Arturo R. Dominguez, MD

To the Editor:

A 56-year-old man with a history of stage IV metastatic penile squamous cell carcinoma treated with penectomy and chemotherapy with 5-fluorouracil and cisplatin presented with several painful ulcerations in the groin, abdomen, and thighs. The lesions initially appeared in the groin and were treated as bacterial abscesses with antibiotics. Over the next few weeks, new lesions appeared on the abdomen and thighs. An additional cycle of chemotherapy led to a reduction in number; however, they again increased within a few weeks. Medications included enoxaparin followed by 3 weeks of warfarin use due to a right leg deep vein thrombosis.

Physical examination revealed multiple 1- to 4-cm, firm, ulcerated nodules on the bilateral inguinal folds, abdomen, and upper thighs, as well as areas of livedo racemosa and noninflammatory retiform purpura with central ulceration (Figures 1 and 2). This retiform purpura was both perilesional and in areas without ulcerations. Laboratory values included the following: sodium, 127 mmol/L (reference range, 136–145 mmol/L); prothrombin time, 16.1 seconds (reference range, 11–15 seconds); white blood cell count, 20.69×10⁹/L (reference range, 4.5–11.0×10⁹/L) with 87% neutrophils (reference range, 54%–62%); hemoglobin, 6.1 g/dL (reference range, 13.5–17.5 g/dL); hematocrit, 18.8% (reference range, 41%–53%); platelets, 474×10⁹/L (reference range, 150–400×10⁹/L); D-dimer, 0.77 mg/L (reference range, ≤0.50 mg/L); fibrinogen, 489 mg/dL (reference range, 150–400 mg/dL); prior urine culture positive for Pseudomonas aeruginosa. He was negative for hepatitis B and hepatitis C viruses as well as HIV, and the lesions were not clinically consistent with herpes simplex virus, as they were not scalloped or circinate. Punch biopsies were obtained from a nodule on the left leg and a purpuric patch on the right leg.

Ischemic cutaneous lesions less commonly occur in the setting of malignancy and can be the result of both direct and indirect systemic effects from the cancer. Malignancies are known to directly trigger vasculopathies in other organs, most commonly the lungs, through 2 primary mechanisms. First, in carcinomatous arteriopathy, metastatic cells promote fibrocellular intimal proliferation of small pulmonary arteries and arterioles leading to stenosis, thrombosis, and obliteration. This mechanism has been described in pulmonary thrombotic microangiopathy secondary to lung carcinoma.⁴ This pathophysiology likely is also what underlies paraneoplastic acral vascular syndromes, which culminate in digital ischemia. Hypothesized mechanisms for this ischemia also range from vasospasm to thromboembolism.⁵ Secondly, in vasculitis carcinomatosa, metastatic tumor cells damage or block vessel walls, resulting in end-organ ischemia. Vasculitis carcinomatosa is a well-known phenomenon in angiocentric and intravascular lymphoid malignancies (typically of B-T or natural killer/T-cell origin) but also has been reported in a case of gastric adenocarcinoma with arterial invasion.⁶ This process is different than carcinoma telangiectoides where malignant cells may be present in the vasculature on histopathology but not trigger thrombosis and ischemic skin necrosis.

Systemic coagulopathies such as disseminated intravascular coagulation (DIC), thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome can occur in the setting of malignancies.⁷ Clinically, all may present with livedo racemosa, noninflammatory retiform purpura, and widespread skin necrosis. In adult patients, purpura fulminans most often is seen in the setting of sepsis and DIC, with accompanying evidence of microangiopathy.⁸ Catastrophic antiphospholipid antibody syndrome can be triggered by malignancy and is characterized by central nervous system, renal, pulmonary, and gastrointestinal complications. Skin involvement such as ulcers, livedo reticularis, and gangrene have been reported.⁹ Other causes of thrombotic vasculopathy include warfarin necrosis, heparin-induced thrombotic thrombocytopenia, calciphylaxis, and angioinvasive infections.⁸ Warfarin necrosis and heparin-induced thrombotic thrombocytopenia typically present days after initiating therapy with the respective medication. Calciphylaxis typically occurs in patients on dialysis, though it may occur in nonuremic patients including those with malignancy.⁸ Patients with malignancies on chemotherapy can become neutropenic and are at risk for ecthyma gangrenosum due to P aeruginosa and other gram-negative rods, Staphylococcus aureus, and angioinvasive fungi.¹⁰

