Search

Thirty-three cases of pediatric Lassa fever were identified at Curran Lutheran Hospital and Phebe Hospital in Liberia between January 1980 and March 1984. All 18 fetal cases died and the case-fatality rate for 15 childhood cases was 27%. We identified four clinical presentations according to age, including a case of congenital Lassa fever, a condition not reported previously. Two cases of Lassa fever were found serologically during a one-month survey of all pediatric admissions at Curran Lutheran Hospital, 2.4% of those children who had serum pairs collected. We also identified a "swollen baby syndrome" consisting of widespread edema, abdominal distention, and bleeding. This distinctive clinical presentation of Lassa fever ended in death in three of four cases and was present in three of the four childhood deaths in this series. Its absence seems to be a good prognostic indicator in children.

Patients admitted with fever to four Liberian hospitals were tested for Lassa fever (LF) by means of the indirect fluorescent antibody technique and by virus isolation. The incidence of LF and presumptive LF among consecutive febrile adult patients was 14% and 17% in two hospitals located in the interior; no cases of LF were found among 24 consecutive patients in a hospital near the coast. In the three inland hospitals the incidence of confirmed or presumptive LF among the patients in whom the diagnosis was seriously considered varied from 13% to 36%. Lassa virus was isolated from 17 patients out of the 59 cases found in this survey. LF is a common cause of fever in northern Liberia. The diagnosis depends upon the readiness of the staff to consider the diagnosis, the collection of blood specimens at appropriate times, and the preservation of sera at sub-freezing temperatures to permit survival of active virus and its subsequent recovery in an appropriate laboratory.

In a study to assess the epidemiological and clinical aspects of endemic Lassa fever (LF) in Liberia at Curran Lutheran Hospital (CLH), 44 cases were diagnosed by virological and serological techniques over a 22-month period. During one calendar month, testing of febrile patients admitted to the medical-surgical ward revealed six cases of LF, 13% of all febrile cases and 17% of those who were tested. As the study progressed the diagnostic skills of the hospital staff improved. The most common mistake was the diagnosis of a case of LF as pneumonia; the most potentially serious diagnostic problem was differentiating LF from typhoid fever, a readily treatable infection. LF may also mimic other diseases such as aseptic meningitis, pelvic inflammatory disease, gastroenteritis or arbovirus infection. We found a previously unreported symptom of LF, rib tenderness typical of costochondritis. The mortality rate in the medical-surgical ward was 5.4%; the over-all case-fatality rate was 13.6%. Women outnumbered men by nearly three to one, and had a higher mortality particularly noted in the pregnant. LF is common at CLH, and as many as 100 cases may occur annually at this hospital.

The efficacy of passive immunization for treatment of Lassa Fever (LF) is believed to depend on the titre of the neutralizing antibody infused. For the purpose of identifying optimal donors of LV-immune plasma, a population of LF-convalescent patients in Liberia was tested for prevalence of neutralizing antibody. Minimally protective titres, expressed as a log10 neutralization index, (LNI), were established in animal models as LNI greater than 2. LNI titres for 26 donors, tested eight or more months after illness, were modest: 16 titred 1 less than LNI less than 2, 4 titred 2 greater than LNI less than 3, and only 4 titred LNI greater than 3. Sequentially obtained plasma from six donors indicated that the LNI response was delayed relative to the indirect fluorescent antibody (IFA) response, that high titres (LNI greater than 3) occurred only after seven months and in only two of six patients. Most of the unselected LV-immune plasma will require concentration to therapeutically useful LNI titres. In a passive immunization experiment, guinea-pigs were protected by a late convalescent plasma (LNI = 4.8, IFA = 320) but not by an early plasma, (LNI = 0.6, IFA = 640), thus supporting the selection of immune plasma on the basis of the LNI. Cross serological testing with LV strains and convalescent plasma from patients in Sierra Leone, Liberia and Nigeria suggested that these LV strains were indistinguishable by cross-IFA, but were readily distinguishable by cross neutralization tests. Geographical matching of LV and plasma origins may thus be a factor in selection of optimal plasma for passive immunization of Lassa fever.

Two hundred thirteen cases of Lassa fever (LF) were diagnosed by virus isolation and seroconversion at Curran Lutheran Hospital in Zorzor, Liberia, between July 1980 and April 1986. An additional 40 cases of probable and presumptive LF were diagnosed on the basis of single serum samples. Of the 246 assessable patients, 23 (9%) died; no data were available for seven patients. Five (16%) of 32 pregnant women and three (43%) of seven immediately postpartum women died. Four (26%) of 15 children less than 12 years died. Case-fatality rates among 125 nonpregnant women and 67 men were approximately 6%. Among 150 patients studied in detail, the case-fatality rate was also 9%. Seventeen (11%) of these patients had abnormal bleeding; of these, six (35%) died. Most platelet counts were at low normal to mildly depressed levels. However, serial counts in seven patients suggested a decrease on about days 10-12 of illness. The symptoms of LF in Liberia are those of a viral syndrome. Edema, sometimes marked, is noted in seriously ill patients. A great variation in mortality and incidence of abnormal bleeding is recorded in reported series of LF; it appears that hemorrhage is a marker for cases with a high mortality. The incidence and severity of hearing defects in LF outbreaks vary. Elucidation of a number of clinical problems in LF requires more information on how strain differences affect the pattern of illness.

