Predictive Factors for Gnathostoma Seropositivity in Patients Visiting the Gnathostomiasis Clinic at the Hospital for Tropical Diseases, Thailand During 2000-2005

SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

PREDICTIVE FACTORS FOR GNATHOSTOMA SEROPOSITIVITY IN PATIENTS VISITING THE GNATHOSTOMIASIS CLINIC AT THE HOSPITAL FOR TROPICAL DISEASES, THAILAND DURING 2000-2005

Valai Bussaratid1, Paron Dekumyoy2, Varunee Desakorn1, Naowarat Jaroensuk3 Busaya Liebtawee3, Wallop Pakdee2 and Yupaporn Wattanagoon1

1Department of Clinical Tropical Medicine, 2Department of Helminthology, 3Hospital for Tropical Diseases, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand

Abstract. This was a retrospective study of patients having Gnathostoma antibody testing at the Hospital for Tropical Diseases, Bangkok during 2000-2005 to investigate predictive factors for Gnathostoma seropositivity in patients attending the Gna- thostomiasis Clinic. Out of 849 patients tested, 531 (62.5%) were Gnathostoma seropositive. The median absolute eosinophil counts were 464 (0-16,796) and 326.5 (0-10,971) cells/mm3 in seropositive and seronegative patients, respectively (p<0.001). Differences in a history of cutaneous swelling, the habit of eating raw meat, eosinophilia (>500 cells/mm3), and the frequency of cutaneous swellings between seropositive and seronegative patients were all statistically significant. Patients with a history of eating raw meat and a history of cutaneous swelling were at 2.1 and 1.8 times more likely to be Gnathostoma seropositive, respectively. Logistic regression analysis showed eosinophilia was not a predictive factor for Gnathostoma seropositivity. Key words: Gnathostoma, predictive factor, seropositivity, Thailand

INTRODUCTION swelling, cutaneous creeping eruption, acute radicular pain with migration signs, Human gnathostomiasis, a tissue and meningoencephalitis (Suntharasamai, parasitic disease caused by advanced third 2003). The most common clinical presen- stage larvae of Gnathostoma spinigerum, is tation in patients attending the Gnathos- prevalent in Thailand. Gnathostomiasis is tomiasis Clinic at the Hospital for Tropi- caused by eating raw fresh water fish, cal Diseases, Bangkok, Thailand, is cuta- frogs, snakes, chickens, ducks or birds con- neous migratory swelling. However, the taining Gnathostoma larvae. Clinical pre- presentation of cutaneous migratory sentations include cutaneous migratory swelling is not specific for gnathostomia- Correspondence: Dr Valai Bussaratid, Depart- sis. The differential diagnoses for cutane- ment of Clinical Tropical Medicine, Faculty of ous migratory swelling includes loiasis, Tropical Medicine, Mahidol University, 420/6 which is prevalent in West and Central Ratchawithi Road, Bangkok 10400, Thailand. Africa (Grove, 2000), paragonimiasis, spar- Tel:+66(0) 2354 9100 ext 2065, 2067 ganosis, other tissue parasitic infestations, E-mail: [email protected] and localized angioedema due to allergic

