In the 1980s, disseminated histoplasmosis due to Histoplasma capsulatum var. capsulatum was described in HIV-infected patients in the United States. Subsequently, it was increasingly reported in immunocompromised patients and was included to the list of opportunistic infection defining the AIDS [1,2].
Histoplasma capsulatum var. capsulatum infection seems to be a major AIDS-defining illness in endemic areas of South and Central America. It is the most frequent opportunistic infection due to HIV in French Guiana, along with tuberculosis, and the first cause of AIDS-related death [3,4].
Nevertheless, reports of HIV-positive H. capsulatum var. capsulatum histoplasmosis are scattered and published cohorts include a limited number of patients, rarely more than 100. Incidence in the greater Caribbean is probably under-reported due to the difficulty of the diagnosis . Moreover, its unspecific symptoms often mimic a Ziehl-negative tuberculosis, especially in low-income settings where invasive procedures and laboratory facilities are scarce .
In Europe, the prognosis of HIV-related H. capsulatum var. capsulatum infection is poor, with a mortality rate reported at 28.1% in the 30 days after starting clinical care. The limited knowledge about this disease in nonendemic settings delays diagnosis and is thus detrimental to its prognosis .
We carried out a descriptive analysis of 200 cases of HIV-related H. capsulatum var. capsulatum infection occurring over 25 years in French Guiana, focusing on epidemiological, clinical and diagnostic aspects. The aim of this work was to help optimize investigations, in particular in high prevalence low-income settings.
French Guiana is a crossroads for South American and Caribbean people migrating for economical or political reasons to this French overseas department. In a population between 30 and 39 years of age, the proportion of immigrants is 37% . The native countries of immigrants may have a high prevalence rate of HIV infection (like Haïti and Guyana) and some are endemic for histoplasmosis. There are 38 new AIDS cases per 100 000 inhabitants/year. With a prevalence greater than 1% in pregnant women, the epidemic is generalized and mature. Immigrants represent up to 80% of the HIV cohort.
The patients included were confirmed HIV-related H. capsulatum var. capsulatum infection diagnosed in the three hospitals of French Guiana, between January 1982 and April 2007. A confirmed case was defined by a positive HIV testing (ELISA and western blot) and an identification of H. capsulatum in tissues or fluids (either direct microscopic examination of May–Grunwald–Giemsa-stained smears or fungal culture or histopathological examination). Patients treated empirically before the reception of the positive result were included retrospectively, using the same inclusion criteria.
Some investigations were systematic (full blood count, creatininemia, CD4 cell count, liver function tests, C-reactive protein, chest radiograph), and others depended on clinical presentation [abdominal ultrasound, computed tomography (CT) scanner, tissue biopsy, endoscopy, lumbar puncture].
We excluded patients relapsing from histoplasmosis and children below the age of 15.
A standardized form was filled retrospectively for data collection, including socio-epidemiologic data, clinical symptoms, results of radiograph, CT scanner and biological results.
The samples collected were broncho-alveolar fluid or bronchial aspiration, bone marrow aspirations, surgical biopsies of adenopathies, liver biopsies, digestive biopsies obtained by endoscopy, skin or mucous biopsies, cerebro-spinal fluid (CSF). Blood smears and cultures were performed for malaria and bacteriological research, but not specifically for histoplasmosis due to technical limitations.
Data were analysed with STATA 8.2 software (Stata Corporation, College Station, Texas, USA). Normally distributed variables were compared using Student's t-test and nonnormally distributed variables using the Mann–Whitney test.
The mean age was younger for women than for men (37.5 vs. 41.4 years, P < 0.004). The sex ratio was 2: 1. Around 35% of the patients were from French Guiana, the same proportion from Haïti, and 22.1% were natives from Brazil, Guyana and Surinam. Patients coming from abroad had lived in French Guiana for 12.4 ± 8.7 years on average. HIV acquisition was through heterosexual sex in 95.2% of cases (Table 1).
Although 85.0% of the patients were included after 1996, when highly active antiretroviral therapy (HAART) became available, 92.0% (161/175) of patients were not receiving HAART at the time histoplasmosis was diagnosed. HAART had been started in 5.7% of the patients (10/175) for less than 6 months. Most cases (77.5%, 152/196) were presenting their first AIDS-defining event.
