederus ustus was described as "very abundant" in 1858 during seasonal flooding along the Colorado River "about the junction of the Gila," the location of modern Yuma.ls This was in an era before massive dam construction on the Colorado River reduced water flow and flooding along the river. Human dermatitis cases in this geographical area have not previously been associated with rove beetles (Justin Schmidt, University of Arizona Poison Control Center, personal communication). Attributions are commonly made to the much larger beetles of the family Meloidae, which are well known in agricultural lands to the east of Yuma and are frequently described by patients presenting with blisters 2 to 10 hours after an exposure (Jude McNally, University of Arizona Health Sciences Center, personal communication). Victims of meloid vesication often recall grasping, restraining, or otherwise molesting large beetles before blister formation.20 The toxin present in meloids, cantharidin, is distinct from the vesicants isolated from paederids.'17 Cantharidin, the active ingredient in the apocryphal aphrodisiac "Spanish fly," typically produces a tingling or burning sensation within 10 minutes, in contrast to the 12 to 36 hours after contact with paederids.s Although meloid beetles remain a possible alternative explanation for our case series, this seems improbable given (a) the troops' lack of reported handling or even seeing large beetles in the habitat, (b) the Sake and see" recognition of the lesions, (c) the described symptomatology of the lesions, (d) our conjunctivitis case (not described, characteristically, in association with meloid beetles), and (e) the remarkable synchronicity of cases with area flooding.

Effects of excess precipitation in the Yuma area evident to the casual visitor in early October included exuberant foliage (with unseasonable flowering of creosote bushes) and prominent alluvial fans on desert watercourses. We suspect that heavy rain, in the context of an overwhelming flood of desert rodent burrows, caused a flight of arthropod inhabitants. Once outside, some insects, including vesicating paederids, were attracted to artificial lights at the remotely sited Marine encampment, resuiting in exposures to personnel. The fact that no vesicating paederids were collected is not surprising because of the ephemeral climatic conditions and because beetles dispersed or returned to their preferred underground habitat by the time the investigators arrived. Success in efforts to capture vesicating beetles has long been noted to be exquisitely weather- and temperature-dependent.21

To our knowledge, this is the first report of suspected rove beetle involvement in an outbreak of vesicular dermatitis in western North America. On-site collections of subfamily Paederinae species and historical descriptions of paederid occurrences in the immediate area in association with seasonal flooding support this proposition. Further studies with temperatureand season-specific insect collections to obtain greater numbers of beetles are necessary. As with the vector mosquitoes of dengue and malaria, global climatologic factors or the El Nino weather pattern may be involved.22

This report is also a study in the broad differential diagnosis for vesicular dermatitis lesions at a time of increased attention to chemical and biological warfare threats.2 Initial concerns of exposure to vesicating war agents, although unfounded, demonstrate the need to identify environmental elements that cause symptoms mimicking those of weapons. Ignorance of such elements can cause unnecessary stress and confusion for personnel deployed to the battlefield where chemical and biological agents are considered potential threats.

Acknowledgments

The authors express appreciation to J. Frank, PhD, for insect identification; LCDR R. Martschinske, MC USN, LT M. Novak, MSC USN, and LT S. Lee, MSC USN, for assistance in field work; CAPT E. Ledbetter, MC USN, for methodological assistance; and CAPT M. Wallace, MC USN, and LCDR S. Cope, MSC USN, for critical review of the manuscript.

References

1 Southwood TRE: Relative methods of population measurement and the derivation of absolute estimates. In Ecological Methods, pp 247-8. London, Chapman and Hall, 1978.

2 Medical aspects of chemical and biological warfare. In Textbook of Military Medicine, pp 197-222. Washington, DC, US Army Office of the Surgeon General, 1997.

3 Frank JH, Kanamitsu K: Paederus, sensu lato (Coleoptera: Staphylinidae): natural history and medical importance. J Med Entomol 1987; 24: 155-91.

4 Morsy TA, Arafa MA, Younis TA, et al: Studies on Paedems aifierti Koch (Coleoptera: with special reference to the medical importance. J Egypt Soc Parasitol 1996; 26: 337-52.

