RAPID ASSESSMENT FOR LYMPHATIC FILARIASIS IN CENTRAL NIGERIA: A COMPARISON OF THE IMMUNOCHROMATOGRAPHIC CARD TEST AND HYDROCELE RATES IN AN AREA OF HIGH ENDEMICITY

INTRODUCTION

Lymphatic filariasis due to infection with Wuchereria bancrofti or Brugia malayi is estimated to infect at least 128 million persons worldwide.¹ The third most endemic country in the world for this disease (after India and Indonesia) is Nigeria, where it is caused by W. bancrofti, and 22.1% of the population is thought to be infected.² In 1997, the World Health Assembly targeted lymphatic filariasis for elimination through a strategy of mass drug administration (MDA). This is currently recommended in communities known or suspected to have an prevalence of lymphatic filariasis greater than 1%.³ In sub-Saharan Africa (including Nigeria), MDA in such communities will use single-dose, annual, simultaneous administration of oral ivermectin (Mectizan®; donated by Merck & Co., Rahway, NJ) and albendazole (donated by GlaxoSmithKline, Research Triangle Park, NC) tablets.⁴

To determine the target areas for MDA, community serologic surveys are recommended using a rapid immunochromatographic card test (ICT) that qualitatively tests finger prick blood for W. bancrofti circulating Og4C3 antigen.⁵ The ICT is rapid, highly sensitive and specific, and unlike blood examinations for parasitemia, does not have to take place at night, when the microfilaria of the parasite circulate. In Africa, the two-stage assessment protocol calls for selection of one or two villages in an administrative unit, then ICT testing of a random sample 100 individuals (males and females) in that village. Any positive individual will result in the decision to provide MDA throughout the administrative unit.⁶ Given the ICT’s expense (costing more than US$1.50 per test), the surveys proposed are limited in their geographic scope. Nigeria, for example, has more than 700 Local Government Areas (LGAs) to be tested, and ICT surveys in just one village per LGA would require well over $100,000 for the purchase of all the required ICT cards.

Hydrocele is the most common clinical manifestation of lymphatic filariasis, and its potential for use as a survey tool has been suggested.⁷ Up to 40% of men actively or previously infected with W. bancrofti exhibit hydrocele, and in regions of northern Ghana and coastal Kenya endemic for this disease hydrocelectomy accounts for 25% of all surgical procedures.^8–10 Hydrocele surveys would be particularly valuable in areas classified as non-endemic by a single village ICT sample where false-negative results lead to a no treatment decision. Such misclassifications might result in entire LGA populations being left out of the MDA program and thus continued transmission of lymphatic filariasis in these regions. If hydrocele surveys had adequate predictive value for lymphatic filariasis (e.g., ICT positivity), they could be conducted more extensively to supplement ICT surveys at a lower cost and in villages where local public health officials believe that this disease is endemic. For these reasons, we undertook a comparative study in Nigeria to evaluate if village hydrocele prevalence could serve as a supplement to the ICT.

MATERIALS AND METHODS

The study was approved by the Internal Review Board of Emory University (Atlanta, GA) and the Ministries of Health of Plateau and Nasarawa States, and carried out between January 1999 and January 2000. Plateau and Nasarawa States are located in central Nigeria in the region having a mean elevation of 1200 meters; the 3.6 million inhabitants are a Hausa-speaking group, and the majority live in rural agricultural areas. The two states have 30 LGAs, 12 of which contain villages that have been given mass ivermectin therapy for onchocerciasis since 1993.^11,12 The two LGAs (Pankshin LGA [Plateau State] and Akwanga LGA [Nasarawa State]) selected for study are LGAs being given mass ivermectin therapy, and were also pilot areas for a lymphatic filariasis MDA program that was launched in 2000 in an integrated fashion with the onchocerciasis program.¹³

After explaining the purposes of the study to the village chief and traditional leadership council and obtaining their permission, all resident adult males of the village (16 years of age and older) were asked to gather at a central point in the village. The study was explained in Hausa, and those willing to participate were asked to form two lines, one with males ≥ 40 years old and another with males < 40 years old. Fifteen males were then randomly selected (every second or third person depending on the number of persons present) from each line, resulting in an overall sample of 30 men of various ages. This method was used to obtain a better distribution of ages to correct for the observation that hydrocele rates increase with age.¹⁴