Almost 90% had had fever, for 30.3 days before the diagnosis was made (standard deviation: 24.9). Table 2 shows that the most usual symptoms were pulmonary and digestive: almost half of the patients complained of abdominal pain or diarrhoea, and the same proportion had pulmonary signs, mainly cough and interstitial syndrome. 13.9% of patients had radiographical abnormalities without pulmonary symptoms. A few had either digestive haemorrhage or tumoural mass or occlusion or subocclusion or ascites or all.
Almost half of the patients presented superficial lymphadenopathies on examination. Patients with deep adenopathies were significantly more likely to present superficial adenopathies [odds ratio (OR) = 1.8, 95% confidence interval (CI): 1.03–3.25, P < 0.007]. Adenopathies were only cervical in 53.0% of the cases and their size was greater than 2 cm in 25.9% of the cases. Inguinal lymph nodes were not considered because their pathological meaning was doubtful.
Neurological symptoms were less common; they were mainly either cognitive deficiency or confusion or both. Skin or mucosal locations were also rare. Lesions affected mainly the face; they were either papules or nodules most of the time.
Cases which occurred within the 6 months following introduction of HAART had mostly fever (9/10), lymphadenopathies (5/10) and hepatomegaly (5/10). Pulmonary (4/10) and neurological symptoms (3/10), cutaneous lesions, digestive symptoms and splenomegaly were less common (2/10 for each).
Other opportunistic infections were associated with histoplasmosis in 36.6% of cases; these were mostly tuberculosis (16 cases, 8.0%), oesophageal candidiasis (15 cases), chronic herpes simplex infection (12 cases), toxoplasmosis (eight cases) and pneumocystosis (seven cases).
Table 3 shows that the mean CD4 cell count was low (63.5 cells/μl) and that more than 80% of patients had a CD4 cell count of 100 cells/μl or less. Nevertheless, a small proportion had a CD4 count greater than 200 cells/μl. Most patients had moderate anaemia with haemoglobin between 8.1 and 11.5 g/dl. Around 40% had thrombocytopenia and around 30% had neutropenia. Liver function tests were abnormal in most patients. Most concerned was the aspartate aminotransaminase concentration (AST), which was increased in 71.9% of cases, with a mean value two to three times the upper limit of the normal range (ULN). Alanine aminotransaminase concentration (ALT) was normal in most cases. Alkaline phosphatase levels (ALP) and ferritinemia were elevated in 61.8 and 81.8% of cases, respectively. Ferritinemia was at least 1000 ng/ml in 66.7% of patients. C-reactive protein was elevated in 95.1% of the patients but less than 100 mg/l in most cases. Mean lactic dehydrogenase (LDH) level was approximately 1000 UI/l.
Severity factors for early death were present in 16.9–63.8% of cases (32 patients with creatininemia at least 120 μmol/l and 92 with LDH at least twice that of ULN ).
Bone marrow aspiration was the most frequently performed investigation (55.5% of the patients included) and led to the highest number of diagnoses. The result was positive in 76.6% of cases, mostly for culture (75.8% for culture and 45.3% for direct microscopic examination, P < 0.001).
Lymph node tissue and liver biopsies were the most contributive with, respectively, 89.6 and 78.6% of positive results. Both were more frequently positive for culture (80.6% for lymph node and 89.2% for liver) than for direct microscopic examination (respectively 41.7%, P = 0.001 and 27.5%, P < 0.001) or histopathology (54.8%, P = 0.016 and 44.9%, P < 0.001). Both were performed in less than 30% of the patients.
Skin/mucosal biopsy had the highest proportion of positivity for direct examination with 78.6% of the results (and around the same rate for histopathology). It was performed in 15.6% of the cases. The ratio of patients with positive skin/mucosal biopsy was 42.8% before 1997 and 3.8% after 1997 (P < 0.001) when a mycologist joined the laboratory. Three quarters of the positive skin biopsies were found before 1997.