Based on clinical, histopathological, and laboratory data, we favored a diagnosis of cutaneous carcinomatous arteriopathy. Vasculitis carcinomatosa was a possibility despite the lack of vasculotropism on histopathology, which may have been due to biopsy site selection. Systemic coagulopathies such as DIC, thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome were unlikely, as the ischemic skin lesions and livedo racemosa were limited to areas adjacent to cutaneous metastases, and the patient lacked other common multiorgan manifestations or laboratory findings. Although our patient was on warfarin, he was on a stable dose for weeks and histopathologic features of subcutaneous thrombosis were not seen. The biopsy also was not consistent with calciphylaxis. Ecthyma gangrenosum was unlikely given the lack of organisms on histopathology and negative skin and blood cultures. Although additional laboratory testing in this patient may have included cryoglobulins and cryofibrinogens, both entities were unlikely due to a lack of ischemic acral lesions.

In conclusion, we present a case of localized thrombotic vasculopathy that likely was secondary to cutaneous carcinomatous arteriopathy in the setting of cutaneous metastatic penile squamous cell carcinoma. The differential diagnosis of retiform purpura, livedo racemosa, and other signs of cutaneous ischemia in patients with metastatic cancer is broad and can be the result of both direct and indirect systemic effects from the cancer. Appropriate workup in these cases should include skin biopsies for histopathology and culture, medication review, and laboratory evaluation for systemic coagulopathies.

To the Editor:

A 56-year-old man with a history of stage IV metastatic penile squamous cell carcinoma treated with penectomy and chemotherapy with 5-fluorouracil and cisplatin presented with several painful ulcerations in the groin, abdomen, and thighs. The lesions initially appeared in the groin and were treated as bacterial abscesses with antibiotics. Over the next few weeks, new lesions appeared on the abdomen and thighs. An additional cycle of chemotherapy led to a reduction in number; however, they again increased within a few weeks. Medications included enoxaparin followed by 3 weeks of warfarin use due to a right leg deep vein thrombosis.

Physical examination revealed multiple 1- to 4-cm, firm, ulcerated nodules on the bilateral inguinal folds, abdomen, and upper thighs, as well as areas of livedo racemosa and noninflammatory retiform purpura with central ulceration (Figures 1 and 2). This retiform purpura was both perilesional and in areas without ulcerations. Laboratory values included the following: sodium, 127 mmol/L (reference range, 136–145 mmol/L); prothrombin time, 16.1 seconds (reference range, 11–15 seconds); white blood cell count, 20.69×10⁹/L (reference range, 4.5–11.0×10⁹/L) with 87% neutrophils (reference range, 54%–62%); hemoglobin, 6.1 g/dL (reference range, 13.5–17.5 g/dL); hematocrit, 18.8% (reference range, 41%–53%); platelets, 474×10⁹/L (reference range, 150–400×10⁹/L); D-dimer, 0.77 mg/L (reference range, ≤0.50 mg/L); fibrinogen, 489 mg/dL (reference range, 150–400 mg/dL); prior urine culture positive for Pseudomonas aeruginosa. He was negative for hepatitis B and hepatitis C viruses as well as HIV, and the lesions were not clinically consistent with herpes simplex virus, as they were not scalloped or circinate. Punch biopsies were obtained from a nodule on the left leg and a purpuric patch on the right leg.

Ischemic cutaneous lesions less commonly occur in the setting of malignancy and can be the result of both direct and indirect systemic effects from the cancer. Malignancies are known to directly trigger vasculopathies in other organs, most commonly the lungs, through 2 primary mechanisms. First, in carcinomatous arteriopathy, metastatic cells promote fibrocellular intimal proliferation of small pulmonary arteries and arterioles leading to stenosis, thrombosis, and obliteration. This mechanism has been described in pulmonary thrombotic microangiopathy secondary to lung carcinoma.⁴ This pathophysiology likely is also what underlies paraneoplastic acral vascular syndromes, which culminate in digital ischemia. Hypothesized mechanisms for this ischemia also range from vasospasm to thromboembolism.⁵ Secondly, in vasculitis carcinomatosa, metastatic tumor cells damage or block vessel walls, resulting in end-organ ischemia. Vasculitis carcinomatosa is a well-known phenomenon in angiocentric and intravascular lymphoid malignancies (typically of B-T or natural killer/T-cell origin) but also has been reported in a case of gastric adenocarcinoma with arterial invasion.⁶ This process is different than carcinoma telangiectoides where malignant cells may be present in the vasculature on histopathology but not trigger thrombosis and ischemic skin necrosis.

