Urine antigen-negative disseminated histoplasmosis mimicking post-transplant lymphoproliferative disorder

  1. Surbhi Gupta ,
  2. Colin Andrew Hinkamp and
  3. Matthew Lo
  1. Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA
  1. Correspondence to Dr Surbhi Gupta; surbhi.gupta@phhs.org

Publication history

Accepted:14 May 2020
First published:11 Jun 2020
Online issue publication:11 Jun 2020

Case reports

Case reports are not necessarily evidence-based in the same way that the other content on BMJ Best Practice is. They should not be relied on to guide clinical practice. Please check the date of publication.

Abstract

A 50-year-old woman with a history of kidney transplant presented with 2 days of abdominal pain after 6 months of recurrent streptococcal pharyngitis, fevers, weight loss and a new rash on her chest and back. Her examination was notable for a unilateral tonsillar exudate and 2–3 mm pink papules with a fine scale over her chest and back. CT of the abdomen and chest demonstrated several large lymph nodes, and laboratory investigation revealed new cytopenias and elevated transaminases. Urine antigen testing for Histoplasma capsulatum was negative, but a fungal complement fixation panel was reactive for Histoplasma antibodies. Skin biopsy revealed intracellular organisms consistent with H. capsulatum. She underwent treatment with liposomal amphotericin B but due to nephrotoxicity, drug interactions and worsening transaminitis, therapy was changed to itraconazole. The diagnosis and management of disseminated histoplasmosis presents multiple challenges, which are of particular importance in patients with a history of renal transplantation.

Background

Histoplasma capsulatum is a dimorphic fungus most prevalent in the USA in the Ohio and Mississippi River valleys, and is the most common of the endemic mycoses in the USA. Exposure occurs via aerosolised microconidia, thought to be associated with soil contaminated by bird and bat guano. In endemic regions it affects an estimated 50%–80% of the population, 90% of whom develop a self-limited asymptomatic infection with the organisms sequestered in calcified granulomas.1 2 In the immunocompromised, or those exposed to a high inoculum, exposure may lead to acute, chronic or disseminated infection.

Disseminated infection presents with a non-specific clinical picture characterised by lymphadenopathy, cytopenias, heterogeneous skin lesions, hepatosplenomegaly, and rheumatological manifestations. This may lead to a significant delay in the diagnosis and treatment of a disease that can reach a 10% mortality rate.2–4 Although antigen testing is most commonly used to identify histoplasmosis, the estimated sensitivity of urine antigen testing is 91%, and in one case series urine antigen was positive in only 79% of patients with disseminated disease.5 6 Thus, false-negative testing can further compound the diagnostic challenge. This case highlights the clinical, diagnostic and therapeutic roadblocks in the management of disseminated H. capsulatum infection in a solid organ transplant recipient.

Case presentation

A 50-year-old woman presented to the hospital with 2 days of generalised abdominal pain at the direction of her primary care physician. In the preceding 6 months she reported recurrent pharyngitis, intermittent fevers, tender cervical lymphadenopathy and 3-5 kg of weight loss. She was seen by her primary care physician numerous times for her recurrent pharyngitis, tested positive for streptococcal pharyngitis, and underwent antibiotic therapy at least twice. She also presented with a new asymptomatic rash to her chest, back and arms that was slowly progressive with new papules appearing every few days, none of which resolved (figures 1 and 2). Two days prior to presentation, she developed generalised abdominal pain for which she presented to her primary care physician, who obtained CT images of her abdomen and pelvis to evaluate her symptoms. She was instructed to present to the emergency department after the imaging identified new retroperitoneal lymphadenopathy and laboratory studies revealed new cytopenias. The remainder of a thorough review of systems was unremarkable. She denied travel outside of her home state of Texas in the last 2 years and had never been to the Ohio or Mississippi River valleys. She also denied bird or bat exposure.

Figure 1

Posterior neck with well-circumscribed pink papules ranging from 1 mm to 5 mm.

Figure 2

Anterior chest with several violaceous papules under the inframammary folds with mild surrounding erythema.