Broncho-alveolar fluid analysis was the second most frequent investigation performed after bone marrow aspiration. The proportion of positive results was 55.6%. Digestive samples concerned 17.7% of patients and were mostly positive on culture from colic biopsies. Although direct examination of blood samples and specific cultures were not performed to search for histoplasmosis, yeasts evoking H. capsulatum were found nine times on blood smears for malaria and 16 times on cultures (blood cultured in search for bacteria and kept less than 1 week). CSF analysis was positive only three times, in all cases on culture.
Overall, the proportion of positive results was higher for cultures (196/317, 61.8%) than for direct microscopic examination (150/355, 42.3%, P < 0.001). After 1997, when fungal culture techniques improved, the proportion of positive cases increased to 45.8% for broncho-alveolar fluids (22/48), 86.8% for bone marrow (66/76), 89.3% for lymph nodes (25/28) and 91.7% for liver biopsies (33/36).
Outcomes and treatment
Early death within the month following the beginning of treatment occurred in 17.5% of cases (31/177). At least 30.7% were dead 6 months after starting treatment (58/189). The diagnosis of histoplasmosis was postmortem for eight patients (one treated unsuccessfully, seven never started specific treatment).
Before the diagnosis was confirmed, 14.3% of the patients (15/105 for which data were available) started the treatment. This empirical treatment was implemented 29.6 days before confirmation was made (standard deviation: 45.0). The outcome at 1 month was available in 14 cases: five of 14 had died (35.7%).
Treatments were: itraconazole, alone (60.3%) or after deoxycholate/liposomal amphotericin B induction (34.9%), fluconazole (1.0%). Some patients started itraconazole first and were switched to liposomal amphotericin B because of persistence of symptoms (3.7%). The mean duration of amphotericin-based treatment was 9.0 days, with a standard deviation of 4.3.
HIV-related H. capsulatum var. capsulatum infection has mainly been described in the United States, but data from southern countries are scattered. To our knowledge, this is the largest series studied to this date; it may give an overview of this disease in South American settings.
Overall, HIV-positive patients presenting histoplasmosis in French Guiana are poorly followed, usually not treated and sometimes even not tested for HIV before presentation. Thus, more than 90% of patients were not receiving HAART at the time of diagnosis (despite the availability of HAART after 1996). For others, as described elsewhere , HIV-related H. capsulatum var. capsulatum infection may become patent during the months following introduction of HAART (10 patients in our series started HAART less than 6 months before diagnosis, whereas only four started HAART more than 6 months before).
Adenopathies were found in 55.3% of our patients whereas they were only present in 19.2–27.8% of cases elsewhere [7,10,11]. We presume that only large adenopathies may be reported in those cases. In the immune restoration period, some patients developed large lymphadenopathies with fever and hepatomegaly, which led to the diagnosis of HIV-related H. capsulatum var. capsulatum infection; these conspicuous symptoms may appear within a few days.
Apart from skin/mucosal involvement, most clinical symptoms are unspecific and may be seen in tuberculosis. The proportion of skin involvement is close to what was described in Panama , but higher than in North American series (1–7%) , and lower than European or Brazilian series (approximately 47%) [7,11]. Genetic differences in the strains of H. capsulatum have been suggested to explain these differences . However, this may also result from the improvement of diagnosis methods . We presume that later stages of HIV-related H. capsulatum var. capsulatum infection are diagnosed because antigen detection techniques are not accessible. This may partly explain the differences in clinical features and outcomes observed in different series.
The biological abnormalities reported in our series are unspecific. However, a high LDH level is a useful clue . LDH concentration also has a prognostic value and was associated with early death if ULN is twice or more . LDH concentrations were notably increased in our study (approximately 1000 IU/l) but a normal value did not exclude the diagnosis. Other clues were an increased AST concentration, combined with an increased alkaline phosphatase concentration and eventually thrombocytopenia. The two former represent severity criteria when greater than 2.5 times of ULN . When available, ferritinemia was frequently 1000 ng/ml or more. Proteinuria and haematuria have been associated with HIV-related H. capsulatum var. capsulatum infection, relative to other opportunistic infections .
