Introduction

Hospitalized patients frequently need multidisciplinary care. Dermatologists may play a pivotal role in managing inpatients, usually more complicated and challenging than many outpatients (1).  A study examining 591 dermatology consultations showed that 78% of the inpatients were misdiagnosed by non-dermatologists (2). Dermatology consultations change the diagnoses made by the primary team in more than 60% of the patients and affect the treatment plan of most (1,3-5). Despite some discrepancy among the published studies, the most frequent requesting service is internal medicine, while hematology and oncology are among the most common subspecialties (4,6).

However, dermatology is among the top five consultations requested from oncology clinics (7). Due to underlying diseases, chemotherapy, stem cell transplantation, and multi-drug treatments, immunosuppression makes this group prone to cutaneous conditions (7,8). Recently, there have been promising advances in cancer therapeutics with many new treatment agents, including immune checkpoint inhibitors, associated with cutaneous immune-related side effects (7-10). Therefore, there is an increasing need for closer collaboration between hematology, oncology, and dermatology departments (7,8).

Few studies have described the characteristics of skin findings in detail in hematology and oncology patients so far. In this study, we aimed to 1) categorize the frequent dermatological diseases in hospitalized patients with malignancy, 2) analyze the role of dermatologists in diagnosing associated dermatological conditions, and 3) demonstrate the effect of dermatologists in the treatment strategies of patients.

Methods

This was a retrospective, cross-sectional study conducted using the medical records of patients diagnosed with hematologic and oncological malignancies who were referred to the dermatology clinic for cutaneous manifestations. The study included inpatient admissions between January 2018 and March 2021. Patients under the age of 18, not hospitalized, or with insufficient medical history were excluded. The study was approved by the University of Health Sciences Turkey, Gülhane Training and Research Hospital Ethics Board (decision no: 2021/123, date: 06.01.2022).

We used the data from the Department of Dermatology and Venereology of the University of Health Sciences Turkey, a tertiary referral care hospital that provides consultant services for inpatients and outpatients. Sociodemographic and medical information, including cancer characteristics (cancer type, current therapeutics) and dermatological findings, were collected from the electronic medical records and patient charts.

Cutaneous conditions evaluated

Cutaneous conditions were classified into ten groups according to etiology:

Skin biopsy results following the dermatologist’s evaluation were recorded. Recommendations by the dermatologist and whether the consultation changed the patient’s immediate treatment plan were also analyzed.

The accuracy of the prediagnoses, if any, specified by the primary care team was also evaluated.

Additionally, we evaluated the effect of Coronavirus disease-2019 (COVID-19) on dermatology consultations requested for hematology and oncology patients.

Statistical Analysis

Statistical analyses were performed by Statistical Package for the Social Sciences for Windows version 22.0 (IBM, Armonk, NY, USA). Categorical variables were presented by frequency and percentage. Normally distributed continuous variables were expressed as mean±standard deviation. Non-normally distributed continuous variables were expressed as median (interquartile ranges, minimum-maximum).

Results

Demographics

A total of 2120 inpatient referrals were made during the study period. Of these, 265 (12.5%) were from the hematology and oncology clinics. Dermatology consultations were ordered most commonly from the internal medicine clinic (26.4%), followed by the intensive care unit (9.5%) and surgical clinics (8.3%). Among the internal medicine clinics, the highest number of referrals were from the hematology and oncology clinics (12.5%), followed by rheumatology (6.5%) and nephrology (4.7%).

The number of total consultations requested by years was 42 in 2018 (21%), 45 in 2019 (22.5%), 81 in 2020 (40.5%), and 53 in the only first three months of 2021 (26.5%). About half of the consultations (48%) (n=96) were requested during the COVID-19 pandemic, corresponding to the last 12 months of the 39-month study period. Of the 265 hematology-oncology consultations requested in total, 65 were follow-up consultations of the same patients. Of these patients, 37.5% had at least one medical comorbidity (Table 1).

Characteristics of underlying malignancies

Various hematologic (n=113) or solid organ malignancies (n=87) were encountered (Table 1). Distant organ metastases were present by 7.5%. The overall mortality rate was 2% (n=4). Among the deceased patients, orolabial herpes (n=2), contact dermatitis, and cutaneous metastasis of testis cancer were the dermatological diagnoses following consultations. These patients died within one week of the consultation request.

Role of the dermatology consultation

In this study, 47 different dermatological diagnoses were made in 10 major disease categories. All consultations were evaluated within 8 h of the consultation request.