Systemic coagulopathies such as disseminated intravascular coagulation (DIC), thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome can occur in the setting of malignancies.⁷ Clinically, all may present with livedo racemosa, noninflammatory retiform purpura, and widespread skin necrosis. In adult patients, purpura fulminans most often is seen in the setting of sepsis and DIC, with accompanying evidence of microangiopathy.⁸ Catastrophic antiphospholipid antibody syndrome can be triggered by malignancy and is characterized by central nervous system, renal, pulmonary, and gastrointestinal complications. Skin involvement such as ulcers, livedo reticularis, and gangrene have been reported.⁹ Other causes of thrombotic vasculopathy include warfarin necrosis, heparin-induced thrombotic thrombocytopenia, calciphylaxis, and angioinvasive infections.⁸ Warfarin necrosis and heparin-induced thrombotic thrombocytopenia typically present days after initiating therapy with the respective medication. Calciphylaxis typically occurs in patients on dialysis, though it may occur in nonuremic patients including those with malignancy.⁸ Patients with malignancies on chemotherapy can become neutropenic and are at risk for ecthyma gangrenosum due to P aeruginosa and other gram-negative rods, Staphylococcus aureus, and angioinvasive fungi.¹⁰

Based on clinical, histopathological, and laboratory data, we favored a diagnosis of cutaneous carcinomatous arteriopathy. Vasculitis carcinomatosa was a possibility despite the lack of vasculotropism on histopathology, which may have been due to biopsy site selection. Systemic coagulopathies such as DIC, thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome were unlikely, as the ischemic skin lesions and livedo racemosa were limited to areas adjacent to cutaneous metastases, and the patient lacked other common multiorgan manifestations or laboratory findings. Although our patient was on warfarin, he was on a stable dose for weeks and histopathologic features of subcutaneous thrombosis were not seen. The biopsy also was not consistent with calciphylaxis. Ecthyma gangrenosum was unlikely given the lack of organisms on histopathology and negative skin and blood cultures. Although additional laboratory testing in this patient may have included cryoglobulins and cryofibrinogens, both entities were unlikely due to a lack of ischemic acral lesions.

In conclusion, we present a case of localized thrombotic vasculopathy that likely was secondary to cutaneous carcinomatous arteriopathy in the setting of cutaneous metastatic penile squamous cell carcinoma. The differential diagnosis of retiform purpura, livedo racemosa, and other signs of cutaneous ischemia in patients with metastatic cancer is broad and can be the result of both direct and indirect systemic effects from the cancer. Appropriate workup in these cases should include skin biopsies for histopathology and culture, medication review, and laboratory evaluation for systemic coagulopathies.

To the Editor:

A 56-year-old man with a history of stage IV metastatic penile squamous cell carcinoma treated with penectomy and chemotherapy with 5-fluorouracil and cisplatin presented with several painful ulcerations in the groin, abdomen, and thighs. The lesions initially appeared in the groin and were treated as bacterial abscesses with antibiotics. Over the next few weeks, new lesions appeared on the abdomen and thighs. An additional cycle of chemotherapy led to a reduction in number; however, they again increased within a few weeks. Medications included enoxaparin followed by 3 weeks of warfarin use due to a right leg deep vein thrombosis.

Physical examination revealed multiple 1- to 4-cm, firm, ulcerated nodules on the bilateral inguinal folds, abdomen, and upper thighs, as well as areas of livedo racemosa and noninflammatory retiform purpura with central ulceration (Figures 1 and 2). This retiform purpura was both perilesional and in areas without ulcerations. Laboratory values included the following: sodium, 127 mmol/L (reference range, 136–145 mmol/L); prothrombin time, 16.1 seconds (reference range, 11–15 seconds); white blood cell count, 20.69×10⁹/L (reference range, 4.5–11.0×10⁹/L) with 87% neutrophils (reference range, 54%–62%); hemoglobin, 6.1 g/dL (reference range, 13.5–17.5 g/dL); hematocrit, 18.8% (reference range, 41%–53%); platelets, 474×10⁹/L (reference range, 150–400×10⁹/L); D-dimer, 0.77 mg/L (reference range, ≤0.50 mg/L); fibrinogen, 489 mg/dL (reference range, 150–400 mg/dL); prior urine culture positive for Pseudomonas aeruginosa. He was negative for hepatitis B and hepatitis C viruses as well as HIV, and the lesions were not clinically consistent with herpes simplex virus, as they were not scalloped or circinate. Punch biopsies were obtained from a nodule on the left leg and a purpuric patch on the right leg.

