This case is a very unusual presentation of a chronic unilateral keratoconjunctivitis in a child. Although the conjunctivitis does improve with the application of topical steroids, it does not fully resolve, indicating an immune-mediated process that could be either infectious or noninfectious. It is important to consider a differential diagnosis for both arms.
Given the indolent nature of the inflammatory response, a virus or atypical mycobacterium top the list of potential infectious causes. In any case of unilateral keratoconjunctivitis, a herpes simplex virus infection should always be placed first on the differential, and it would be reasonable to try acyclovir orally (10 mg/kg to 20 mg/kg) to determine whether there is any response. Although the previous cultures turned out negative, virus and atypical microorganisms are difficult to grow. Obtaining a surgical tissue biopsy and sending the sample for a full infectious polymerase chain reaction (PCR) panel would result in the most accurate infectious diagnosis. If a specific organism is identified, focused treatment can be administered based on the pathogen. However, given the lack of infectious animal contacts and a lack of lymphadenopathy, an infectious etiology seems unlikely.
Most noninfectious autoimmune conditions causing a conjunctivitis generally present with an accompanying systemic condition, either rheumatologic or neoplastic. The first step would be to do an in-depth systemic review with the patient and obtain a detailed family history of autoimmune diseases and malignancies. Systemic symptoms can help narrow the final diagnosis. Underlying leukemia, a common pediatric cancer, can also give atypical presentations of both infectious and noninfectious conjunctivitis. Basic blood work, including a complete metabolic panel and complete blood count are reasonable starting screens for hematologic cancer. A QuantiFERON-TB Gold (Qiagen N.V.) and rapid plasma reagin blood tests that rule out infections tuberculosis and syphilis are essential before starting any systemic immunosuppression.
Given this child’s clinical picture, an underlying rheumatologic condition and autoimmune conjunctivitis seems the most likely diagnosis. Although not classic in appearance, some of the lesions look as if they could be granulomas. A noninfectious granulomatous inflammation such as sarcoidosis comes to mind; however, another immunologic diagnosis from the tissue pathology is equally as likely. At the time of biopsy, multiple samples should be sent in formalin for infectious PCR and in fresh tissue for immune analysis. Given the chronicity of the condition, waiting for results of the conjunctival biopsy for a more definitive diagnosis before starting any systemic treatment seems prudent. Whether the etiology is infectious, neoplastic, or autoimmune, addressing the underlying systemic condition would be the treatment. If the final answer is autoimmune, then starting this patient on systemic steroids and immunosuppression with the help of a rheumatologist would be the next step.