Her relevant previous medical history included a deceased donor kidney transplant performed 3 years ago. Her immunosuppression regimen included tacrolimus 1.5 mg two times per day, mycophenolate mofetil 540 mg two times per day and chronic prednisone 5 mg daily. Her tacrolimus levels had remained within therapeutic range, specified to be 6–8 ng/mL by her nephrologist, in the last year.

Investigations

She was well-appearing, afebrile and haemodynamically stable. Her physical examination revealed a unilateral tonsillar exudate, mild right upper quadrant abdominal tenderness and numerous discrete 2–3 mm violaceous pink papules with a fine scale over her chest and back. The remainder of her examination was unremarkable.

The patient’s primary care physician had initiated a preliminary investigation prior to her presentation to the hospital. Outpatient laboratory studies performed days prior to admission were notable for new cytopenias, with a platelet count of 105×109/L and leukocyte count of 4×109/L, which were lower than her baseline platelet count around 200×109/L and white cell count around 7×109/L. Due to a concern for post-transplant lymphoproliferative disorder, she underwent CT imaging of her chest, abdomen and pelvis which demonstrated extensive retroperitoneal and mediastinal lymphadenopathy (figures 3 and 4). Ground glass opacities were also noted in the right lower lobe of her lung.

Figure 3

CT of the abdomen and pelvis with contrast demonstrating upper retroperitoneal and central mesenteric lymphadenopathy. The arrow marks a 15.9×17.8 mm lymph node.

Figure 4

CT of the chest without contrast demonstrating mediastinal lymphadenopathy.

At the time of her presentation to the hospital, her laboratory studies demonstrated a normal complete blood count lacking cytopenias, while her comprehensive metabolic panel was significant for an aspartate transaminase (AST) that peaked at 65 units/L, alanine transaminase (ALT) that peaked at 104 units/L and alkaline phosphatase that peaked at 134 units/L. Her initial serum creatinine was 1.46 mg/dL, elevated from her baseline level of 1.2 mg/dL. Due to the initial concern for post-transplant lymphoproliferative disorder, her initial work-up consisted of Epstein-Barr virus, Cytomegalovirus and BK virus testing which were negative. Additionally, her serum and urine protein electrophoreses did not demonstrate an M-component.7

She began to have fevers on her first day of hospitalisation and was started on broad-spectrum antibiotics which were stopped after her bacterial blood and urine cultures, drawn prior to antibiotic administration, had remained negative for 48 hours. Throughout the hospital course her blood counts gradually declined and she developed pancytopenia with nadirs of leukocyte count as low as 3.8×109/L, haemoglobin of 109 g/L and platelet count of 133×109/L.

Given her evolving hospital course, a broader work-up was commenced. Her bacterial and mycobacterial blood cultures and an interferon-gamma release assay (QuantiFERON-TB Gold) were negative. She had a transthoracic echocardiogram which demonstrated normal valvular function and ejection fraction without evidence of vegetations. HIV testing was also negative.

A biopsy of a skin lesion from her left arm was performed (figures 5 and 6), which demonstrated a diffuse infiltrate with numerous histiocytes and containing small round microorganisms. H. capsulatum was suspected and confirmed by fungal culture of the tissue sample.

Figure 5

Left arm skin biopsy demonstrating diffuse infiltrate with numerous histiocytes on methenamine silver stain.

Figure 6

Left arm skin biopsy demonstrating intracellular yeast measuring 2–5 µm in diameter with a clear space around them with a few foci of narrow-based budding on H&E stain at 400× original magnification.

Interestingly, while her urine antigen was repeatedly negative, her serum fungal complement fixation panel demonstrated positive Histoplasma serologies at titers of 1:256 for the mycelial phase antigen and 1:16 for the yeast phase antigen. Also included on the fungal panel were Coccidioides and Blastomyces, which were both negative. Serum antigen testing was not done. Fungal blood cultures, drawn after broad spectrum antibiotics but prior to antifungals, demonstrated rare mould growth and identified as Histoplasma, which helped to confirm the diagnosis. Biopsy of an enlarged lymph node was not considered, as the suspicion for a neoplastic process was now lower and the lymphadenopathy was more likely to be secondary to disseminated infection.