1316 Vol 41 No. 6 November 2010 FACTORS FOR GNATHOSTOMA SEROPOSIVITY reactions. Several serological tests for the times with tap water until no appearance detection of antibodies against Gnathos- of blood. The livers were chopped into toma antigens have been developed for the small pieces and digested with artificial confirmation of clinical gnathostomiasis. gastric juice (1% pepsin-HCl solution) at (Suntharasamai et al, 1985; Dharmkrong- 37ºC. After two hours of incubation, the At et al, 1986; Tada et al, 1987; Maleewong suspension was passed through a screen et al, 1988; Tapchaisri et al, 1991). and collected. The suspension collected Unfortunately, all currently available was left to sediment, then washed several serodiagnostic tests, including the times with distilled water and examined immunoblot technique, are not helpful for for the presence of larvae under a stereo- differentiating recent and past Gnathostoma microscope. All larvae were washed sev- infection. This study aimed to investigate eral times with distilled water and kept at predictive factors for Gnathostoma sero- -75ºC, or used directly in a further step. positivity in patients attending the Gna- Parasites were ground with alumina pow- thostomiasis Clinic. der in distilled water and the suspension was sonicated using probe No. 418, at magnification No. 4 (Sonicator Heat Sys- MATERIALS AND METHODS tems, Lakewood, NJ), for 10 minutes at in- This was a retrospective case-control tervals of 1 minute. The suspension was study carried out at the Gnathostomiasis then centrifuged at 10,000 rpm for 60 min- Clinic, Hospital for Tropical Diseases in utes, and the supernatant was collected Bangkok and the Immunodiagnostic Unit and determined for protein content deter- for Helminthic Infections, Department of mined by Coomassie Blue Plus Protein Helminthology, Faculty of Tropical Medi- Assay Reagent Kit (Pierce, Rockford, IL). cine, Mahidol University, Bangkok, Thai- Immunoblot technique land. A 13% separating gel of SDS-polyacry- The study population consisted of 849 lamide gel carried the SDS-treated proteins patients whose medical records were re- in a single well. After electrophoresis viewed for clinical baseline and laboratory transfer, a nitrocellulose sheet was treated data, including presenting symptoms, di- with blocking buffer containing 2% skim agnosis, eating habits, immunoblot analy- milk, 0.02% NaN3-PBS, pH 7.4. The blot sis results, CBC results and stool exami- was cut into small strips, which contained nations (both simple smear and formalin 15 µg of antigen per strip. The strips were ether concentration techniques) done then reacted with serum samples diluted within 4 weeks prior to or after the per- to 1:50 in 0.05% Tween20, 0.02% NaN3- formance of immunoblot analysis. Eosino- PBS, pH 7.4, on a rocking platform at room philia was defined as an absolute eosino- temperature for 14-16 hours. The strips phil counts ≥500 cells/mm. were incubated with diluted peroxidase con- jugate anti-human IgG (SouthernBiotech, Antigen preparation Birmingham, AL) (1:1,000) in 0.05% Third-stage larvae of G. spinigerum Tween20-PBS, pH 7.4, for 2 hours, washed, were collected from the livers of freshwa- and color developed with reagent contain- ter eels purchased from local vendors in ing 2, 6-dichlorophenol indophenol. The Bangkok. The eel livers were separated excess reaction was removed by washing from the intestines and washed several with distilled water (Dekumyoy et al,

Vol 41 No. 6 November 2010 1317 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

Table 1 Factors that may predict the results of Gnathostoma antibody detection by immunoblot.

Characters Gnathostoma p-value Seropositive Seronegative No. (%) No. (%)

Sex Male 190 (61.1) 121 (38.9) 0.568 Female 340 (63.3) 197 (36.7) History of cutaneous swellinga Present 311 (69.0) 140 (31.0) <0.001 Absent 219 (55.2) 178 (44.8) History of eating raw meatb Yes 468 (66.0) 241 (34.0) 0.001 No 25 (43.9) 32 (56.1) Eosinophiliac Yes 203 (66.8) 101 (33.2) 0.044 No 232 (59.0) 161 (41.0) Frequency of cutaneous swellingd ≤1/month 140 (77.3) 41 (22.7) 0.049 >1/month 56 (65.1) 30 (34.9) Location of cutaneous swellinge Head and neck 48 (60.0) 32 (40.0) 0.053 Extremity 139 (74.7) 47 (25.3) Trunk or mixed location 30 (66.7) 15 (33.3) Result of stool examination for intestinal parasitef Positive 23 (69.7) 10 (30.3) 0.462 Negative 164 (61.4) 103 (38.6) aThere was a patient whose clinical manifestation on cutaneous swelling was not recorded. b There were 766 patients with known eating habits; 493 and 273 patients in Gnathostoma seropositive and seronegative, respectively. c There were 697 patients who had CBC results; 435 and 262 patients were Gnathostoma seropositive and seronegative, respectively. dThere were 267 patients, who had more than 1 episodes of cutaneous swelling, whose frequencies of cutaneous swelling were recorded. eThere were 311 patients whose locations of cutaneous swelling were recorded. fThere were 187 and 113 Gnathostoma seropositive and seronegative patients, respectively.