Ischemic cutaneous lesions less commonly occur in the setting of malignancy and can be the result of both direct and indirect systemic effects from the cancer. Malignancies are known to directly trigger vasculopathies in other organs, most commonly the lungs, through 2 primary mechanisms. First, in carcinomatous arteriopathy, metastatic cells promote fibrocellular intimal proliferation of small pulmonary arteries and arterioles leading to stenosis, thrombosis, and obliteration. This mechanism has been described in pulmonary thrombotic microangiopathy secondary to lung carcinoma.⁴ This pathophysiology likely is also what underlies paraneoplastic acral vascular syndromes, which culminate in digital ischemia. Hypothesized mechanisms for this ischemia also range from vasospasm to thromboembolism.⁵ Secondly, in vasculitis carcinomatosa, metastatic tumor cells damage or block vessel walls, resulting in end-organ ischemia. Vasculitis carcinomatosa is a well-known phenomenon in angiocentric and intravascular lymphoid malignancies (typically of B-T or natural killer/T-cell origin) but also has been reported in a case of gastric adenocarcinoma with arterial invasion.⁶ This process is different than carcinoma telangiectoides where malignant cells may be present in the vasculature on histopathology but not trigger thrombosis and ischemic skin necrosis.

Systemic coagulopathies such as disseminated intravascular coagulation (DIC), thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome can occur in the setting of malignancies.⁷ Clinically, all may present with livedo racemosa, noninflammatory retiform purpura, and widespread skin necrosis. In adult patients, purpura fulminans most often is seen in the setting of sepsis and DIC, with accompanying evidence of microangiopathy.⁸ Catastrophic antiphospholipid antibody syndrome can be triggered by malignancy and is characterized by central nervous system, renal, pulmonary, and gastrointestinal complications. Skin involvement such as ulcers, livedo reticularis, and gangrene have been reported.⁹ Other causes of thrombotic vasculopathy include warfarin necrosis, heparin-induced thrombotic thrombocytopenia, calciphylaxis, and angioinvasive infections.⁸ Warfarin necrosis and heparin-induced thrombotic thrombocytopenia typically present days after initiating therapy with the respective medication. Calciphylaxis typically occurs in patients on dialysis, though it may occur in nonuremic patients including those with malignancy.⁸ Patients with malignancies on chemotherapy can become neutropenic and are at risk for ecthyma gangrenosum due to P aeruginosa and other gram-negative rods, Staphylococcus aureus, and angioinvasive fungi.¹⁰

Based on clinical, histopathological, and laboratory data, we favored a diagnosis of cutaneous carcinomatous arteriopathy. Vasculitis carcinomatosa was a possibility despite the lack of vasculotropism on histopathology, which may have been due to biopsy site selection. Systemic coagulopathies such as DIC, thrombotic thrombocytopenia purpura, and catastrophic antiphospholipid antibody syndrome were unlikely, as the ischemic skin lesions and livedo racemosa were limited to areas adjacent to cutaneous metastases, and the patient lacked other common multiorgan manifestations or laboratory findings. Although our patient was on warfarin, he was on a stable dose for weeks and histopathologic features of subcutaneous thrombosis were not seen. The biopsy also was not consistent with calciphylaxis. Ecthyma gangrenosum was unlikely given the lack of organisms on histopathology and negative skin and blood cultures. Although additional laboratory testing in this patient may have included cryoglobulins and cryofibrinogens, both entities were unlikely due to a lack of ischemic acral lesions.

In conclusion, we present a case of localized thrombotic vasculopathy that likely was secondary to cutaneous carcinomatous arteriopathy in the setting of cutaneous metastatic penile squamous cell carcinoma. The differential diagnosis of retiform purpura, livedo racemosa, and other signs of cutaneous ischemia in patients with metastatic cancer is broad and can be the result of both direct and indirect systemic effects from the cancer. Appropriate workup in these cases should include skin biopsies for histopathology and culture, medication review, and laboratory evaluation for systemic coagulopathies.

