1562. Impact of Skin Biopsy on Diagnosing Infections and Changing Treatment in Cancer Patients with New Skin Rash
Session: Poster Abstract Session: Viruses and Bacteria in Immunocompromised Patients
Friday, October 5, 2018
Room: S Poster Hall
Posters
  • IDWeekposterfinal5.pdf (425.6 kB)
  • Background:

    Skin lesions in immunosuppressed cancer patients have a broad differential of infectious and non-infectious causes. Rash may be an early indication of serious systemic infections that are otherwise difficult to diagnose; hence, skin biopsy with culture and histopathology plays a vital role in establishing a diagnosis. Our study aims to determine the yield of skin biopsy in identifying infections and its impact on diagnosis and therapy.

    Methods:

    We performed a retrospective review of all cancer patients admitted to University of Maryland from August 2010 to October 2017 who had a skin biopsy for new rash. We classified the skin lesion as infectious if the biopsy pathology or culture showed a pathogenic organism.

    Results:

    Of 269 patients biopsied for new skin lesions, 43 (16%) were caused by infection and 226 (84%) were non-infectious. Among non-infectious causes, 63% were due to graft vs. host disease, 9% cancer, 9% drug reaction, 4% Sweet syndrome, and 29% were non-diagnostic. The median WBC count trended toward significantly lower in the infectious group (1100/mcL) vs. the non-infectious group (2700/mcL; p=0.08).

    Of the 43 infectious lesions, 21 (49%) were fungal, 13 (30%) bacterial, 7 (16%) viral and 1 (2%) mycobacterial. 67% patients had absolute neutrophil counts <1000/mcL, 40% were febrile, and 28% had had a stem cell transplant. The majority of infections (58%) were identified by skin biopsy alone.

    Change in diagnosis after biopsy was significantly more likely in patients with infectious cause of skin lesions than non-infectious (47% vs. 28%, respectively, p<0.02). Patients with a biopsy-confirmed infectious cause were 5 times (95% CI 2.70-10.22) more likely to have a change in therapy post biopsy compared to patients with a non-infectious cause. The sensitivity and specificity of provider diagnosis prior to biopsy was 86% and 81%, respectively. The positive predictive value of pre-biopsy provider diagnosis was low at 46%.

    Conclusion:

    Skin biopsy of new rash in immunocompromised cancer patients frequently reveals systemic infections (especially fungal) and often leads to a change in diagnosis and therapeutic management.

    Niyati Jakharia, MD, Infectious Diseases, University of Maryland Medical Center, Baltimore, MD, Kristen Stafford, PhD, MPH, Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD and David J. Riedel, M.D., M.P.H., Infectious Disease, Institute of Human Virology and University of Maryland School of Medicine, Baltimore, MD

    Disclosures:

    N. Jakharia, None

    K. Stafford, None

    D. J. Riedel, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.