Bloodstream Infection in an Immunocompromised Host After Unusual Presentation of Photo Recall Dermatitis

Introduction

Photo recall dermatitis (PRD) is a phototoxic inflammatory reaction, a subset of radiation recall dermatitis (RRD), that occurs in areas of prior UV-damaged skin following systemic administration of certain medications, particularly antineoplastic agents such as gemcitabine and docetaxel [1]. This phenomenon causes a pattern of reactivation in previously injured skin. The most common malignancies associated with RRD are breast and gastrointestinal malignancies [2,3]. While the pathophysiology of PRD or RRD is not fully understood, reports show that when ultraviolet radiation damages DNA in the skin, there is a local UV-induced immunosuppression. This is likely mediated by the release of immunomodulating cytokines, which promote more T regulatory cells at the location of the injury, contributing to local immunosuppression [4]. A recall reaction seems to require a cytotoxic agent that acts by various mechanisms to cause radiosensitization, in this case in the skin, leading to an exaggerated reaction when that tissue is exposed to a certain amount of UV or other types of radiation [5]. Studies suggest the reaction is most intense if the interval between chemotherapy agents (actinomycin D in this study) and radiation is brief [6].

Coagulase-negative staphylococci (CoNS) are common skin commensals often dismissed as blood culture contaminants. In immunocompromised patients, particularly those with malignancy and neutropenia, they may represent true bloodstream infection. Despite this biologic plausibility, bloodstream infection complicating photo recall dermatitis has not been previously described. CoNS other than S. epidermidis (NonSE-CoNS) account for about 33% of CoNS blood stream infections. In a study of 252 cases, NonSE-CoNS were statistically associated with malignancy and immunosuppression [7]. Acute inflammatory reactions such as photo recall dermatitis can disrupt the skin barrier, potentially increasing the risk of secondary infection. The purpose of this report is to address this gap in the literature by presenting a case of photo recall dermatitis complicated by bacteremia.

Case Report

The infectious disease service was consulted on a White woman in her 80s with a past medical history of locally advanced pancreatic adenocarcinoma first diagnosed in 2021. Her oncologic treatment included 7 cycles of FOLFIRINOX (5-fluorouracil, leucovorin, irinotecan, and oxaliplatin) from November 2021 to February 2022, complicated by neuropathy and cytopenias. She received 5 consecutive days of radiation treatments to her pancreatic mass in November 2021, then later underwent a distal pancreatectomy, splenectomy, and portal vein resection in June 2022. She developed local recurrence and lung metastases after the surgery, followed by capecitabine between November 2022 and July 2023, but the lung mass grew while she was receiving therapy, so she was treated with gemcitabine alone from August 2023 to November 2023, at which time nab-paclitaxel was added to the regimen, and she was on both medications from November 2023 to January 2024. In January 2024, chemotherapy was stopped due to continued progression in a dominant lung nodule. In March 2024 she underwent a wedge resection of the right middle and right lower lobes for metastases. She then experienced further disease progression with enlarging nodules in the right lung, so in June 2024 she was enrolled in a clinical trial evaluating a DNA methyltransferase-inhibiting nucleoside analog drug. On June 24, she started receiving 16 mg daily oral 5-aza-4′-thio-2′-deoxycytidine (Aza-TdC), which was given from June 24 to 28, followed by a 2-day break (per protocol), then daily doses from July 1 to 5. On July 6 she called the clinic to report new red and pruritic rashes on the soles of her feet, which cleared up with topical hydrocortisone. On the first in-person follow-up visit, on July 16, she was noted to have grade 3 neutropenia with an absolute neutrophil count (ANC) of 660 cells/μL, so her study drug was held.

On Monday, July 22, 2024 she was seen by Dermatology after the rash recurred on July 19, 2024. She reported that 2 weeks prior she developed a painful sunburn while sitting at her pool without sunscreen. Upon evaluation by Dermatology, she reported that her new rash seemed to be in an identical distribution to her recent sunburn (mid-thighs, superior lower legs, upper arms, and around the neck). Her rash was warm, sensitive, and mildly pruritic. She reported that over that weekend she had felt lethargy, fatigue, and fevers associated with rigors. In the clinic, she was found to be febrile to 38.1°C and neutropenic with an ANC of 700 cells/μL, so she was admitted to the hospital for expedited treatment and workup.

On admission, she had a body temperature of 38.2°C, a heart rate of 76 beats per minute, blood pressure of 124/47 mmHg, and oxygen saturation of 96% on room air. On exam, she was not in distress, with no sores in the mouth, but she had dry oral mucosa and redness to bilateral eyes and conjunctiva. The lungs were clear to auscultation. Her right chest wall port-a-catheter site examination found no tenderness, warmth, erythema, or swelling. Her cardiac and abdominal exam results were benign. Her skin exam showed an erythematous, blotchy, blanching rash on the lateral left arm (Figure 1A) and the left side of her bilateral thighs and knees (Figure 1B). The entire rash was limited to areas of sun-exposed skin.

She had no recent travel, had not eaten any unusual or raw foods, and had no ill contacts. She had had a right chest port for approximately 2 years, but it had not been accessed recently as her medications were all orally administered. She had no history of central-line bloodstream infections (CLABSI) or issues relating to her port. She had no history of skin infections but reported 6 years before this admission she had a positive nasal screening test for methicillin-resistant Staphylococcus aureus (MRSA). A respiratory pathogen panel polymerase chain reaction (PCR) was negative for all 23 pathogens tested. In the setting of a neutropenic fever with a significant rash, vancomycin 1000 mg every 12 hours and ceftazidime 2 g every 8 hours were started empirically. Granulocyte-colony stimulating factor (G-CSF) was also started, which led to a rapid improvement in her neutropenia. Peripheral blood cultures on admission grew a non-aureus species of Staphylococcus, while blood cultures from the port-a-catheter were negative. Staphylococcus hominis was subsequently identified in 1 of 4 peripheral blood culture bottles. After starting intravenous antibiotics, she had significant diaphoresis, and had resolution of her fever after 3 days of antibiotics. Despite only 1 bottle growing S. hominis, she was considered to have a true bacteremia due to the timeline of her rash, leading to classic symptoms of bacteremia, followed by an expected response to appropriate antibiotics. Due to the S. hominis susceptibility data (Table 1) and resolution of neutropenia, the vancomycin and ceftazidime were stopped, and oral cephalexin 500 mg 3 times daily was started. Her repeat blood cultures remained negative.