Practice Gap

Invasive fungal infections are a leading cause of morbidity and mortality among neutropenic, immunocompromised, and critically ill patients. Candida species are the most common cause of fungemia, with portals of entry into the bloodstream including the gastrointestinal tract, contaminated intravascular catheters, and localized foci of infection.¹ Diagnosis of invasive candidiasis remains challenging due to an absence of specific clinical signs and symptoms, varying from a mild fever that is unresponsive to antibiotics to florid sepsis. When present, clinical clues may include chorioretinitis; muscle abscesses; and skin eruptions, characteristically with Candida tropicalis. Cutaneous manifestations of disseminated Candida infections appear in only 13% of affected patients.¹ The lesions typically present as 5- to 10-mm pink dermal papules or painless pustules on an erythematous base and may be singular, localized, or diffuse in distribution. Body regions normally involved are the trunk, arms, and legs, rarely the head and neck.¹ Cutaneous lesions often develop at a time when patients are febrile, are not responding to antibiotics, and are clinically deteriorating.

A 15-year-old adolescent boy with pre–B-cell acute lymphoblastic leukemia was admitted with febrile neutropenia for presumed septic shock secondary to an unknown infectious etiology. The patient was started on broad-spectrum intravenous antibiotics, and blood cultures were obtained. On the second day of hospitalization, he developed approximately 10 to 15 discrete, 3- to 6-mm, pink to violaceous papules scattered on the chest and arms (Figure 1). Over several hours, the number of lesions increased to more than 50 with involvement of the legs (Figure 2). A punch biopsy of lesional skin from the left dorsal wrist demonstrated a circumscribed abscess of yeast in the papillary dermis, which was highlighted by periodic acid–Schiff staining with minimal associated inflammation (Figure 3). Blood and tissue cultures persistently grew C tropicalis. The patient was started on intravenous liposomal amphotericin B but died on day 5 of hospitalization after developing endocarditis.

Early and reliable diagnosis of Candida species fungemia is of critical importance to successful treatment, particularly with the emergence of multidrug-resistant strains such as Candida auris.² In patients with apparent cutaneous manifestations, a lesional punch biopsy for culture and histopathologic evaluation is recommended in addition to blood culture; however, organisms may or may not be present in large numbers, and they may be difficult to identify on routine hematoxylin and eosin–stained tissue sections. To enhance the likelihood of highlighting the fungus within the sample, the pathologist must be made aware of the presumptive diagnosis of disseminated candidiasis so that special techniques can be utilized, such as periodic acid–Schiff stain.

Although positive blood culture is the gold standard for candidemia diagnosis, only 30% to 50% of patients with disseminated candidiasis had positive blood cultures at autopsy.¹ Another study showed the sensitivity of blood culture for the detection of invasive fungal infection to be as low as 8.3%.³ In cases with positive blood cultures, the median time to positivity is 2 to 3 days, but it can take as long as 8 days, thus limiting its clinical utility in acutely ill patients.⁴ Given the low sensitivity and prolonged time required for culture growth of most fungal organisms, novel assays for rapid, non–culture-based diagnosis of systemic fungal infections hold substantial clinical promise moving forward.

The Technique

One of the more promising non–culture-based fungal diagnostic methodologies is an antigen assay based on the detection of serum (1,3)-β-D-glucan (BDG), a major cell wall constituent of most pathogenic fungi. This assay is not specific for Candida species and can be positive for Aspergillosis species, Fusarium species, Coccidioides immitis, Histoplasma capsulatum, and Pneumocystis jirovecii pneumonia, among others; therefore, it functions as a general biomarker for fungi in the bloodstream.^4,5 (1,3)-β-D-glucan assay can be useful as an adjunct for blood cultures and punch biopsy, especially when cultures are negative or the results remain outstanding. The results of the BDG assay are available in less than 24 hours at minimal cost, and the test is approved by the US Food and Drug Administration for use as an aid in invasive fungal disease diagnosis. In a meta-analysis of 11 studies, BDG sensitivity was 75%.⁴ In a study based on autopsy cases from 6 years, BDG specificity was 98.4% with positive and negative predictive values of 86.7% and 97.1%, respectively.³ Optimal results were achieved when 2 consecutive tests were positive.⁴ The serum assay output is based on spectrophotometer readings, which are converted to BDG concentrations (negative, <60 pg/mL; indeterminate, 60–79 pg/mL; positive ≥80 pg/mL).⁵ Although we cannot be certain, utilizing the BDG assay in our patient may have led to earlier treatment and a better outcome.