After individual verbal consent was obtained, the patients were assigned an ID number and then taken to a private room or screened examination area and asked in Hausa (which was often translated from Hausa to the local language) their names, ages, occupations, and if they had a swollen scrotum. They were then asked to partially disrobe, and one of two Ministry of Health nurses (IG or JU) trained to differentiate hydrocele from inguinal hernia performed a physical examination focused on the scrotum and inguinal area. Clinical hydrocele (referred to herein as hydrocele) was diagnosed based on the finding of a nontender, soft, fluid-filled mass whose superior limit could be defined the examining finger. Transillumination was difficult due to heavy scrotal pigmentation and lighting conditions and was not included as a criteria. Clinical hydrocele was distinguished from indirect inguinal hernia (which would change with cough or straining, show an inguinal swelling at the internal ring, and whose superior limit of mass in scrotum not defined with the examining finger), epididymitis (tenderness), tumor (painless, hard and irregular), scrotal lymphedema (internal scrotal contents normal on palpation), and inguinal adenopathy (hanging groin). In the initial village examinations, a physician (FOR or ESM) was present to verify that the nurses were capable of an accurate clinical distinction of hydrocele from these other entities. If a hydrocele was present, its size was determined (in the event of bilateral hydrocele, the largest hydrocele was measured and recorded). Initially, the physicians and nurses attempted to measure the circumference of the hydrocele using a disposable tape measure, but this proved difficult and poorly reproducible. After the first 50 villages, the procedure was altered and examiners were asked to subjectively equate hydrocele size (volume) to the sizes of three fruits kept by the team during the surveys (an orange, coconut, and papaya). The sizes were then converted to a standard centimeter circumference at the time of data entry. Data were not recorded on presence of inguinal hernia or surgical scars.

Immediately after the physical examination, ICT testing was performed following kit instructions (ICT Diagnostics, Balgowlah, Australia, product number FLO.1; patent now sold and produced as NOW®, ICT filariasis kits; Binax, Portland, ME). The patient’s left index finger was cleaned with 70% isopropanol and then punctured using a sterile lancet. The initial sample of blood was removed using a cotton swab, and sufficient fresh blood was then obtained to fill a 100-μl capillary tube. The blood was then transferred from the capillary tube to the pad on the ICT test kit card; two drops of the accompanying reagent were then added to the pad and the card was sealed. The results of each ICT test card were read after 15 minutes. A positive result was when two pink lines appeared on the card’s window, and a negative result was when a single line was seen. Test results with the individual’s ID number were recorded both on the card, and on each individual’s data sheet. Village summary data sheets were prepared at the end of each visit on which the 30 ID numbers and their corresponding ICT results (positive/negative) and hydrocele results (positive/negative) were recorded.

Data were entered into a computer and statistical analysis was carried out in Epi Info 6.04d (Centers for Disease Control and Prevention, Atlanta, GA) and SAS version 8.0 (SAS Institute, Inc., Cary, NC) software packages. Correlation analyses were carried out for ICT and hydrocele results at both the individual and village levels (e.g., village ICT prevalence compared with village hydrocele prevalence). Villages were considered to have endemic filariasis if at least one individual from the village sample had a positive ICT test result for filarial antigen. Individual analysis also included evaluating the relationships between age, history of scrotal swelling, ICT, hydrocele presence, and size of hydrocele. Chi-square tests were used to compare dichotomous variables; continuous variables were compared using the Kruskal-Wallis (non-parametric) test.

RESULTS

One hundred fifty villages were visited during the study period: six villages were excluded as result of having sample sizes of less than 30 men, leaving a total of 144 villages for analysis (86 in Akwanga LGA and 58 villages in Pankshin). Village ICT and hydrocele rates were analyzed from village summary forms for 4,320 examinations in 144 villages; individual data forms (with age, history, physical examination, and ICT results) were available for 4,120 individuals (95% of all participants).

