Outbreak of Hepatitis E Virus Infection in Darfur, Sudan: Effectiveness of Real-Time Reverse Transcription-PCR Analysis of Dried Blood Spots

Hepatitis E virus (HEV) is a spherical, nonenveloped, single-stranded RNA virus (2, 16) that belongs to the new genus, Hepevirus (14). This pathogen is responsible for at least 50% of acute non-A non-B hepatitis in developing countries, where it causes sporadic infection but also large epidemics usually associated with fecal contamination of water (4-6, 10, 11, 15). A large outbreak of hepatitis E was reported in June 2004 in the internally displaced population camps of Darfur, in West Sudan, and across the border in Chad: at least 5,000 HEV infections were recorded from June to December 2004 (9). A task force, set up by the nongovernmental organizations Médecins Sans Frontières (Doctors Without Borders) and Epicentre in refugee camps in Darfur and by the World Health Organization and Centers for Disease Control in refugee camps in Chad, conducted a field study of this large epidemic. The samples were drawn from both clinical patients and asymptomatic individuals (3, 9, 13). The French National Reference Laboratory (CNR) for hepatitis E monitored the biological and virological investigations. The aims of this study were to assess the feasibility of amplifying HEV RNA from dried blood spots (DBS) in comparison with amplification from serum and to describe the biological profiles of patients and asymptomatic individuals.

Eighty-nine displaced persons living in the Chad Goz Amer camp (median age, 28 years; range, 7 to 65 years) and 92 persons in the Sudanese Mornay camp (median age, 25 years; range, 3 to 75 years) were investigated in August and September 2004. Two groups were defined in each camp. The first was composed of patients considered to be HEV infected, i.e., patients presenting or having presented jaundice since 1 July 2004 and with negative test results for a malarial diagnosis (n = 36 in Goz Amer; n = 56 in Mornay). The time before onset of jaundice was reported by the physicians of medical teams after investigation of patient histories. The second group included individuals that were asymptomatic from the start of the outbreak (n = 53 in Goz Amer; n = 36 in Mornay). Two types of biological samples were collected for each individual. First, whole-blood samples were collected by venipuncture and were centrifuged locally to isolate sera. Aliquots were conserved between +4 and +8°C in the field and at −80°C at CNR. Second, finger-prick whole-blood samples were collected on filter paper. The filter papers were thoroughly air dried and were stored at ambient local temperatures (28 to 40°C) in individual paper bags to prevent contamination and preserve long-term stability. As various medical teams were involved, two types of filter paper were used: Whatman paper in Goz Amer and Isocode Stix in Mornay (Schleicher and Schuell Bioscience, Inc.).

Samples were shipped to France, and serum samples were tested for anti-HEV immunoglobulin G (IgG) and IgM with enzyme-linked immunosorbent assay HEV IgG and HEV IgM kits (Genelabs Diagnostics and Abbott). HEV RNA was amplified from DBS and serum samples from each refugee. Paper sample areas were eluted in 220 μl of phosphate-buffered saline-Tween buffer. A final elution or serum volume of 200 μl was used for nucleic acid extraction with a MagnA pure RNA isolation kit (Roche Diagnostics). After a retrotranscription step, HEV RNA was amplified and detected by consensus TaqMan real-time reverse transcription-PCR (RT-PCR) performed on a LightCycler as described previously (8). Samples with a crossing point inferior to 43, without amplification of the negative control, were considered positive (8). Samples with discordant results for serum and DBS were tested in duplicate in two independent real-time PCR assays (retrotranscription and PCR). HEV RNA amplification results for sera and DBS were concordant in 90.6% of cases (Table 1), with no statistically significant differences (McNemar chi-square test). The kappa correlation coefficient was calculated as 0.808. The results of amplification from DBS and sera were in agreement at 86.5% for Whatman paper and 94.6% for Isocode Stix. No significant difference between the two papers was observed in this study (P = 0.1). Thus, DBS are easy to collect and store and have been successfully used for the amplification of HEV RNA by real-time RT-PCR. However, in nine cases, HEV RNA was amplified in serum samples but not in DBS samples. Six of the nine samples were stored on Whatman paper, and the amount of blood on the filter was less than 50 μl. Three of the nine samples were collected on Isocode Stix in suitable volumes, and discrepancies may be explained by persistence of natural PCR inhibitors in the paper matrix or by RNases introduced by finger contact when handling the papers during collection. HEV amplification sensitivity is dependent on standardizing the quantity of blood and on the careful handling of filter papers with gloves, especially for this RNA virus. In eight cases, DBS samples were positive and serum samples were negative for HEV RNA amplification; these cases occurred in samples from patients with clinical and serological evidence of acute hepatitis (n = 3) and in samples from asymptomatic refugees in which anti-HEV IgM was detected in the serum (n = 4). These cases, which provided the best results for HEV amplification from DBS samples, may be explained by the degradation of RNA in sera due to an absence of cooler conditions during transport (less than −20°C).

The second objective of this study was to describe the various biological profiles identified. Among the group defined as hepatitis E cases, all the specimens were positive for at least one of the HEV markers HEV RNA, anti-HEV IgG, and anti-HEV IgM (Table 2). HEV RNA was detected in 65.1% of cases; among these, 51 of 58 had detectable anti-HEV IgM. The Abbott IgM anti-HEV assay had a sensitivity of 86% in comparison with the results of RT-PCR. These results are consistent with data reported under outbreak conditions (7, 12). Interestingly, HEV RNA amplification was positive in 7 of 12 (58.3%) DBS specimens with negative anti-HEV IgM results, allowing the diagnosis of HEV infection. Conversely, among the cases in which HEV RNA amplification was negative (34.9%), 26 of 31 (83.9%) had detectable anti-HEV IgM. These observations underline the importance of using different biological tests for the HEV diagnosis, as none of the tests are sufficiently sensitive for use alone. Interestingly, HEV viremia was detectable with TaqMan HEV RNA real-time RT-PCR for more than 39 days after the onset of jaundice. Another point to underline is the large proportion of asymptomatic individuals displaying a biological profile consistent with acute hepatitis E infection (43.5%), with a detectable viremia for 31.5% of cases (Table 3). No statistically significant differences were observed when comparing the ages of individuals with icteric infections and those with asymptomatic HEV infections; however, clinical observations have documented the frequency of symptoms among older populations (1). An accurate evaluation of the spread of disease and the infection attack rate also requires biological samples from a cross section of asymptomatic individuals.

DBS have already been used for PCR detection of hepatotropic viruses, but this study is the first report using real-time nucleic acid amplification of DBS collected under tropical conditions in refugee camps during an outbreak. DBS requires only 50 μl of blood, which is an advantage for children. DBS samples do not require a centrifuge and can be mailed unrefrigerated with a low biohazard risk. Thus, RT-PCR of DBS samples may provide an accurate and reliable tool for conducting field studies of outbreaks in developing countries.

• American Society for Microbiology