2002). For standard control, a strip of each Statistical analysis blotted nitrocellulose sheet was treated Clinical and laboratory data were ana- with monoclonal antibody (kindly pro- lyzed by SPSS for windows. Data were vided by Prof Wanpen Chaicumpa, De- summarized as mean (SD) for normally partment of Parasitology, Faculty of Medi- distributed quantitative data, median cine, Siriraj Hospital, Mahidol University, (range) for non-normally distributed Thailand) against 24 kDa antigen detected quantitative data and number (percentage) with peroxidase conjugated anti-mouse for categorical data. Group comparison IgG (KPL, Galthersbury, MD) at 1:1,000. was done with the t-test for quantitative

1318 Vol 41 No. 6 November 2010 FACTORS FOR GNATHOSTOMA SEROPOSIVITY

Table 2 Predictive factors for Gnathostoma seropositivity.

Predictive factors Odds ratio 95% CI p-value

History of eating raw meata 2.106 1.14-3.88 0.017 History of cutaneous swellinga 1.840 1.32-2.56 <0.001 Eosinophiliaa 1.176 0.83-1.65 0.354 aNot present was used as a reference. data with normal distribution, the Mann- sue disease, it was not known whether Whitney U-test for non-normal distribu- cutaneous swelling was present or absent. tion, and the chi-square test or Fisher’s Of 766 patients whose eating habits were exact test where appropriate, for categori- known, 709 (92.6%) ate raw meat and only cal data. Logistic regression was per- 57 (7.4%) did not eat raw meat. formed to confirm an association with Of the 849 patients tested for Gnathos- various predicting factors with G. toma antibodies by immunoblot technique, spinigerum seropositivity. 531 (62.5%) and 318 (37.5%) patients were This study was approved by the Eth- Gnathostoma seropositive and seronega- ics Committee of Faculty of Tropical Medi- tive, respectively. The median ages of pa- cine, Mahidol University. tients in the former and latter groups were 37 (8-79) and 37 (5-79) years, respectively (p = 0.722). The median absolute eosino- RESULTS phil count was significantly higher in the Eight hundred forty-nine patients seropositive group [464 (0-16,796) vs 326.5 / 3 p were enrolled in this study; the median age (0-10,971) cells mm <0.001]. was 37 (5-79) years. Three hundred eleven There were significant differences be- (36.6%) patients were male. Of those 849 tween the Gnathostoma seropositive and patients, 451(53.3%) patients had cutane- seronegative groups in the presence of ous swellings. The clinical diagnoses of cutaneous swelling, raw meat eating hab- patients with no cutaneous swelling in- its, eosinophilia (absolute eosinophil count cluded urticaria (23.8%), tactile hallucina- >500 cells/mm3), and frequency of cutane- tions/tingling sensation (9.2%), delusions ous swellings. There was no significant of parasitosis (1.3%), localized pain in an difference in location of cutaneous swell- extremity (0.7%), anxiety (0.1%), pruritus ing between Gnathostoma seropositive and without skin lesions (2.4%), dermatitis seronegative patients (Table 1). (2.4%), skin nodules (0.7%), headaches Logistic regression analysis revealed (0.9%), arthralgia/arthritis (0.8%), asymp- patients with a history of eating raw meat tomatic eosinophilia (0.6%), allergies and a history of cutaneous swellings were (0.6%), myalgia (0.5%), healthy (1.9%), 2.1 and 1.8 times more likely to be Gna- creeping eruption (0.2%), cellulitis (0.1%), thostoma seropositive, respectively. How- hordeolum (0.1%), parotitis (0.1%), cancer ever presence of eosinophilia was not as- (0.1%), and feeling numbness (0.1%). In sociated with Gnathostoma seropositivity one patient with possible connective tis- (Table 2).