The most common diagnoses were cutaneous infections (33.5%) and drug eruptions (13.5%), followed by dermatitis (11.5%), cutaneous metastasis (7.5%) (Figure 1a, 1b), xerosis and pruritus (5.5%), steroid-induced acne (5%), deep vein thrombosis and thrombophlebitis (%3), chemotherapy skin side effects (%3) (Figure 1c, 1d), GVHD (2.5%), and miscellaneous skin disorders (11.5%) (Table 2).

Drug reaction was the second most common diagnosis. Steven Johnson syndrome (SJS)-toxic epidermal necrolysis (TEN) overlap (Figure 1e) due to the use of phenytoin-valproic acid and fluconazole was observed in two patients with lung and bladder cancers, respectively. Urticarial drug reaction (n=1) and symmetrical drug-related intertriginous and flexural exanthema (n=1) were recorded following administration of ciprofloxacin and piperacillin-tazobactam, respectively. The use of multiple antibiotics (≥3 antibiotics simultaneously) was recorded in three patients diagnosed with fixed drug eruptions. In 8 (40%) patients with maculopapular drug eruptions (n=20), simultaneous dual broad-spectrum antibiotic use was recorded; a combination of meropenem-teicoplanin was the most common suspicious dual-antibiotic (n=4). Meropenem (n=7), teicoplanin (n=5), piperacillin-tazobactam (n=3), and trimethoprim-sulfamethoxazole (n=3) were the most common individual suspicious antibiotics leading to a maculopapular (morbilliform) drug eruption. The majority (73.9%) of patients with maculopapular drug eruption had a hematological malignancy twice as common in patients with solid organ malignancies (13.3% vs. 5.7%).

Hematology and oncology services mentioned a suspicious diagnosis of their own in 34.5% (n=69) of dermatology consultations but no preliminary diagnosis in 65.5% of the patients. The dermatology team confirmed the preliminary diagnosis by the hematologist or medical oncologist in 70.1% of 69 patients. The highest diagnostic accuracy was observed in GVHD and cutaneous side effects of cancer therapeutics (100%). However, stasis dermatitis was consulted as cellulitis in 30% of the cases, and 44.4% of the cutaneous metastases were mistaken for zona zoster or folliculitis. Steroid-induced acne, which can be observed in malignancy patients receiving high-dose steroids according to the treatment schedule, was the least accurately recognized dermatological diagnosis by primary physicians (Table 3).

Evaluation of the histopathological results

Histopathological examination was used in 19.5% (n=39) of the patients to confirm the dermatological diagnosis. Dermatology consultations led to changes in the medical management of 63 patients (31.5%) through the modification of antibiotic treatment (9%), current chemotherapy sessions (17.5%), or chemotherapeutic options and treatment plans (5%). Treatments recommended for the medical management of patients are listed in Table 4.

Discussion

In this study, the most common cutaneous disorder among dermatology inpatient consultations in hematology and oncology clinics was cutaneous infections followed by drug eruptions. Since oncology and hematology inpatients are immunosuppressed and predisposed to neutropenia due to cancer therapeutics, they are more susceptible to mucocutaneous infections and drug reactions due to multiple drug use and prophylaxis schemes (8-13). Although GVHD (2.5%) and cutaneous side effects of chemotherapeutics (3%) have been commonly reported in previous studies (8,10-13), they were rarely observed in our study. The current study included only the patients consulted in the dermatology department; therefore, all inpatients in hematology and oncology clinics were not examined. Since our hospital is a tertiary referral center for bone marrow transplantation and cancer treatment, hematologists and oncologists might be more familiar with skin conditions specifically related to their specialties, such as GVHD and the side effects of chemotherapeutics. Therefore, they might have managed some other patients without the need for a dermatology consultation. The correct preliminary diagnoses for these two dermatological diseases also support our suggestion.