Differential diagnosis

Initially, given the non-specific constellation of signs and symptoms, history of transplant, and immunosuppressed state, we maintained a broad differential diagnosis which included bacterial, mycobacterial and fungal infections including opportunistic pathogens, subacute HIV, endocarditis, lymphoma and post-transplant lymphoproliferative disorder. We felt her abdominal imaging with bulky retroperitoneal and mediastinal lymphadenopathy, with several large lymph nodes, was most consistent with post-transplant lymphoproliferative disorder, which served as our initial working diagnosis.

Treatment

After establishing the diagnosis of disseminated histoplasmosis, therapy was initiated with liposomal amphotericin B. However, within 24 hours of the infusion she developed acute renal injury, which is a well-characterised adverse effect of the drug, as well as rising transaminase levels. Her transiently reduced renal function and presumed drug–drug interactions with her immunosuppression caused elevated levels of tacrolimus into the supra-therapeutic range, as high as 16.0 ng/mL. Thus, she was unable to complete the planned induction therapy of amphotericin B for 2 weeks, and instead started liquid itraconazole. The liquid formulation was chosen initially due to its higher bioavailability, but unfortunately her insurance would not approve the liquid formulation.8 She was instead discharged with capsules of itraconazole with the intent to treat for 12 months with close follow-up in the infectious diseases clinic. At the time of discharge, her renal function had begun to recover and her tacrolimus levels returned to within the therapeutic range. Serum itraconazole levels were monitored during the duration of therapy and were consistently within the therapeutic range of >1.0 µg/mL.

Outcome and follow-up

In the months since discharge, the patient continues to follow-up at our infectious diseases clinic and is tolerating itraconazole well. Her complete blood count, liver function tests, and renal function have returned to her baseline. Her symptoms including abdominal pain, fevers, weight loss and anterior cervical lymphadenopathy have resolved. Imaging to assess the retroperitoneal and mediastinal lymphadenopathy since initiating therapy has not been repeated. No fungal studies have been repeated since discharge.

Discussion

Solid organ transplant recipients on chronic immunosuppressive therapy are at higher risk for a number of illnesses after their transplant including haematological malignancies, opportunistic infections and transplant rejection. In a multicentre study looking at endemic fungal infections in solid organ transplant recipients, there were 5989 transplant recipients included, of whom 22 patients developed Histoplasma infection. Of these patients, 14 (64%) had disseminated infection. None of the identified patients had cutaneous involvement.3

Cutaneous lesions in the setting of disseminated disease is surprisingly uncommon, but seems to occur more frequently in renal transplant patients. In a case series by Sun et al, out of the 69 identified cases of dissemination, 17 (25%) patients exhibited skin lesions.9 This is higher than the 17% rate seen in generally immunocompromised patients observed in another retrospective study.5 The cutaneous manifestations of Histoplasma can vary greatly, ranging from plaques to purpura, but when identified can provide a convenient source for biopsy and culture, which was the case in our patient.9 Our patient also exhibited what initially seemed to be a tonsillar exudate, but in hindsight may have represented mucocutaneous ulcers, another manifestation of cutaneous histoplasmosis. The significance of her symptomatic pharyngitis with concomitant positive rapid streptococcal antigen testing done at her primary care physician’s office is unclear, though could have been attributed to a coincidental infection or colonisation.

Although its presentation is heterogeneous, disseminated histoplasmosis in immunocompromised patients most commonly presents with fever (66%), hepatosplenomegaly (55%) and respiratory symptoms (35%). Many patients also present with thrombocytopenia (52%), elevated liver enzymes (elevated ALT in 41%, elevated AST in 38%) and elevated alkaline phosphatase (51%).5 Our patient did not have respiratory symptoms, but did have ground glass opacities in her right lower lobe on CT imaging of the chest. Additionally, while she had elevated liver enzymes and alkaline phosphatase, she lacked any cytopenias on initial presentation to the hospital.