Vol 41 No. 6 November 2010 1319 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

DISCUSSION tending the Gnathostomiasis Clinic. Hope- fully, these predictive factors will help cli- Patients came to the Gnathostomiasis nicians confidently diagnose gnathosto- Clinic for various reasons. Cutaneous swell- miasis. ing, pruritus and skin rashes are common The study found that 219 (55.2%) of clinical presentations. We used serological patients with no cutaneous swelling were testing to help confirm the diagnosis but it Gnathostoma seropositive. These included may not be able to distinguish past from 2 patients with creeping eruption, 3 pa- present infection. In this study we looked tients with asymptomatic eosinophilia, 115 at risk behaviors, clinical manifestations patients with urticaria, and 99 patients and laboratory findings to find associations with other clinical diagnoses. The rationale with Gnathostoma seropositivity. for Gnathostoma seropositive may be sub- We found around 90% of patients at- clinical infection, unrecognized cutaneous tending the Gnathostomiasis Clinic had a swelling or false positivity. history of raw meat consumption, hence This study showed a history of eating they were at risk for many parasitic infec- raw meat and a history of cutaneous swell- tions including gnathostomiasis. The preva- ing were predictive factors for Gnathostoma lence of gnathostomiasis in Thailand has seropositivity. The findings are consistent been estimated to be 0.4 % (Suntharasamai, with current clinical practice. 1987). Eosinophilia, or absolute eosinophil In Thailand, the differential diagnosis count ≥ 500 cells/mm3, are not necessary for cutaneous swelling includes allergic for diagnosing gnathostomiasis because it related localized angioedema, gnathosto- is not always present (Sirikulchayanonta miasis, insect bite reactions, and trauma and Viriyavejkul, 2001; Dekumyoy et al, related cutaneous swelling. With gnathos- 2002; Magana et al, 2004). Eosinophilia is tomiasis, a cutaneous swelling usually not specific for gnathostomiasis, since it lasts longer than 24 hours and is often may be associated with various diseases migratory. To diagnose gnathostomiasis, or conditions (Tefferi, 2005). However, this a history of eating raw meat, especially study revealed significant differences in raw fresh water fish, snake, frog, or poul- absolute eosinophil counts and eosino- try, which are intermediate hosts or philia between Gnathostoma seropositive paratenic hosts for Gnathostoma, and Gna- and seronegative patients. Nevertheless, thostoma seropositivity by immunoblot are eosinophilia was not a predictive factor for required, in addition to the clinical symp- Gnathostoma seropositivity. tom of cutaneous migratory swelling. A disadvantage of this study was the Unfortunately, cross-reactivity with data were incomplete. First, cutaneous other parasites in some patients with a di- swelling was not usually evident on physi- agnosis of paragonimiasis, trichinosis, cal examination but relied on the patients’ opisthorchiasis, cysticercosis, strongyloidi- verbal report. Second, there was no infor- asis, or fascioliasis, has been reported with mation regarding duration of each swell- the immunoblot test (Tapchaisri et al 1991; ing episode. In case of rapidly resolved Laummuanwai et al, 2007). Therefore, we cutaneous swelling, allergy associated lo- aimed to investigate predictive factors for calized angioedema should be considered. Gnathostoma seropositivity in patients at- Third, some patients were seen after a

