Hemorrhage Incognito

A 51-year-old man with a history of end-stage renal disease secondary to hypertension and type II diabetes mellitus underwent a deceased donor kidney transplant. The donor's cause of death was anoxic brain injury after aspirating a piece of food. The recipient received thymoglobulin and methylprednisolone for induction immunosuppression, then tacrolimus, mycophenolic acid, and prednisone for maintenance immunosuppression. The patient had significant post-transplant hemorrhagic complications requiring surgical management. He was taken to the operating room (OR) for peri-renal transplant hematoma evacuation on post-transplant days 1 and 4. Each OR visit demonstrated peri-transplant hematomas without arterial or venous anastomosis failure. On post-transplant day 10, the patient developed a pruritic, scaly, maculopapular rash on the abdomen, bilateral lower extremities, and flanks (Figure 1). Based on clinical presentation and history, dermatology evaluated the patient and found that the presentation was most consistent with dermatitis secondary to topical antiseptic versus dermatophyte infection. Potassium hydroxide preparation of the skin was negative for hyphae, and the patient was started on triamcinolone at 0.1% BID. A skin biopsy showed acute spongiotic dermatitis with subcorneal intraepidermal pustulosis, neutrophilic exocytosis, and superficial dermal mononuclear and polynuclear cellular infiltrate. The patient's clinical presentation and pathology results favored a diagnosis of acute localized exanthematous pustulosis (ALEP) at the time. The patient had two additional episodes of significant hemorrhage requiring packed red blood cell (RBC) transfusions and returned to the OR on post-transplant days 11 and 13. Once again, no discrete source of bleeding was identified. On post-transplant day 18, the patient developed bilious emesis, and a computed tomography (CT) of the abdomen revealed evidence of a small bowel obstruction (SBO). He was taken to the OR for lysis of adhesions, and peri-transplant hematoma was again encountered. Cultures were obtained from the hemorrhagic fluid (Movie S1). On post-transplant day 32, the patient was discharged to an acute rehabilitation facility. His medications at the time of discharge included tacrolimus, mycophenolic acid, prednisone, valganciclovir, and trimethoprim-sulfamethoxazole. Approximately 48 h after discharge, the patient became hemodynamically unstable, requiring transfer to a local hospital. The patient underwent an exploratory laparotomy and was found to have an intraabdominal hemorrhage from the inferior pole artery of the graft. Intraoperative culture from an evacuated hematoma grew Serratia marcescens. Complications from this episode of hemorrhagic shock included acute kidney injury and ischemic hepatopathy. The patient was started on cefepime, and following clinical stabilization, the patient was transferred back to our facility. The transplant infectious diseases service was consulted for antibiotic management on post-transplant day 34. Physical exam revealed a temperature of 35.5°C, heart rate of 71 beats per minute, blood pressure of 138/59 mmHg, respiratory rate of 10 breaths per minute, and SpO2 99% on room air. He appeared non-toxic, alert, and oriented. The cardiopulmonary exam was unremarkable. His abdomen was soft, non-tender, and non-distended. The abdominal drain in the right lower quadrant (RLQ) had serosanguinous fluid. The RLQ surgical incision was healing well. An extensive, confluent, erythematous, scaly rash with associated ulceration and desquamation was present on the abdomen and upper thighs with a well-demarcated, raised edge. The remainder of the exam was unremarkable. Full laboratory test results are shown in Table 1. The patient developed altered mental status after a few days of cefepime therapy, raising concerns for cefepime neurotoxicity. Cefepime was held, and the patient was started on levofloxacin. He developed worsening pancytopenia and was found to have an elevated ferritin level of 5951 ng/mL. Blood cultures were negative. A repeat CT of the abdomen and pelvis was negative for evidence of intraabdominal abscess. The operative cultures from the post-transplant day 18 lysis of adhesions and washout were now growing a mold. The cultures obtained on post-transplant day 34 at the outside hospital also grew a mold in addition to Serratia. The organisms from both cultures appeared morphologically similar. Mold infections in the early (<1 month) post-transplant period are uncommon but can be highly morbid. The identity of the mold has implications for both diagnosis and management. From a diagnostic perspective, knowledge of the organism will clarify the clinical syndrome and its likely pathogenesis, which could influence subsequent testing in the patient and/or investigation of potential exogenous sources of infection. From a management perspective, knowledge of the organism influences the choice of optimal antifungal therapy, the need for adjunctive source control interventions, and potential risk mitigation strategies for recipients of other organs and tissues from the same donor if the infection is suspected to be donor-derived. Accurate laboratory identification of molds is often constrained by variables that cannot be easily accelerated, including growth characteristics of the organism on solid media and its development of recognizable reproductive structures (e.g., conidia). Furthermore, test methodologies for rapid identification of bacterial organisms, such as mass spectrometry, may not be deployed as widely for mold identification, and molecular methods are usually only available through reference or research laboratories. Given these constraints, definitive identification of pathogenic molds is often delayed. Thus, to expedite decision-making, clinicians must be able to anticipate the most likely pathogen(s) based on epidemiologic risk factors, features of the clinical syndrome, and microbiologic characteristics of the organism in the clinical laboratory. I will consider each of these aspects of the case separately to help prioritize potential pathogens. The epidemiology of invasive fungal infections (IFI) among solid organ transplant (SOT) recipients varies by organ type, time after transplant, and antimicrobial prophylaxis strategy. Invasive candidiasis is the most common IFI across organ groups, with mold infections being a distant second [1]. Mold infections are also comparatively less common in the kidney compared with liver or thoracic organ recipients [1, 2]. The most common mold infections among kidney recipients include aspergillosis and mucormycosis. 