Once clinical suspicion is raised, testing for disseminated histoplasmosis can be difficult. Although there are several ways to diagnose disseminated histoplasmosis, initial testing comprises of serum or urine antigen detection.6 Urine antigen can take up to 7 days to finalise and has a reported sensitivity of 90%. However, the measure of sensitivity of testing varies in different case series and has been reported to be as low as 79%, suggesting that even in disseminated disease, urine antigen testing may not always be reliable.5 This was seen in this patient, who had a negative urine antigen despite repeated testing. Alternatively, culture may be used for confirmatory testing, most often taken from a bronchoalveolar lavage specimen, blood or bone marrow biopsy. In other cases, histopathology may be obtained, usually from the liver, lung or a lymph node.3 In one case series, half of the patients with post-transplant histoplasmosis had both positive blood cultures and histopathology, while the remaining half had H. capsulatum seen in at least two organs.10

Likely a result of the heterogeneous clinical presentation and unreliable screening tests, the median time from symptom onset to diagnosis is roughly 30 days.3 Although estimates of mortality vary, in one case series of 22 renal transplant patients with disseminated histoplasmosis, the mortality rate was as high as 29%. This is higher than the reported 10% mortality rate in recipients of other solid organ transplants.4 Thus, early detection is imperative for a more favourable outcome.

The Infectious Disease Society of America’s 2007 updated guidelines for treatment of Histoplasma recommend initiating treatment for disseminated disease with liposomal amphotericin B for 1–2 weeks, followed by itraconazole therapy for 12 months.11 Liquid itraconazole has higher bioavailability and achieves 30% higher blood concentrations than capsules and is therefore the preferred agent. Itraconazole notably can have widely varying blood concentrations and thus should be monitored by serum concentration about 2 weeks after initiating therapy. The recommended target is >1.0 µg/mL, and due to the long half-life of about 24 hours, serum collection does not have to be timed in relation to the last dose of itraconazole. Levels over 10 µg/mL may be associated with toxicity. While there is no recommended interval for the frequency of drug monitoring, repeating the serum concentration can be considered when there is a change in itraconazole dose or there is concern for inadequate compliance or treatment failure.10 11

While treatment is often successful, it can have numerous complications that the clinician should be aware of, and that were of particular concern in our patient. Liposomal amphotericin B is associated with kidney injury. Further, alternate therapies such as the azoles inhibit CYP3A4 and P glycoprotein, which are involved in the primary mechanisms of action of sirolimus and calcineurin inhibitors such as tacrolimus.12 Thus, drug interactions with crucial immunosuppressive medications in transplant patients can be particularly problematic and require close monitoring.

Disseminated histoplasmosis is a rare disease that is associated with significant morbidity and mortality. Physicians may face several obstacles in the management of these patients along any step from identification to diagnosis to treatment. With the growing population of patients with organ transplants, knowledge of these pitfalls can serve to improve recognition and management of this vulnerable population.

Learning points

  • A rare clinical manifestation of disseminated Histoplasma infection is cutaneous disease, which may occur more frequently in patients with renal transplants.

  • Histoplasma urine antigen is the most sensitive test; however, it is important to consider additional testing such as tissue culture and biopsy if antigen testing is negative and there is high clinical suspicion.

  • Liquid itraconazole is the preferred therapy, after 1–2 weeks of initiation with liposomal amphotericin B, due to its higher bioavailability, but may not always be financially feasible. In these cases, itraconazole capsules may be used.

  • Due to the highly variable serum concentrations of itraconazole when undergoing therapy for histoplasmosis, therapeutic drug monitoring is recommended at least 2 weeks after initiation of therapy with a recommended target serum concentration of >1.0 µg/mL.

Acknowledgments

Thank you to Mitchell von Itzstein, MD who provided valuable follow-up information and obtained consent from the patient.

Footnotes

  • Contributors SG contributed to the acquisition, analysis and interpretation of data, drafting and revising of the manuscript. CAH contributed to the acquisition, analysis and interpretation of data, drafting and revising of the manuscript. ML contributed to the analysis and interpretation of data and revising of the manuscript. All authors agreed to the final version of the manuscript and agreed to be accountable for the article.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References

Use of this content is subject to our disclaimer