The sera of 844 Liberian hospital staff memebers were positive for Lassa Virus (LV) antibodies in a survey using the indirect fluorescent antibody technique (IFAT). In two hospitals in Lofa County near the Sierra Leone border, the prevalence, 15.4%, was significantly higher than the 8.4% in seven others. There were near differences between the prevalence among laboratory workers, 15.3%, and other workers, 7.7%, and between midwifery students, 21.2%, and midwives, 4.2%, suggesting their infection from patients or their blood products. However, the over-all prevalence among those with patient contacts was the same as that among those without direct patient contact; most LV infections were apparently acquired from sources other than patients in hospital. This finding, the lack of evidence of hospital outbreaks and the presence of comparable prevalences in all age groups suggest that LV infections occur on a continuing basis in this population. In one hospital the comparison of the results of IFAT and complement fixation tests revealed some who reacted by one technique and not by the other. In one person the titre by IFAT had dropped from 1:32 to undetectable levels in two years. This finding prompts caution in the interpretation of results.

Six villages in Lofa County, north-west Liberia, and one near the coast were surveyed for the presence of indirect fluorescent antibodies (IFA) to Lassa virus (LV). Prevalences were similar among males and females, and among various age groups. The prevalence of IFA positive sera, 6.4%, in two roadside villages was significantly higher than in two matched villages "in the bush", 1.9%. It was also higher in Gbanwei, a roadside village which did not maintain traditional sanitary measures, than in Zuwulo, similarly located but with maintenance of clean-swept areas without shrubbery or rubble between the houses. In another pair of villages, the one adjacent to a Mission Clinic with a very high prevalence of IFA positive staff members had significantly higher prevalence, 14.1%, than did the other, a roadside village with 5.1% seropositives. LV antibodies were also found in 4.3% of the inhabitants of a small coastal village near Robertsfield International Airport. Though LV infections are more common in villages in which traditional practices have been modified, they are present even in villages which are relatively unchanged. In the former they appear to be continuous while sporadic in the latter. The prevalences of IFA in the villages with the highest rates are about one third of what is found in personnel of hospitals near them, suggesting that hospital staff members acquire infections from patients as well as from the communities in which they live.

Sixty-three virus isolates were obtained by inoculation of Vero cells with sera from 50 hospital in-patients in Liberia with acute febrile illnesses. 57 of the isolates were presumptively identified as Lassa virus (LV) by direct fluorescent antibody (DFA) staining of inoculated Vero cells. These, and six additional isolates obtained only by titration of supernatant fluids from inoculated Vero cells, were definitively identified as LV in a neutralization test. Two additional LV isolates were obtained from a patient's sera from Nigeria. By cross-neutralization tests, the Nigerian LV strains were serologically identical to the prototype Nigerian LV strain (PP) but were distinct from both a reference LV strain from Sierra Leone (SL), and from the Liberian (LIB) strains isolated in this study. The LIB and SL strains were closely related to each other, but not to the Nigerian LV strains. LIB LV strains were tested for virulence in strain 2 and 13 guinea-pigs, and a spectrum of virulence was observed which correlated only approximately with disease severity for human patients. Two human-lethal isolates killed all inoculated strain 2 and 13 guinea-pigs, whereas nine isolates from mildly ill patients were benign for guinea-pigs. Yet some LV isolates from severely ill or lethally infected patients, especially those from pregnant women and infants, were totally benign for guinea-pigs. These data suggest that antigenically distinct LV strains exist in nature, and that antigenically indistinguishable LV isolates may differ in virulence potential for various hosts.

Serological testing of hospital personnel by the indirect fluorescent antibody (IFA) technique was used to indicate the distribution of Lassa virus (LV) activity in Liberia. Determination of the places of origin of the staff members as well as the sites of the hospitals indicated that LV is active in throughout Liberia. Prevalences of IFA varied from 3.8% at the J. J. Dossen Hospital on the coast in the south-east to 22.3, 23.5 and 40.4% in Lofa County hospitals inland in the north-west. Rises in LV antibody prevalences, high prevalences and relatively high IFA titres in hospital personnel suggest the LV activity is particularly high in Lofa, Grand Cape Mount and Nimba Counties.