1320 Vol 41 No. 6 November 2010 FACTORS FOR GNATHOSTOMA SEROPOSIVITY single episode of non-migratory cutaneous thostomiasis. Parasitol Res 2007; 101: 703- swelling. These factors caused problems 8. when an investigator looked for a classic Maleewong W, Morakote N, Thamasonthi W, clinical presentation of cutaneous gnathos- Charuchinda K, Tesana S, Khamboon- tomiasis. Fourth, not all patients were in- ruang C. Serodiagnosis of human gna- terviewed in detail for habits of eating raw thostomiasis. Southeast Asian J Trop Med intermediate or paratenic hosts for gna- Public Health 1988; 19: 201-5. thostomiasis. In theory, eating raw meat, Magana M, Messina M, Bustamante F, Cazarin such as pork or beef, does not put one at J. Gnathostomiasis: clinicopathologic risk for gnathostomiasis. Therefore the study. Am J Dermatopathol 2004; 26: 91-5. data collection may not reflect true risk Sirikulchayanonta V, Viriyavejkul P. Various exposure to Gnathostoma parasite. morphologic features of Gnathostoma spinigerum in histologic sections: report of Selection bias should be considered as 3 cases with reference to topographic well, because the majority of the patients study of the reference worm. Southeast seen at the Gnathostomiasis clinic were re- Asian J Trop Med Public Health 2001; 32: ferral cases and were very much con- 302-7. cerned about gnathostomiasis. Suntharasamai P, Desakorn V, Migasena S, Bunnag D, Harinasuta T. ELISA for REFERENCES immunodiagnosis of human gnathostomiasis. Southeast Asian J Trop Med Public Dekumyoy P, Sanduankiat S, Nuamtanong S, Health 1985; 16: 274-9. et al. A seven year retrospective evaluation of gnathostomiasis and diagnostic Suntharasamai P. Gnathostomiasis. In: Weatherall DJ, Ledingham JGG, Warrell specificity of immunoblot. Proceedings of DA, eds. Oxford textbook of medicine. 2nd the First International Meeting on Gna- ed. London: Oxford University Press, thostomiasis. Bangkok: Faculty of Tropi- 1987: 5.558-61. cal Medicine, Mahidol University, 2002: 45-8. Suntharasamai P. Gnathostomiasis. In: Warrell Dharmkrong-At A, Migasena S, Suntharasamai DA, Cox TM, Firth JD, Benz, Jr EJ, eds. th P, Bunnag D, Priwan R, Sirisinha S. En- Oxford textbook of medicine. 4 ed. New zyme-linked immunosorbent assay for York: Oxford University Press, 2003: 815- detection of antibody to Gnathostoma anti- 7. gen in patients with intermittent cutane- Tada I, Araki T, Matsuda H, Araki K, Akahane ous migratory swelling. J Clin Microbiol H, Mimori T. A study on immunodiag- 1986; 23: 847-51. nosis of gnathostomiasis by ELISA and Grove DI. Tissue nematodes (trichinosis, dra- double diffusion with special reference to cunculiasis, filariasis). In: Mandell GL, the antigenicity of Gnathostoma doloresi. Bennett JE, Dolin R, eds. Mandell, Douglas, Southeast Asian J Trop Med Public Health and Bennett’s principles and practice of in- 1987; 18: 444-8. fectious diseases. 5th ed. Philadelphia: Tapchaisri P, Nopparatana C, Chaicumpa W, Churchill Livingstone, 2000: 2947-8. Setasuban P. Specific antigen of Gnathos- Laummaunwai P, Sawanyawisuth K, Intapan toma spinigerum for immunodiagnosis of PM, Chotmongkol V, Wongkham C, human gnathostomiasis. Int J Parasitol Maleewong W. Evaluation of human IgG 1991; 21: 315-9. class and subclass antibodies to a 24 kDa Tefferi A. Blood eosinophilia: a new paradigm antigenic component of Gnathostoma in disease classification, diagnosis, and spinigerum for the serodiagnosis of gna- treatment. Mayo Clin Proc 2005; 80: 75-83.

Vol 41 No. 6 November 2010 1321