A main limitation in the study of the performance of different diagnostic methods was the absence of a gold standard. Lyse-centrifugation blood culture and antigen detection techniques are not performed because of technical and financial constraints, and PCR is being validated. This study was retrospective and all patients were not screened identically; Table 4 reflects the practice of French Guiana physicians. Data published on diagnostic methods are scattered and results are often not detailed (see Table 5; some reports mix AIDS and non-AIDS cases). Despite these limitations, the French Guyanese setting may be a good reflection of the South American and Caribbean situations which may have a limited access to laboratory facilities, and no access to the antigen detection techniques.
As suggested by others , bone marrow analysis was a helpful diagnostic tool, especially for fungal culture, which was more contributive than direct examination (75.8 vs. 45.3%). It was performed in the majority of cases and gave the highest number of diagnoses. Although its sensitivity may vary with culture methods (Table 5), we would recommend this procedure because it is easily performed at the bedside and usually free of complications.
Although lymph nodes were not always biopsied, the cultures of lymph nodes and liver samples were contributive investigations that could also screen for differential diagnoses. This surgical procedure may be avoided if a simpler investigation like skin or bone marrow analysis gives a quick diagnosis.
Criteria associated with death within the first 30 days of treatment are dyspnoea, thrombocytes less than 100 000 cells/μl, LDH level greater than twice of ULN . These criteria were present in a majority of patients. Early mortality was 17.5%, which may be lower than in nonendemic setting (it was reported at 28.1% in Europe ).
For severe cases, liposomal amphotericin B (3.0 mg/kg) is recommended for 1–2 weeks, followed by itraconazole . In our series, duration of liposomal amphotericin B was based on the clinician's appreciation of the patient's status. In some cases, treatment was switched after 3–4 days only, which was sufficient to reach apyrexia.
To our knowledge, empirical treatment has never been described in the literature. Nevertheless, considering that H. capsulatum may need several weeks to grow in culture, our experience is that empirical treatment is a life saving strategy. The rationale seems even stronger in endemic settings where fungal culture is not available. In severe cases, amphotericin B is a good choice because improvement is generally observed within a few days [17,18]. It is better than itraconazole which has a slower effect (1–2 weeks). Liposomal amphotericin should be preferred when available [17–19]. If there is no clinical improvement, antituberculosis treatment may be started quickly after a course of 3–7 days of amphotericin B-containing regimen.
We warmly acknowledge the contributions of Dr Pradinaud, Dr Sainte-Marie, Dr Fior, Dr Mottard, Dr Bourbiguot, Dr Louvel, Dr Randrianjohany, Dr Alvarez, Christiane Clairvoyant, Hélène Lucas, the former Senior House Officers, the interns of the Dermatology Unit of Cayenne's Hospital, and the patients included.
F.H. contributed by providing original idea, design of the study, data collection, statistical analysis and writing. Statistical analysis, manuscript correction and validation were done by M.N., C.A. and B.C. helped in biological diagnosis and validation. M.P.-D., M.E.G., T.V., V.V., A.M., C.M., E.C. contributed by data collection and validation. P.C. contributed in setting-up the database, data collection, manuscript correction and validation.
2. Centers for Disease Control and Prevention (CDC). Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults; 1993. http://www.who.int/hiv/pub/guidelines/hivstaging/en/
. [Accessed 23 October 2007]
3. Couppié P, Sobesky M, Aznar C, Bichat S, Clity E, Bissuel F, et al
and acquired immunodeficiency syndrome: a study of pronostic factors. Clin Infect Dis 2004; 38:134–138.
4. Lewden C, Sobeski M, Cabie A, Couppié P, Boulard F, Bissuel F, et al
. Causes of death among HIV
-infected adults in French Guiana
and the West Indies in the area of highly active antiretroviral therapy (HAART). Med Mal Infect 2004; 34:286–292.
5. Nacher M, Aznar C, Blanchet D, Demar M, El Quedj M, Vaz T, et al
. AIDS-related disseminated histoplasmosis
in the greater Caribbean: how frequent is it? AIDS 2006; 20:951–952.
6. Couppié P, Aznar C, Carme B, Nacher M. American histoplasmosis
in developing countries with a special focus on patients with HIV
, treatment and prognosis. Curr Opin Infect Dis 2006; 19:443–449.