During her hospital stay, she was examined by the Dermatology service, who diagnosed the rash as photo recall dermatitis due to it recurring in a similar distribution to her sunburn, and they recommended topical triamcinolone. The rash significantly improved on topical steroids.

She was discharged on July 26, 2024 in her usual state of health on oral cephalexin 500 mg 3 times daily to complete a 14-day course for an uncomplicated bloodstream infection presumed to have entered her bloodstream through a break in her skin, given her rash. She was treated for uncomplicated blood stream infection from a skin source and not a central–line-associated BSI because her central catheter blood cultures were negative. The preponderance of evidence suggested a skin source because the sensitivity of culture or microbiologic confirmation of skin infections has low sensitivity [8], yet skin infections are common in neutropenic patients [9]. The challenging and interesting part of this case was that her rash represented photo recall dermatitis from sun exposure while on an experimental chemotherapy agent. The patient chose to withdraw from further participation in the clinical trial and continues to be monitored for adverse events but has not had any further infectious complications.

Discussion

Photo recall dermatitis is a phototoxic inflammatory reaction occurring in areas of prior ultraviolet-induced injury after exposure to systemic medications, most commonly chemotherapeutic agents such as methotrexate, gemcitabine, and docetaxel [10]. Although it resembles radiation recall dermatitis, it is triggered specifically by prior UV exposure rather than ionizing radiation. The reaction typically occurs days to weeks after a sunburn and within days to 1 week after administration of the systemic agent, often with overlapping timing of exposures [11]. Clinical manifestations usually include erythema, edema, desquamation, urticaria-like lesions, and, occasionally, vesiculation. Management centers around topical or systemic corticosteroids. It rarely recurs with rechallenge of the suspected medication and seems to be a phenomenon of the timing of exposure to both UV and the systemic agent [12].

The pathophysiology is not fully understood, but prior UV injury is believed to result in persistent local immune dysregulation. In immunocompromised patients – particularly those with chemotherapy-induced neutropenia – such inflammatory skin reactions can have additional clinical implications. Radiation and photo recall reactions have been reported with numerous chemotherapies and, less commonly, antibiotics [2,12–14]. However, no prior reports describe this reaction pattern with hypomethylating nucleoside analogs, making this presentation clinically important.

Coagulase-negative staphylococci (CoNS) are frequently dismissed as blood culture contaminants, particularly when only a single bottle is positive. However, they can cause clinically meaningful infections in immunosuppressed hosts when supported by compatible clinical features and response to therapy. S. hominis is most often associated with central–line-associated bloodstream infections (CLABSI), but attribution to skin sources cannot be confirmed without site cultures or molecular comparison [15].

Few published cases have described superimposed infection complicating photo recall dermatitis or radiation recall. A previous report described the case of a 44-year-old patient with cholangiocarcinoma who had received radiation therapy then gemcitabine, cisplatin, and nab-paclitaxel before presenting with arm swelling, erythema, leukocytosis to 29 000/μL, a temperature of 37.9°C, and 1 of 2 blood cultures bottles positive for Staphylococcus hominis. As in the present case, there was concern that this represented a contaminant of the blood cultures, but unlike our case, this patient had only a partial response to antibiotic therapy alone (daptomycin and cefepime) and their patient eventually responded to steroids [16]. Prior reports of S. hominis bacteremia in this setting remain inconclusive due to the high likelihood of culture contamination when only 1 bottle is positive.

In contrast, our patient had true rigors, soaking diaphoresis, neutropenia, and subsequent clinical improvement with antimicrobial therapy and hematologic recovery. The skin could be considered a potential source since there was not biopsy-proven invasion at the site of the RRD. The temporal relationship between rash recurrence, systemic symptoms, and response to antibiotics supports the interpretation of true bacteremia rather than contamination.

To the best of our knowledge, no prior case reports have documented a confirmed bloodstream infection complicating photo recall dermatitis in a neutropenic patient. While prior reports describe recall reactions mimicking infection, documented bacteremia temporally associated with photo recall dermatitis has not been established. This case therefore may represent the first published report of Staphylococcus hominis bloodstream infection complicating photo recall dermatitis, expanding the recognized clinical spectrum of this dermatologic entity.

Conclusions

This case highlights several important clinical considerations. First, clinicians should recognize that photo recall dermatitis can occur in immunocompromised patients receiving chemotherapy or investigational agents and can present with erythema localized to previously sun-exposed areas. Second, in neutropenic patients with inflammatory skin reactions accompanied by fever or rigors, bloodstream infection should be actively evaluated and not automatically attributed to contamination, even when CoNS species are isolated in just 1 culture bottle. Management should include antibiotics adequate for neutropenic fever coverage in addition to coverage skin flora like Staphylococci, including CoNS and/or MRSA, dermatological evaluation, and treatment of inflammatory reactions as appropriate. While coagulase-negative Staphylococci (CoNS) are colonizers of the skin and commonly cause contamination in blood cultures, in the correct clinical scenario they can represent true infections and require appropriate treatment.