2 However, these typically manifest at least one month after transplant, when the cumulative effect of immunosuppression is more pronounced. For example, in one registry analysis of IFI among SOT recipients, the median time to onset of invasive mold infection in kidney recipients was 406 days (interquartile range 149–1758 days) [3]. Given these observations, the early onset of infection in this patient (<2 weeks after transplant) strongly suggests that it is linked to the transplant procedure itself. There are many potential time points in the organ procurement and transplantation process continuum where pathogenic organisms may be acquired and evolve into peri-transplant infections [4]. Considerations include donor-derived infection, contamination of the organ following procurement but before implantation, nosocomial infection in the recipient, or pre-existing infection or colonization in the recipient whose trajectory is modified by surgery or immunosuppression. Donor-derived mold infections are rare events and have been associated with specific risk factors, including the immune status of the donor, donors who were victims of drowning/near-drowning events, and transplant tourism [5, 6]. One transmission event occurred in the setting of a pre-hospital “in-the-field” tube thoracostomy that inadvertently caused organ contamination [7]. None of these risk factors appear relevant to this case. Nosocomial mold infections related to contaminated air (e.g., operating room ventilation), fluid (e.g., irrigation fluid), or surgical paraphernalia (e.g., packing, dressing, etc.) have all been described. They may need to be considered if the mold is ultimately identified as an environmental organism (Suleyman). Thus, based just on the epidemiologic features of this case and published reports of peri-transplant/peri-operative mold infections, potential pathogens include Aspergillus, mucormycetes, Fusarium, Scedosporium, and dematiaceous molds. The patient's clinical syndrome consists of two key features that are useful for further refining the microbiologic differential diagnosis–recurrent allograft-related hemorrhage and rash. Recurrent allograft-related hemorrhage indicates a localized vascular process, which could involve the kidney parenchyma or larger vessels. Fungal arteritis/mycotic aneurysm after a kidney transplant is a rare and typically graft-threatening complication, one that usually arises because of active bloodstream infection in the donor [8]. Most of these infections are due to Candida species. However, cases of Aspergillus renal arteritis have been described. Notably, most of these presented with acute graft dysfunction, abdominal pain, and frank vascular anastomotic disruption, leading to life-threatening hemorrhage that necessitated urgent graft nephrectomy. [9] Similarly, donor-derived mold infections acquired through organs procured from near-drowning victims also manifest with graft complications and can even disseminate to the lungs or nervous system. [10] These observations stand in contrast to the case patient, who has experienced recurring hemorrhage but has intact vascular anastomosis, no signs of disseminated disease, and has survived for over a month post-transplant without anti-fungal therapy. These features suggest a more indolent organism, such as Scedosporium or dematiaceous fungi, and make highly virulent molds–such as mucormycetes, Fusarium, and, to some extent, Aspergillus–less likely. The dermatologic component of the patient's presentation is not characteristic of any specific fungal infection. However, it could be a clue to the syndrome's pathogenesis and implicate additional organisms. Although the histopathology was interpreted as suggestive of ALEP, several features of the rash are atypical for this diagnosis, namely the absence of an obvious trigger (most published cases are due to medications, not topical agents) and the lack of clinical improvement despite the withdrawal of the presumed offending agent (topical antiseptic) [11]. If the rash represents a (non-infectious) drug reaction, localized skin barrier disruption close to the surgical incision may have served as the portal of entry for an environmental pathogen or skin commensal. Alternatively, the rash could represent a coexistent fungal skin/soft tissue infection, perhaps inoculated from the superficial layers of the skin into the deeper tissues and peri-allograft space at the time of surgery. Given these potential mechanisms of infection, dermatophytes such as Trichophyton should also be considered in addition to the indolent organisms noted previously. Notably, however, allograft infection and associated hemorrhage due to these organisms have not been previously described. Finally, we can infer some features of the pathogen based on the limited microbiologic information available. The organism recovered from operative cultures at postoperative days 18 and 34 and took days to more than a week to grow. Furthermore, since a fungal infection was not suspected when these cultures were obtained, the organism was recovered without any prespecified nutritional supplementation or optimized growth conditions. These characteristics again support the hypothesis of an indolent organism. In sum, based on the epidemiologic, clinical, and microbiologic features of the case, this patient is likely experiencing an invasive infection due to an indolent mold that was