5 Williams AN: Rove beetle blistering (Nairobi eye]. J R Army Med Corps 1993;139: 17-9.

6. Gelmetti C, Grimalt R: Paederus dermatitis: an easily diagnosable but misdiagnosed eruption. Eur J Pediatr 1993; 152: 6-8.

7. Harwood RF, James MT: Family Staphylinidae. In Entomology in Human and Animal Health, pp 442-3. New York, Macmillan, 1979.

8. Borroni G, Brazzelli V, Rosso R, et al: Paederus fuscipes dermatitis: a histopathological study. Am J Dermatopathol 1991: 13: 467-74.

9. Todd RE, Guthridge SL, Montgomery BL: Evacuation of an aboriginal community in response to an outbreak of blistering dermatitis induced by a beetle (Paederus australis). Med J Aust 1996; 164: 238-40.

10. George AO: Paederus dermatitis: a mimic-contact dermatitis. Int J Dermatol 1993; 29: 212-3.

11. Fox R: Paederus (Nairobi fly) vesicular dermatitis in Tanzania. Trop Doct 1993; 23: 17-9.

12. Armstrong RK, Winfield JL: Staphylinidae dermatitis in Okinawa. J Med Entomol 1968: 5: 362.

13. Allen AM: Skin diseases in Vietnam. In Internal Medicine in Vietnam, Vol 1. pp

134-6. Washington, DC, Center of Military History, 1989. 14. Kamaladasa SD, Perera WD, Weeratunga L: An outbreak of Poederus dermatitis in a suburban hospital in Sri Lanka. Int J Dermatol 1997; 36: 34-6. 15. Hugh-Jones M: P. crebinpunctatus and P. sebaeus emerge in aftermath of Kenya floods. ProMED Digest 1998; 98: 27.

16. Minton SA, Bechtel HB: Arthropod envenomation and parasitism. In Wilderness Medicine: Management of Wilderness and Environmental Emergencies, p 752. Edited by Auerback PS. St. Louis, MO, Mosby, 1995. 17. McCrae AW, Visser SA: Paederus (Coleoptera: Staphylinidae in Uganda. I. Outbreaks, clinical effects, extraction and bioassay of the vesicating toxin. Ann Trop Med Parasitol 1975; 69: 109-20.

18. Le Conte JL: Description of new species of Coleoptera, chiefly collected by the United States and Mexican boundary commission, under Major W. H. Emory,

U.SA. Proc Acad Nat Sci Phila 1858; 59-89.

Austin EP: On the species of Sun/us and Perus found in the United States. Proc Boston Soc Nat Hist 1876; 4-11.

20. Smith RL: Venomous Animals of Arizona, pp 44-6. Tucson, University of Arizona, 1989.

Swarts WB, Wanamaker JF: Skin blisters caused by vesicant beetles. JAMA 1946: 131: 594-5.

Taubes G: Apocalypse not. Science 1997; 278: 1004-6.

Guarantor: CAN Patrick E. Olson, MC USN

Contributors: LCDR David M. Claborn, MSC USN*; LCDR James M. Polo, MC USN^; CAPT Patrick E. Olson, MC USN*; LCDR Kenneth C. Earhart, MC USN^; LCDR Sterling S. Sherman, MC USN^^

*Navy Environmental and Preventive Medicine Unit No. 5, Departments of Entomology and Epidemiology, Naval Station, Box 368143, San Diego, CA 92136.

^ Naval Medical Center San Diego, Departments of Dermatology and Medicine, San Diego, CA 92134.

^^ Marine Expeditionary Force, Preventive Medicine Officer, Marine Corps Base

Camp Pendleton, Box 555300, Camp Pendleton, CA 92055. Case 1 was presented as a dermatology clinical case report at the 13th Annual

Academic Research competition, Naval Medical Center San Diego, April 3, 1998 (presented by J.M.P.).

This manuscript was received for review in May 1998 and was accepted for publication in July 1998.

Reprint & Copyright by Association of Military Surgeons of U.S., 1999.

Copyright Association of Military Surgeons of the U.S. Mar 1999
Provided by ProQuest Information and Learning Company. All rights Reserved