Clinical and epidemiological characterization of histoplasmosis cases in a nonendemic area, Connecticut, United States

Abstract

We performed a retrospective analysis of histoplasmosis cases diagnosed at our institution in New Haven, Connecticut, from 2005 to 2015. Among 12 cases of active histoplasmosis, seven were immunosuppressed and five had human immunodeficiency virus (HIV). Eleven patients reported travel to potentially endemic areas at a median of 105 days prior to presentation; travel to the Caribbean was most common (n = 6). Median time to diagnosis from symptom onset and first Histoplasma antigen testing were 41 and 28 days, respectively. Consistent with reports from other non-endemic areas, our findings suggest that the epidemiology of histoplasmosis may differ in Connecticut, potentially contributing to delayed diagnoses.

histoplasmosis, nonendemic, epidemiology, delay, travel

Histoplasma capsulatum, the etiologic agent of histoplasmosis, is highly endemic to the Mississippi and Ohio River valleys of North America. The geographical extent of histoplasmosis, however, may be farther reaching than the regions currently understood to be endemic. In an era of mass travel and migration, even transient passage into endemic areas can lead to disease acquisition with the possibility of reactivation many years later, in nonendemic regions. As an increasing number of people receive lifesaving but immunosuppressive (IS) therapies, the population at risk for histoplasmosis continues to grow. Moreover, several reports of histoplasmosis outside of high prevalence areas, in persons who have never traveled, suggest that pockets of previously unrecognized endemicity may exist both within¹ and outside of North America.^2,3 Connecticut has not been previously thought to be endemic for histoplasmosis. Between 1958 and 1965, less than 1% of navy recruits from New Haven, Connecticut, tested positively with the histoplasmin skin test.⁴ In this report, we examine clinical and epidemiological characteristics of histoplasmosis cases diagnosed in Connecticut, United States.

Between July 2014 and June 2015, we diagnosed three cases of active histoplasmosis while on clinical duties at the Yale New Haven Hospital in New Haven, Connecticut, United States. To better characterize the clinical and epidemiologic features of histoplasmosis in the nonendemic area of Connecticut, we sought to review cases at our institution spanning the past decade. We performed a retrospective chart review of microbiologically and histopathologically confirmed cases of histoplasmosis diagnosed at the Yale New Haven Hospital (York and St. Raphael campuses) in New Haven, Connecticut, from 2005 to 2015. Medical records were reviewed for data on year of diagnosis, demographics, comorbidities, exposure and travel history, clinical presentation, diagnostics, Histoplasma antigen levels, treatment, and outcomes. In accordance with the European Organization for Research and Treatment of Cancer and the Mycosis Study Group (EORTC-MSG) criteria for a proven, endemic fungal infection, detection of H. capsulatum by histopathology or culture was required for a definite diagnosis of histoplasmosis. Active histoplasmosis was defined as the presence of clinical symptoms in the setting of proven fungal disease. Treatment failure was defined as clinical, radiographic, or microbiologic evidence of disease progression while on adequate therapy, leading to death, clinical worsening, or relapse. The search strategy included: (1) pathology records for histopathologic evidence of histoplasmosis, (2) clinician-referred cases, (3) microbiology records for H. capsulatum-positive cultures, and (4) electronic medical records for relevant ICD-9 codes. After elimination of duplicates, a total of 20 cases were identified including 12 cases of active histoplasmosis. In eight patients, an incidental diagnosis was made after biopsy of asymptomatic pulmonary nodules, most commonly as part of malignancy screening (n = 4). The diagnosis was made by histopathology in all eight with only one culture positive specimen. Cases were considered ‘incidental’ because patients were asymptomatic. None received antifungals, and there was no evidence of active disease on follow-up.

Among 12 patients with active histoplasmosis, the median age was 54 years, and 10 were male. There were six Hispanics, four whites, one black, and one Asian. Seven patients (58%) had underlying IS conditions including human immunodeficiency virus (HIV) (n = 5; four had CD4 < 100 cells/μl), bone marrow transplantation (n = 1), combined kidney and liver transplantation with post-transplant lymphoproliferative disorder (n = 1), and systemic lupus erythematosus (n = 1).