DISCUSSION

Our data demonstrate that if a village had a hydrocele rate greater than 20%, it was also ICT positive. However, below that point, and for 75% of villages surveyed, the hydrocele survey results did not appear useful for predicting ICT positivity. We were disappointed in the poor (albeit statistically significant) correlation of hydrocele and ICT village prevalence, and the finding that ICT prevalence was often very high in the face of very low hydrocele prevalence. The poor correlation was not consistent with observations by Gyapong and others,⁷ who found an excellent correlation between hydrocele and (nocturnal) microfilaremia prevalence in Ghana.

It is possible that the many years of mass treatment with ivermectin (for onchocerciasis) in these LGAs influenced the results obtained. This consideration would assume a disparate impact of ivermectin treatment on filarial hydrocele prevalence (an irreversible condition unchanged by ivermectin treatment) compared with ICT prevalence rates (possibly reduced by treatment). We do not believe this to be the case. First, if ivermectin treatment had the proposed effect, we would expect more ICT-negative villages. In fact, only 8% of villages had ICT negative samples, compared with 21% of villages with ICT rates greater than 40%. Second, Eberhard and others found that Og4C3 antigen levels (e.g., ICT positivity) do not decrease with ivermectin treatment in the same manner as does microfilaremia prevalence and intensity, perhaps due to the fact that ivermectin does not kill the adult W. bancrofti worms (and ICT detects circulating antigen from both microfilaria and adult worms).¹⁵ The impact of ivermectin treatment activities on mapping of lymphatic filariasis is an important one for Nigeria, since more than 16,500,000 treatments were provided in Nigeria in 2001 (Nigeria National Onchocerciasis Task Force, Federal Ministry of Health, unpublished data). Thus, future ICT surveys proposed for the lymphatic filariasis program in Nigeria, similar to the ones in this study, will often occur in areas already exposed to MDA. It would appear from our results that ICT results in lymphatic filariasis mapping activities will not be influenced by the ongoing onchocerciasis ivermectin treatment program.

This study sought to identify hydrocele, which was a clinical diagnosis based on the finding of a discrete, nontender, fluctuant mass around the testes that was not explained by indirect inguinal hernia or scrotal lymphedema. In our experience, nurses were as capable as physicians in not confusing these and other conditions (such as inguinal adenopathy, hanging groin, and acute epididymitis) with hydrocele. However, malignancy, varicocele, chylocele, spermatocele/cyst, or chronic epididymitis could have been recorded as hydrocele in our study, and some hydrocele could have been other than filarial in etiology (e.g., traumatic, tuberculous, idiopathic, etc.).

We recommend that if hydrocele surveys are used, they be based on actual physical examinations and not questionnaires, since we found that 32% of the men with physical finding of hydrocele did not report scrotal swelling when asked. In addition, questionnaire surveys for hydrocele would also fail to distinguish hydrocele from other easily differentiable causes of scrotal swelling. Physical examinations could be carried out by personnel trained to correctly distinguish hydrocele from other forms of scrotal pathology. Careful study is required to determine the best way to effectively provide such training.

Patients with clinical disease from lymphatic filariasis are often antigen or microfilaria negative. In a review of 23 published studies, Michael and others found similar rates of disease in both microfilaremic and amicrofilaremic persons.¹⁶ A report by Addiss and others found that 35% of 57 Haitian men with hydrocele were Og4C3 antigen negative.¹⁷ Consistent with these reports, we also found individuals with clinical hydrocele to be no more likely to have positive filarial antigen test results than men without hydrocele. However, this finding was initially quite confusing to the field staff, who expected that the ICT test results should be positive in hydrocele patients (e.g., as a diagnostic test for lymphatic filariasis). Staff confusion was abated by explaining to them that lymphatic filariasis often occurs after the body has mounted an effective immune reaction; that is, manifestations are residual effects of past filarial infection, e.g., burnt-out disease. Thus, persons with hydrocele and other manifestations of lymphatic filariasis may no longer have living worms, or circulating worm antigens, in their bodies to give a positive reaction on the ICT test. Nurses and physicians were satisfied by this explanation.

In conclusion, we found that only a very high village prevalence of hydrocele consistently identified the coexisting prevalence of lymphatic filariasis antigenemia (as measured by ICT positivity) in central Nigeria. We recommend further research to determine if and how hydrocele surveys can be used in lymphatic filariasis elimination programs as a supplement or alternative to ICT mapping for targeting mass treatment programs.