The most common diagnoses in our study were cutaneous infections. More than half of them did not relate to any significant risk of mortality and morbidity, such as localized fungal or bacterial infections. The most frequently observed subgroups were zona zoster, cellulitis, and tinea unguium. Interestingly, these infections are some of the essential topics of dermatology education in medical school, which could be better recognized by the primary team and primarily described in the consultation note. Nevertheless, the diagnostic accuracy of cellulitis was relatively low by the primary physicians. Stasis dermatitis was present approximately in one of every three patients consulted dermatology with the diagnoses of cellulitis. One study at two separate centers showed that 20-35% of patients admitted for cellulitis by the emergency department were mistakenly diagnosed as cellulitis by physicians other than dermatologists and infectious diseases specialists (14). Although stasis dermatitis is a clinical mimicked of cellulitis, bilateral, chronic, and non-tender erythema with usually long-standing pitting edema mainly indicates stasis dermatitis (1,15). A history of trauma in the affected area or accompanying tinea pedis or unguium with tenderness, warmth, swelling, and indistinct borders in a unilateral erythematous plaque, points to cellulitis. Additionally, rapid onset, presence of systemic symptoms such as fever, immunosuppression, and satisfactory response to antibiotics are findings in favor of cellulitis. However, in indistinguishable cases, Doppler ultrasonography could be considered (15,16). These observations suggest that postgraduate dermatology training programs help physicians better assess and manage patients with underlying malignancies and build more proper collaboration with dermatologists.

Drug reactions are another common dermatological condition with a higher risk of morbidity and mortality in patients with underlying malignancies. The frequency of adverse drug reactions in hematology consultations varies between 13 and 38% (8,17,18). In our study, drug reactions were recorded in 13.5% of 200 patients, and most of them were morbilliform drug eruptions (10%) in line with previous literature 9-17.3% (8,10,18). As in our study, most morbilliform drug reactions were uncomplicated conditions. Morbilliform or maculopapular drug eruptions usually present 1 to 2 weeks after initial exposure of the causative drug, may also show up sooner on rechallenge. Although most cases regress within 1-2 weeks, dermatology consultation is critical in confirming the diagnosis and excluding imitators such as viral exanthema seen in the immunosuppressive patient group (19,20). A dermatologist can help determine the suspected drug(s). More importantly, dermatologists may help to recognize the early signs of severe cutaneous drug reactions (SJS and TEN) such as mucosal involvement, skin tenderness, blistering, dusky red and coalescent macular exanthema, atypical target lesions, and Nikolsky sign (ready removal of the epidermis with slight unrelated pressure) (20,21). Therefore, dermatology consultations should be considered in the early period for patients with diffuse maculopapular eruptions.

The current study demonstrates the possible dermatological diagnoses encountered by physicians working with patients with hematology and oncological malignancies. Although 47 different dermatological diagnoses were made in 10 main disease categories, most patients (72.5%) were managed with topical therapy, and only 39% required systemic medications. These data suggested that most skin conditions could be governed without aggressive treatments.

This study presents the 39-month experience of an advanced tertiary center. The last 12 months of the study period corresponds to the COVID-19 pandemic, which has rapidly affected the world and caused severe morbidity and mortality. While our center was mainly in charge of managing patients with COVID-19 infection in this period, hematology and oncology services were at the forefront of the departments that tried maintaining inpatient service properly. In this period, dermatology consultations are essential due to various skin manifestations of COVID-19 infections including urticarial, purpuric, erythema multiforme-like, chickenpox-like rash, acro-ischemia, chilblain-like eruptions, and pityriasis rosea like eruptions (22-25). Intense anxiety among patients and physicians and cutaneous side effects of several systemic medications are likely to increase dermatology consultations during the pandemic period. All these factors are potential causes of the increasing number of dermatology consultations in hematology and oncology inpatients during the pandemic period.

The retrospective design and lack of long-term follow-up are the main limitations of this study. The major strength is documenting a three-year experience of a large population from an advanced tertiary center.

Conclusion

In conclusion, the current study indicates that dermatology inpatient consultations of hematological or solid organ malignancies play an essential role in determining skin manifestations associated with underlying malignancy or treatment of the disease. A close collaboration between hematologists, oncologists, and dermatologists helps manage patients with underlying malignancies appropriately. 

Ethics

Ethics Committee Approval: The study was approved by the University of Health Sciences Turkey, Gülhane Training and Research Hospital Ethics Board (decision no: 2021/123, date: 06.01.2022).

Informed Consent: Retrospective study.

Peer-review: Externally peer-reviewed.

Authorship Contributions

Concept: P.E., S.N.Ç., G.A., Design: P.E., S.N.Ç., G.A., Data Collection or Processing: P.E., S.N.Ç., G.A., Analysis or Interpretation: P.E., S.N.Ç., G.A., Literature Search: P.E., S.N.Ç., G.A., Writing: P.E., S.N.Ç., G.A.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.