Exposure history was elicited in four patients (33%) including previous landscaping or soil exposure in three and waterway exposure in one. Eleven patients reported travel to or prior residency in potentially endemic areas. Travel to the Caribbean was most common (Puerto Rico in five, Guatemala in one) followed by the Midwest (n = 3) and East Coast (North Carolina; n = 2). Other travel included to the Southwestern USA (n = 2) and Southeast Asia (n = 1). The median interval time from travel to diagnosis was 105 days (6 days to 50 years). In terms of clinical characteristics, constitutional symptoms were present in six (50%) and respiratory symptoms in three (25%). Notable physical exam findings included oral lesions in three (25%) and skin lesions in two (17%). Significant laboratory abnormalities were elevation of transaminases (n = 7) and anemia (n = 6). Histoplasmosis was limited to the lungs in four patients and disseminated in eight patients (see Table 1). The most common extra-pulmonary sites were the oral mucosa (n = 3) and bone marrow (n = 3). Other sites included the blood (n = 2), spleen (n = 2), liver (n = 2), skin (n = 2), adrenal glands (n = 1), bone (n = 1), central nervous system (n = 1), and colon (n = 1). Five patients were admitted to the intensive care unit (ICU), two of whom developed acute respiratory distress syndrome. Pathology consistent with H. capsulatum and positive cultures for H. capsulatum were observed in 11 and 5 cases, respectively. The most common source of a definitive diagnosis was a lung or mediastinal tissue biopsy (n = 4). Histoplasma antigen was positive in 6 of 11 patients tested (Table 2).

Table 1.

Baseline and clinical characteristics for patients with active histoplasmosis.

Characteristic	N = 12
Age (median ± SD) (years)	54 ± 16
Sex
Male	10
Female	2
Race
Hispanic	6
White	4
Black	1
Asian	1
Underlying conditions	10
Chronic conditions	4
Solid organ malignancy	3
Diabetes mellitus	2
Cirrhosis	1
Immunosuppressive conditions	7
HIV	5
AIDS (with CD4 < 100)	4
Allogenic HSCT	1
Solid organ transplant with PTLD	1
SLE	1
Charlson comorbidity score (median ± SD)	6 ± 3
Exposure history	4
Landscaping/soil exposure	3
Waterway exposure	1
Travel to endemic areas	11
Caribbean/Latin America	6
Puerto Rico	5
Guatemala	1
Midwest	3
East Coast (North Carolina)	2
Southwest	2
South-East Asia	1
Clinical presentation
Constitutional symptoms	6
Fever	5
Fatigue	3
Weight loss	3
Night sweats	1
Oral lesions	3
Respiratory symptoms	3
Cough	2
Pleuritis	1
Respiratory failure	1
Cutaneous lesions	2
Papular Rash	2
Ulcer	1
Acute mental status change	2
Gastrointestinal symptoms	2
Diarrhea	1
Hematochezia	1
Generalized edema	2
Laboratory abnormalities
LFTs abnormalities	7
Anemia (Hb < 12)	6
LDH > 200	6
Abnormal WBC	4
Leukopenia	3
Leukocytosis	1
Hyponatremia	4
Radiology findings
Lung nodules or lymphadenopathy	10
Splenomegaly	2
Liver lesions	2
Adrenal mass	1
Bone or soft tissue abnormalities	1
Brain lesions	1
Gastrointestinal lesions	1

AIDS, acquired immunodeficiency virus; HIV, human immunodeficiency virus; LDH, lactate dehydrogenase; LFT, liver function tests; PTLD, post-transplant lymphoproliferative disorder; SD, standard deviation; SLE, systemic lupus erythematosus.

Table 2.

Diagnostics and treatment outcomes for patients with active histoplasmosis.