7. Spinello A, Magni C, Nebuloni M, Parravicini C, Corbellino M, Sollima S, et al
among human immunodeficiency virus-infected people in Europe. Report of 4 cases and review of the literature. Medicine 2006; 85:22–36.
9. Nacher M, Sarazin F, El Quedj M, Vaz T, Alvarez F, Nasser V, et al
. Increased incidence of disseminated histoplasmosis
following highly antiretroviral therapy initiation. J Acquir Immune Defic Syndr 2006; 41:468–470.
10. Gutierrez ME, Canton A, Sosa N, Puga E, Talavera L. Disseminated histoplasmosis
in patients with AIDS in Panama: a review of 104 cases. Clin Infect Dis 2005; 40:1199–1202.
11. Severo LC, Oliveira F, Irion K, Porto N, Londero AT. Histoplasmosis
in Rio Grande do Sul, Brasil: a 21 year experience. Rev Inst Med Trop Sao Paulo 2001; 43:183–187.
12. Karimi K, Wheat J, Connoly P, Cloud G, Hajjeh R, Wheat E, et al
. Differences in histoplasmosis
in patients with acquired immunodeficiency syndrome in the United States and Brazil. J Infect Dis 2002; 186:1655–1660.
13. Couppié P, Clity E, Nacher M, Aznar C, Sainte-Marie D, Carme B, et al
. AIDS-related oral and/or cutaneous histoplasmosis
: a descriptive and comparative study of 21 cases in French Guiana
. Int J Dermatol 2002; 42:571–576.
14. Corcoran GR, Al-Abdely H, Flanders CD, Geimer J, Patterson TF. Markely elevated serum lactate deshydrogenase levels are a clue to the diagnosis
of disseminated histoplasmosis
in patients with AIDS. Clin Infect Dis 1997; 24:942–944.
15. Wheat JL, Chetchotisakd P. Factors associated with severe manifestations of histoplasmosis
in AIDS. Clin Infect Dis 2000; 30:877–881.
16. Daher EF, Silva GB Jr, Barros FA, Takeda CF, Mota RM, Ferreira MT, et al
. Clinical and laboratory features of disseminated histoplasmosis
patients from Brazil. Trop Med Int Health 2007; 12:1108–1115.
17. Wheat LJ, Freifeld AG, Kleiman MB, Baddley JW, McKinsey DS, Lyod JE, et al
. Clinical practice guidelines for the management of patients with histoplasmosis
: 2007 update by the Infectious Diseases Society of America. Clin Infect Dis 2007; 45:807–825.
18. Couppié P, Aznar C, Carme B, Nacher M. American histoplasmosis
in developing countries with a special focus on patients with HIV
, treatment, and prognosis. Curr Opin Infect Dis 2006; 19:443–449.
19. Johnson PC, Wheat LJ, Cloud GA, Goldman M, Lancaster D, Bamberger DM, et al
. Safety and efficacy of liposomal amphotericin B compared with conventional amphotericin B for induction therapy of histoplasmosis
in patients with AIDS. Ann Intern Med 2002; 137:105–109.
20. Wheat LJ, Connoly-Stringfield PA, Baker RL, Curfman MF, Eads ME, Israel KS, et al
. Disseminated histoplasmosis
in the acquired immune deficiency syndrome: clinical findings, diagnosis
and treatment, and review of the literature. Medicine (Baltimore) 1990; 69:361–374.
21. Hajjeh RA, Pappas PG, Henderson H, Lancaster D, Bamberger DM, Skahan KJ, et al
. Multicenter case-control study of risk factors for histoplasmosis
in human immunodeficiency virus infected persons. Clin Infect Dis 2001; 32:1215–1220.
22. Leimann BC, Pizzini CV, Muniz MM, Albuquerque PC, Monteiro PCF, Reis RS, et al
in a Brazilian center: clinical forms and laboratory tests. Rev Iberoam Micol 2005; 22:141–146.
Keywords:© 2008 Lippincott Williams & Wilkins, Inc.
diagnosis; French Guiana; histoplasmosis; HIV; presumptive treatment