A disadvantage of the BDG assay is the potential for false-positive results, which have been reported in lung transplant recipients with respiratory mold colonization and patients with other systemic bacterial infections.⁴ False-positive results also have been associated with use of ampicillin-clavulanate and piperacillin-tazobactam antibiotics and human blood products, hemodialysis, and severe mucositis, thus reaffirming the importance of judicious interpretation of BDG assay results by the clinician.^4,6 There also is a potential for false-negative results, as the BDG assay does not detect certain fungal species such as Cryptococcus species and Blastomyces dermatitidis, which produce very low levels of BDG, or zygomycetes (Absidia, Mucor, and Rizopus species), which are not known to produce BDG.⁶

Practice Implications

In the setting of invasive fungal infections, a high degree of clinical suspicion is paramount due to the often subtle nature of cutaneous manifestations. A positive BDG assay can be used to identify high-risk patients for empiric antifungal therapy, prompting early intervention and improved outcomes in these acutely ill patients. The BDG assay’s excellent negative predictive value is useful in ruling out invasive Candida infections and may justify stopping unnecessary empiric antifungal therapy.⁴ For the dermatology hospitalist, incorporation of the BDG assay as a noninvasive screening tool may allow for more rapid initiation of appropriate antifungal therapy while awaiting confirmatory skin biopsy or culture results in disseminated candidemia and other invasive fungal infections.

Practice Gap

Invasive fungal infections are a leading cause of morbidity and mortality among neutropenic, immunocompromised, and critically ill patients. Candida species are the most common cause of fungemia, with portals of entry into the bloodstream including the gastrointestinal tract, contaminated intravascular catheters, and localized foci of infection.¹ Diagnosis of invasive candidiasis remains challenging due to an absence of specific clinical signs and symptoms, varying from a mild fever that is unresponsive to antibiotics to florid sepsis. When present, clinical clues may include chorioretinitis; muscle abscesses; and skin eruptions, characteristically with Candida tropicalis. Cutaneous manifestations of disseminated Candida infections appear in only 13% of affected patients.¹ The lesions typically present as 5- to 10-mm pink dermal papules or painless pustules on an erythematous base and may be singular, localized, or diffuse in distribution. Body regions normally involved are the trunk, arms, and legs, rarely the head and neck.¹ Cutaneous lesions often develop at a time when patients are febrile, are not responding to antibiotics, and are clinically deteriorating.

A 15-year-old adolescent boy with pre–B-cell acute lymphoblastic leukemia was admitted with febrile neutropenia for presumed septic shock secondary to an unknown infectious etiology. The patient was started on broad-spectrum intravenous antibiotics, and blood cultures were obtained. On the second day of hospitalization, he developed approximately 10 to 15 discrete, 3- to 6-mm, pink to violaceous papules scattered on the chest and arms (Figure 1). Over several hours, the number of lesions increased to more than 50 with involvement of the legs (Figure 2). A punch biopsy of lesional skin from the left dorsal wrist demonstrated a circumscribed abscess of yeast in the papillary dermis, which was highlighted by periodic acid–Schiff staining with minimal associated inflammation (Figure 3). Blood and tissue cultures persistently grew C tropicalis. The patient was started on intravenous liposomal amphotericin B but died on day 5 of hospitalization after developing endocarditis.

Early and reliable diagnosis of Candida species fungemia is of critical importance to successful treatment, particularly with the emergence of multidrug-resistant strains such as Candida auris.² In patients with apparent cutaneous manifestations, a lesional punch biopsy for culture and histopathologic evaluation is recommended in addition to blood culture; however, organisms may or may not be present in large numbers, and they may be difficult to identify on routine hematoxylin and eosin–stained tissue sections. To enhance the likelihood of highlighting the fungus within the sample, the pathologist must be made aware of the presumptive diagnosis of disseminated candidiasis so that special techniques can be utilized, such as periodic acid–Schiff stain.

Although positive blood culture is the gold standard for candidemia diagnosis, only 30% to 50% of patients with disseminated candidiasis had positive blood cultures at autopsy.¹ Another study showed the sensitivity of blood culture for the detection of invasive fungal infection to be as low as 8.3%.³ In cases with positive blood cultures, the median time to positivity is 2 to 3 days, but it can take as long as 8 days, thus limiting its clinical utility in acutely ill patients.⁴ Given the low sensitivity and prolonged time required for culture growth of most fungal organisms, novel assays for rapid, non–culture-based diagnosis of systemic fungal infections hold substantial clinical promise moving forward.