presumably introduced into the deep soft tissues and peri-allograft space at or around the time of the transplant procedure. Based on this analysis, the likeliest pathogens include indolent hyaline molds such as Scedosporium, dematiaceous fungi, or dermatophytes such as Trichophyton. Non-invasive testing for fungal infection was obtained, and all tests returned negative, including a Histoplasma urine antigen, Blastomyces urine antigen, Cryptococcal antigen, and galactomannan. The patient was empirically started on isavuconazole. The transplant infectious diseases service recommended liposomal amphotericin B, but this treatment was deferred due to concerns about kidney graft function. The patient continued to have altered mental status, pancytopenia, and elevated inflammatory markers, so a bone marrow biopsy was performed, which revealed evidence of hemophagocytosis. He was diagnosed with hemophagocytic lymphohistiocytosis (HLH), and the hematology service recommended treatment with dexamethasone. The mold isolated on post-transplant days 18 and 34 was finally identified as Trichophyton rubrum. A third culture isolated during another hematoma evacuation on post-transplant day 52 also contained T. rubrum. Following another event of symptomatic hemorrhage from the renal graft and worsening coagulopathy, the patient underwent graft nephrectomy post-transplant day 58. Histopathology of the graft was strongly suggestive of an angioinvasive fungal infection within the kidney with expansion into the surrounding vessels and perinephric soft tissue (Figure 2). A specimen from the right renal artery revealed evidence of acute arteritis and an organizing fungal thrombus. Septate hyphae compatible with T. rubrum were seen with Grocott's methenamine silver stain (Figure 3). After the nephrectomy, the patient was started on liposomal amphotericin B for treatment of both the severe mold infection and the newly diagnosed concurrent C. glabrata bloodstream infection. Following an episode of acute decompensation involving shock and worsening hypoxemic respiratory failure, the family decided to pursue comfort care. The patient subsequently expired on post-transplant day 92. In summary, we presented a kidney transplant recipient whose post-transplant course was complicated by recurrent bleeding, necessitating multiple takebacks in the setting of a severe abdominal rash. Initially, the recurrent bleeding complications proved challenging to understand. Ultimately, the findings suggested the introduction of a relatively benign skin infection to the abdominal cavity during multiple surgeries. This conclusion is supported by the isolation of T. rubrum from three independent operative abdominal cultures. This unique clinical presentation of a dermatophyte, T. rubrum, infecting a deep organ space following renal transplantation. Cutaneous infection with T. rubrum can be incredibly challenging to diagnose due to non-specific clinical presentation and propensity to produce a false negative on potassium hydroxide preparation [12]. A literature review yielded several case reports of missed or delayed diagnoses. Nowowiejska et al. [13] report a case of a 57-year-old male who presented with erythematous-exfoliative lesions in his groins and on the buttocks, initially diagnosed as eczema but later diagnosed as tinea incognito. Dovalovsky et al. [14] report a case of a 38-year-old female who presented with an itchy rash on bilateral medial thighs for the past several months​, initially thought to be psoriatic plaques. Seo et al. [15] report a case in which Tinea corporis mimicked ALEP, a diagnosis that was entertained for our patient as well. Our presentation highlights the challenging cutaneous diagnosis and a rare report of deep invasion with T. rubrum. In most invasive dermatophytosis case reports, the disease is limited to deep dermal involvement with or without invasion of local lymphatics. Mularoni et al. [16] reported a case of invasive T. rubrum infection to the dermal-epidermal junction in a patient who underwent liver transplantation 17 years prior. Marconi et al. [17] demonstrated biopsy-proven microvascular invasion of T. rubrum in a patient with cirrhosis due to hemochromatosis and alcohol abuse who presented with multiple cutaneous lesions over the face and extremities. Frankly disseminated and visceral disease has rarely been reported. In cases of disseminated dermatophytosis, the majority occur in immunocompromised hosts such as our patient. CARD9 deficiency has also been associated with the development of deep dermatophytosis, likely due to its involvement in immune defense against fungal pathogens [18]. Our patient's immune system was severely compromised, and he had a particular deficiency in cell-mediated immunity after receiving thymoglobulin and subsequent treatment with high-dose steroids for possible HLH. One case report described a case of cutaneous disease of the face secondary to Trichophyton rubrum with dissemination to the lung and chest wall in a kidney transplant patient; dissemination was assumed based on the clinical response of lung lesions to antifungal therapy [19]. Histopathologically-confirmed angioinvasive disease with visceral involvement, as we report, appears to be quite rare. Comparative data for the treatment of invasive dermatophytosis is not available. However, as in other invasive mold infections in immunocompromised hosts, the most effective treatment regimen is multifaceted, involving antifungal therapy, immunosuppression reduction, and source control surgery. As far as medical treatment is concerned, antifungals with excellent in vitro activity against T. rubrum include terbinafine, triazoles besides fluconazole, and amphotericin formulations; there is no data on the use of combination therapy, but 20% of patients with disseminated infection reported by Marconi et al. were treated with multiple agents [17]. The authors declare no conflicts of interest. Data are available on request from the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.