Characteristic	N = 12
Source of diagnosis
Lung/Mediastinal tissue biopsy	4
Blood culture	2
Bone marrow biopsy	2
Bronchoalveolar lavage	2
Bone/Soft tissue biopsy	1
Oral ulcer biopsy	1
Colonic biopsy	1
Adrenal biopsy	1
Lymph node biopsy	1
Positive culture	5
Tissue	3/11
Blood	2
Cerebrospinal fluid	0/2
Positive histopathology	9/11
Positive cytology	2/7
Histoplasma Antigenemia	6/11
Urine Ag	6/11
Blood Ag	1/1
Cerebrospinal fluid Ag	1/1
Amphotericin B use	8
Liposomal AMB	5
Conventional AMB	3
Duration (median ± SD) (weeks)	2.5 ± 4
Azole use	10
Itraconazole	9
Voriconazole	1
Clinical outcome
Success	8
Death*	2
Escalation of treatment	1
Relapse	1
Duration of follow-up (median ± SD) (months)	2.5 ± 6
Time to diagnosis (median ± SD) (days)
From symptom onset**	41 ± 75
From initial evaluation	28 ± 30
Time to Histoplasma antigen testing (median ± SD) (days)	29 ± 35
Delay in diagnosis (≥14 days from admission)	7
Extensive workup (≥5 imaging modalities and/or invasive procedures)	9
Discharged at least once or death without diagnosis	6

AMB, amphotericin B; SD, standard deviation.

*Both deaths occurred in highly immunocompromised patients (HIV; solid organ transplantation) receiving amphotericin B.

**Time to diagnosis from symptom onset for the two patients who died were 41 and 57 days.

All patients with disseminated disease except for one patient with colonic involvement (n = 8) were treated initially with amphotericin B for a median of 2.5 weeks. Azoles were used initially for pulmonary and mild presentations (n = 4) and as step down therapy. Clinical improvement with treatment occurred in eight patients (67%). Failure included death attributable to histoplasmosis (n = 2), clinical worsening (n = 1), and relapse post-treatment (n = 1).

Since histoplasmosis is relatively uncommon in Connecticut, we sought to measure correlates of disease unfamiliarity among providers including time to diagnosis and extent of diagnostic testing. The median time to diagnosis from symptom onset and initial evaluation were 41 days (15–226) and 28 days (4–106), respectively. The median time to first Histoplasma antigen testing was 29 days (2–114). Extensive workup (≥5 imaging or invasive procedures) was performed in nine patients, while six were discharged (n = 4) or died (n = 2) without diagnosis.

This small Connecticut series of histoplasmosis cases in two New Haven hospitals over the past decade (2005–2015) suggests that the epidemiology of histoplasmosis may differ in nonendemic areas. Most patients were immunosuppressed with the most common cause being HIV/AIDS. This is consistent with reports from New York^5–7 and in contrast to endemic areas in which most cases involve immunocompetent patients with heavy soil exposure. Moreover, in endemic areas, most affected immunosuppressed individuals suffer from conditions other than HIV/AIDS such as hematologic malignancies or autoimmune disorders requiring immunosuppressive agents.⁸ The high prevalence of disseminated disease in this series likely parallels the prevalence of IS but may also be partially attributable to referral bias to a tertiary health center. Interestingly, as with series from Atlanta, Georgia,⁹ and New York,^6,7 travel to the Caribbean/Latin America, particularly Puerto Rico, was more common than travel to endemic areas in the continental United States, possibly a reflection of the substantial immigrant makeup of coastal cities. Active histoplasmosis cases in this series may have been due to reactivation disease remotely following initial exposure, though inoculum-dependent re-exposure has been proposed as a more common mechanism in immunosuppressed patients.^10,11 Nontraditional risk factors combined with healthcare provider disease unfamiliarity may have contributed to delays in diagnosis, extensive testing, and poor outcomes, consistent with previous findings from Montana and Idaho.¹² Taken together, these data suggest the need for a better understanding of epidemiologic features and an increased awareness among healthcare providers in these regions. Moreover, histoplasmosis should be considered in a patient with compatible signs and symptoms regardless of their known travel history.

Declaration of interest

There was no funding to support the preparation of this paper.

Lawrence Joseph Wheat, MD, is founder of MiraVista Diagnostics Indianapolis.

All other authors report no conflicts of interest.

The authors alone are responsible for the content and the writing of the paper.