The Technique

One of the more promising non–culture-based fungal diagnostic methodologies is an antigen assay based on the detection of serum (1,3)-β-D-glucan (BDG), a major cell wall constituent of most pathogenic fungi. This assay is not specific for Candida species and can be positive for Aspergillosis species, Fusarium species, Coccidioides immitis, Histoplasma capsulatum, and Pneumocystis jirovecii pneumonia, among others; therefore, it functions as a general biomarker for fungi in the bloodstream.^4,5 (1,3)-β-D-glucan assay can be useful as an adjunct for blood cultures and punch biopsy, especially when cultures are negative or the results remain outstanding. The results of the BDG assay are available in less than 24 hours at minimal cost, and the test is approved by the US Food and Drug Administration for use as an aid in invasive fungal disease diagnosis. In a meta-analysis of 11 studies, BDG sensitivity was 75%.⁴ In a study based on autopsy cases from 6 years, BDG specificity was 98.4% with positive and negative predictive values of 86.7% and 97.1%, respectively.³ Optimal results were achieved when 2 consecutive tests were positive.⁴ The serum assay output is based on spectrophotometer readings, which are converted to BDG concentrations (negative, <60 pg/mL; indeterminate, 60–79 pg/mL; positive ≥80 pg/mL).⁵ Although we cannot be certain, utilizing the BDG assay in our patient may have led to earlier treatment and a better outcome.

A disadvantage of the BDG assay is the potential for false-positive results, which have been reported in lung transplant recipients with respiratory mold colonization and patients with other systemic bacterial infections.⁴ False-positive results also have been associated with use of ampicillin-clavulanate and piperacillin-tazobactam antibiotics and human blood products, hemodialysis, and severe mucositis, thus reaffirming the importance of judicious interpretation of BDG assay results by the clinician.^4,6 There also is a potential for false-negative results, as the BDG assay does not detect certain fungal species such as Cryptococcus species and Blastomyces dermatitidis, which produce very low levels of BDG, or zygomycetes (Absidia, Mucor, and Rizopus species), which are not known to produce BDG.⁶

Practice Implications

In the setting of invasive fungal infections, a high degree of clinical suspicion is paramount due to the often subtle nature of cutaneous manifestations. A positive BDG assay can be used to identify high-risk patients for empiric antifungal therapy, prompting early intervention and improved outcomes in these acutely ill patients. The BDG assay’s excellent negative predictive value is useful in ruling out invasive Candida infections and may justify stopping unnecessary empiric antifungal therapy.⁴ For the dermatology hospitalist, incorporation of the BDG assay as a noninvasive screening tool may allow for more rapid initiation of appropriate antifungal therapy while awaiting confirmatory skin biopsy or culture results in disseminated candidemia and other invasive fungal infections.

Practice Gap

Invasive fungal infections are a leading cause of morbidity and mortality among neutropenic, immunocompromised, and critically ill patients. Candida species are the most common cause of fungemia, with portals of entry into the bloodstream including the gastrointestinal tract, contaminated intravascular catheters, and localized foci of infection.¹ Diagnosis of invasive candidiasis remains challenging due to an absence of specific clinical signs and symptoms, varying from a mild fever that is unresponsive to antibiotics to florid sepsis. When present, clinical clues may include chorioretinitis; muscle abscesses; and skin eruptions, characteristically with Candida tropicalis. Cutaneous manifestations of disseminated Candida infections appear in only 13% of affected patients.¹ The lesions typically present as 5- to 10-mm pink dermal papules or painless pustules on an erythematous base and may be singular, localized, or diffuse in distribution. Body regions normally involved are the trunk, arms, and legs, rarely the head and neck.¹ Cutaneous lesions often develop at a time when patients are febrile, are not responding to antibiotics, and are clinically deteriorating.

A 15-year-old adolescent boy with pre–B-cell acute lymphoblastic leukemia was admitted with febrile neutropenia for presumed septic shock secondary to an unknown infectious etiology. The patient was started on broad-spectrum intravenous antibiotics, and blood cultures were obtained. On the second day of hospitalization, he developed approximately 10 to 15 discrete, 3- to 6-mm, pink to violaceous papules scattered on the chest and arms (Figure 1). Over several hours, the number of lesions increased to more than 50 with involvement of the legs (Figure 2). A punch biopsy of lesional skin from the left dorsal wrist demonstrated a circumscribed abscess of yeast in the papillary dermis, which was highlighted by periodic acid–Schiff staining with minimal associated inflammation (Figure 3). Blood and tissue cultures persistently grew C tropicalis. The patient was started on intravenous liposomal amphotericin B but died on day 5 of hospitalization after developing endocarditis.

