The Relationship Between Strongyloidiasis and Human Retroviral Infections—So What is it?

Abstracts & Commentary

Synopsis: Two recently published papers indicate potential relationships between severe forms of strongyloidiasis and retroviral infections such as HIV and HTLV-1. However, unless coinfections with more than one retrovirus are excluded, the literature may soon become filled with conflicting case reports and series as illustrated by the simultaneous publication of the following two papers.

Sources: Ferreira MS, et al. Strongyloidiasis and infections due to human immunodeficiency virus: 25 cases at a Brazilian teaching hospital, including seven cases of hyperinfection syndrome. Clin Infect Dis 1999;28:154-155; Gotuzzo E, et al. Strongyloides stercoralis hyperinfection associated with human T cell lymphotropic virus type-1 infection in Peru. Am J Trop Med Hyg 1999;60:146-149.

What is the relationship between human retroviral infections and strongyloidiasis, if there actually is any such relationship to be concerned with? The expectation had been that strongyloidiasis would emerge as an important opportunistic nematode infection during the AIDS pandemic. Well, has it? Can one demonstrate the association by simply looking at concurrent HIV-1 infections and strongyloidiasis? It’s apparently not that simple. Perhaps we need to be looking more closely at regions of the world in which other retroviruses, such as HTLV-1 and strongyloidiasis, are endemic. Ferreira and colleagues note that relatively few cases of disseminated strongyloidiasis have been recognized in patients with AIDS, even when the literature from tropical countries is examined. They report 25 cases of strongyloidiasis in a population of HIV-infected patients, including seven with hyperinfection observed over a 77-month interval at a Brazilian teaching hospital. In their population of 650 adults and adolescents, with at least two positive ELISA tests demonstrating HIV seropositivity, larvae of S. stercoralis were found in the stools of 23, sputum of two, and/or on postmortem examination in five patients, for a total of 25 cases (3.85%). Most patients were male (80%), and fever, diarrhea, or cough were commonly observed, although not necessarily related to strongyloidiasis since other opportunistic infections were also present. Larva currens were observed in four individuals, and only four patients had elevated eosinophil counts of 400/µl.

Seven cases, all of whom died, were diagnosed as strongyloidiasis with hyperinfection, based upon either postmortem findings or detection of larvae in sputum of two patients. Although postmortem examinations did show extensive involvement of the gastrointestinal tract and lungs, larvae were not found outside their normal migration pattern. What contribution the presence of strongyloidiasis made to their demise was unclear to Ferreira et al. Results of blood cultures, if any, were not reported.

Albeit not disseminated strongyloidiasis, Ferreira et al attributed the presence of extensive infection, higher parasitic burdens, and increased migration of filariform larvae from the gastrointestinal tract into the lungs, to depressed immunity secondary to HIV infection. No data were reported in this study regarding the serological status of these patients for HTLV-1 infection, and the prevalence of strongyloidiasis at this teaching hospital, as it appears in a group of HIV-seronegative patients, was not reported.

Gotuzzo and associates conducted a study in Lima, Peru, at the Cayetano Heredia Hospital where they investigated a group of patients who presented with strongyloides hyperinfection defined as a systemic illness with chronic diarrhea, abdominal pain, weight loss, cough, edema, hypoproteinemia, and anemia with two or more organs involved (usually lung, intestines, liver, and the central nervous system) and stools demonstrating Strongyloides larvae, along with at least one sputum. In this case-control study, the seropositivity rates for HTLV-1 in patients with hyperinfection (18/21, 85.7%) were significantly higher than those obtained in two control groups. These included healthy age- and sex-matched controls (1/21 seropositive, 4.7%) and patients with intestinal strongyloides infection (6/52 seropositive, 9.7%). As in the first study by Ferreira et al, more than 90% of their reported cases of strongyloides hyperinfection did not demonstrate eosinophilia. More importantly, Gotuzzo et al specifically state in their discussion that none of their patients had concurrent HIV infection.

Comment by Frank J. Bia, MD, MPH

It is probably well known to our readers that S. stercoralis filariform larvae can autoinfect their human hosts and lead to chronic infections with an ability to both hyperinfect and disseminate within immunosuppressed patients. Transplant recipients, those receiving corticosteroid therapy, or those with underlying immunosuppressive states, such as severe malnutrition, are also at risk. As mentioned earlier, it had been expected that in the AIDS era strongyloidiasis would emerge as a major opportunistic infection, in both its hyperinfective and disseminated forms. The latter scenario would potentially carry enteric bacteria out of the gastrointestinal tract, threatening yet another serious opportunistic parasitemia and bacteremia.

This has not been the case. In fact, any association between more severe forms of strongyloidiasis and retroviruses appears to be with a different retrovirus, namely HTLV-1. It is not known why this occurs, but there are several problems in the report by Ferreira et al from Brazil. They have no control group of HIV-uninfected patients and do not report the prevalence of various forms of strongyloidiasis in such a group for their region of Brazil. It may conceivably be no different from their study group of HIV-infected patients. More importantly, they do not indicate the serological status of their patients with strongyloidiasis with regards to HTLV-1 infection.

Why is this important? The patients reported in the first article by Ferreira et al were largely male drug users and a third were either homosexual or bisexual. As early as 1989, Cortes and colleagues had reported seroepidemiological data on coinfections with HIV-1, HIV-2, and HTLV-1 on 704 patients in Brazil. They reported the prevalence of HTLV-1 infections ranged from 1% in rural female prostitutes to 13% in HIV-1 seropositive men with bleeding disorders in Rio de Janeiro, Brazil.1 In fact, combined HIV-1 and HTLV-1 infections occurred in between 1% to 11% of some male groups when studied by them more than 10 years ago. By 1995, Domingues et al had reported 45 cases of HTLV-1 associated myelopathy/tropical spastic paraparesis in Sao Paulo, Brazil.2 However, as recently as 1996, Broutet and colleagues reported the prevalence of HTLV-1 antibodies in more than 2700 persons tested during 1993-1994 in the northeastern city of Fortaleza, Brazil. The prevalence ranged from 0.12% in pregnant women to 1.21% in female commercial sex workers.3 It is not possible to show an association between any form of strongyloidiasis and HIV-1 or HIV-2 infection without also providing information about the serological status of such patients with regards to HTLV-1 infection. In contrast, the discussion provided by Gotuzzo et al indicates that none of their cases of S. stercoralis hyperinfection associated with HTLV-1 seropositivity were coinfected with HIV. That is important information since the potential for coinfection exists in association with high-risk sexual behavior, Japanese origin, and Quechua ethnicity—all known risk factors for HTLV-1 infection in Peru, as pointed out by Gotuzzo et al. Clearly, local social conditions, ethnicity, and behavioral risk factors influence the potential for coinfections. However, if we are to make any sense out of purported associations between potentially opportunistic parasitic infections and retroviral infections, studies must rigorously define the serological status of their study groups with regard to known retroviruses.


    1. Cortes E, et al. HIV-1, HIV-2, and HTLV-1 infection in high-risk groups in Brazil. N Engl J Med 1989;321: 830-832.

    2. Domingues RB, et al. Human T lymphotropic virus type 1-associated myelopathy/tropical spastic paraparesis in Sao Paulo, Brazil. Clin Infect Dis 1995;20: 1540-1542.

    3. Broutet N, et al. Prevalence of HIV-1, HIV-2 and HTLV antibody in Fortaleza, Ceara, Brazil 1993-1994. Int J STD AIDS 1996;7:365-369.