Early and reliable diagnosis of Candida species fungemia is of critical importance to successful treatment, particularly with the emergence of multidrug-resistant strains such as Candida auris.² In patients with apparent cutaneous manifestations, a lesional punch biopsy for culture and histopathologic evaluation is recommended in addition to blood culture; however, organisms may or may not be present in large numbers, and they may be difficult to identify on routine hematoxylin and eosin–stained tissue sections. To enhance the likelihood of highlighting the fungus within the sample, the pathologist must be made aware of the presumptive diagnosis of disseminated candidiasis so that special techniques can be utilized, such as periodic acid–Schiff stain.

Although positive blood culture is the gold standard for candidemia diagnosis, only 30% to 50% of patients with disseminated candidiasis had positive blood cultures at autopsy.¹ Another study showed the sensitivity of blood culture for the detection of invasive fungal infection to be as low as 8.3%.³ In cases with positive blood cultures, the median time to positivity is 2 to 3 days, but it can take as long as 8 days, thus limiting its clinical utility in acutely ill patients.⁴ Given the low sensitivity and prolonged time required for culture growth of most fungal organisms, novel assays for rapid, non–culture-based diagnosis of systemic fungal infections hold substantial clinical promise moving forward.

The Technique

One of the more promising non–culture-based fungal diagnostic methodologies is an antigen assay based on the detection of serum (1,3)-β-D-glucan (BDG), a major cell wall constituent of most pathogenic fungi. This assay is not specific for Candida species and can be positive for Aspergillosis species, Fusarium species, Coccidioides immitis, Histoplasma capsulatum, and Pneumocystis jirovecii pneumonia, among others; therefore, it functions as a general biomarker for fungi in the bloodstream.^4,5 (1,3)-β-D-glucan assay can be useful as an adjunct for blood cultures and punch biopsy, especially when cultures are negative or the results remain outstanding. The results of the BDG assay are available in less than 24 hours at minimal cost, and the test is approved by the US Food and Drug Administration for use as an aid in invasive fungal disease diagnosis. In a meta-analysis of 11 studies, BDG sensitivity was 75%.⁴ In a study based on autopsy cases from 6 years, BDG specificity was 98.4% with positive and negative predictive values of 86.7% and 97.1%, respectively.³ Optimal results were achieved when 2 consecutive tests were positive.⁴ The serum assay output is based on spectrophotometer readings, which are converted to BDG concentrations (negative, <60 pg/mL; indeterminate, 60–79 pg/mL; positive ≥80 pg/mL).⁵ Although we cannot be certain, utilizing the BDG assay in our patient may have led to earlier treatment and a better outcome.

A disadvantage of the BDG assay is the potential for false-positive results, which have been reported in lung transplant recipients with respiratory mold colonization and patients with other systemic bacterial infections.⁴ False-positive results also have been associated with use of ampicillin-clavulanate and piperacillin-tazobactam antibiotics and human blood products, hemodialysis, and severe mucositis, thus reaffirming the importance of judicious interpretation of BDG assay results by the clinician.^4,6 There also is a potential for false-negative results, as the BDG assay does not detect certain fungal species such as Cryptococcus species and Blastomyces dermatitidis, which produce very low levels of BDG, or zygomycetes (Absidia, Mucor, and Rizopus species), which are not known to produce BDG.⁶

Practice Implications

In the setting of invasive fungal infections, a high degree of clinical suspicion is paramount due to the often subtle nature of cutaneous manifestations. A positive BDG assay can be used to identify high-risk patients for empiric antifungal therapy, prompting early intervention and improved outcomes in these acutely ill patients. The BDG assay’s excellent negative predictive value is useful in ruling out invasive Candida infections and may justify stopping unnecessary empiric antifungal therapy.⁴ For the dermatology hospitalist, incorporation of the BDG assay as a noninvasive screening tool may allow for more rapid initiation of appropriate antifungal therapy while awaiting confirmatory skin biopsy or culture results in disseminated candidemia and other invasive fungal infections.