Conclusions

The diagnosis of sarcoidosis in melanoma patients has been made in a diverse array of clinical settings, including sequential and/or concurrent diagnoses, and following initiation of anti-melanoma therapy. To date, sarcoidosis development has been reported in untreated melanoma patients and those managed with interferon, peptide vaccine, vemurafenib, ipilimumab and nivolumab [5–14, 16–21]. To our knowledge, this is the first reported incident of sarcoidosis with combination ipilimumab and nivolumab immunotherapy.

Recent data suggest an incidence of sarcoidosis in melanoma patients of roughly 0.58%, which is substantially higher than the 0.005% incidence of the general population [16]. Our patient had mediastinal and left hilar lymphadenopathy on baseline scans with FDG-uptake comparable with other lesions. The mediastinal and hilar lymph nodes did not respond to immunotherapy treatment, despite response in all other sites present on baseline evaluation, and the adenopathy progressed with the development of the biopsy-proven cutaneous sarcoid lesions. FNA of the subcarinal lymph node present prior to treatment demonstrated granulomatous inflammation, raising the likelihood that the sarcoid process had already started prior to initiation of immunotherapy.

The clinical manifestations of sarcoidosis are diverse, most commonly with pulmonary and cutaneous findings, and are patient-specific with no pathognomonic pattern [16]. As the clinical findings can vary greatly and can occur long after immune therapy is completed, clinicians must be aware of the possibility of granuloma formation that mimics metastatic disease on imaging and examination. Tissue remains the gold standard for diagnosis.

Sarcoidosis is a multi-system granulomatous disorder of unclear etiology characterized by the presence of CD4+ Th1-like cells that initiate granuloma formation in various tissues [22]. These CD4+ Th1 cells secrete predominantly interleukin-2 and interferon-γ to enhance macrophage TNF-α production which results in a compartmentalized accumulation of multinucleated giant cells and epithelioid cells surrounded by CD8+ T-cells, regulatory T-cells, and fibroblasts [22, 23]. Th17 cells are a subset of CD4+ effector T cells that produce inteleukin-17 and are linked with tissue inflammation and autoimmunity [23–25]. Von Euw et al. demonstrated that treatment with the anti-CTLA4 antibody tremelimumab resulted in an increase in Th17 CD4+ cells in peripheral blood [25]. Patients with irAEs from tremelimumab therapy showed an increase in Th17 cells in their post-treatment blood assessments. In a potential link with anti-CTLA4 therapy and sarcoidosis development, bronchoalveolar lavage samples and lung biopsy specimens from patients with sarcoidosis also contain high numbers of Th17 effector CD4+ cells [23, 24].

PD-L1 staining on the sarcoid tissue from our patient was diffusely positive. In 2014, Braun et al. demonstrated increased expression of PD-1 on CD4+ T cells in both peripheral blood and BAL fluid samples from patients with sarcoidosis [26]. There was also upregulation of PD-L1 in sarcoid granulomas. The authors further showed that blockade of the PD-1 pathway, in vitro, restored the proliferative capacity of sarcoid CD4+ cells to the normal range, a pattern seen in spontaneous clinical resolution of sarcoidosis. Consistent with this, Xu et al. also demonstrated diffuse, strong, membranous positivity for PD-L1 in 100% and PD-L2 positivity in 86% of evaluated sarcoidosis cases [27]. These results suggest a possible therapeutic role for anti-PD-1 therapy in patients with sarcoidosis, though this idea is countered by the cases of sarcoid development in the setting of nivolumab and pembrolizumab therapy. Additional research into these questions will be necessary to fully understand the relative role of these therapeutic agents in the cause and management of sarcoidosis.

In our case, there were multiple variables contributing to the clinical scenario and the driver for the sarcoid progression was unclear. She had received both ipilimumab and nivolumab in combination around the time of the clinical change, but in retrospect also likely had sarcoidosis at baseline. We opted for a break in treatment to allow a period of observation without further immunotherapy or initiation of steroids. In 2016, Johnson et al. detailed the use of ipilimumab in a series of 30 patients with pre-existing autoimmune disease, including two patients with underlying sarcoidosis [28]. One of these patients experienced a sarcoid-exacerbation with hypercalcemia and renal insufficiency managed with steroids and the other patient developed an ocular irAE managed with topical steroids. Our patient had no clinical or radiographic change on observation and after discussion of the risks and benefits we elected to continue with nivolumab. To date, there has been neither change in her lymphadenopathy nor the development of any evidence of progression of sarcoidosis or melanoma. With routine use of combination and sequential anti-CTLA-4 and anti-PD1 therapy in melanoma patients, this clinical scenario will likely be encountered with more frequency and providers will be challenged to decide on appropriate anti-melanoma therapy.

In our patient with known metastatic melanoma, had biopsies not been performed the growth of sarcoid would have erroneously led to the conclusion that disease had progressed, which may have resulted in premature cessation of therapy. The differential for new mediastinal adenopathy and cutaneous lesions in melanoma patients on combination ipilimumab/nivolumab immunotherapy should include sarcoidosis. Biopsy for evaluation of new lesions developing on immunotherapy, particularly in the setting of mixed responses, is favored to help guide decision making. Continued anti-PD1 therapy can be considered for appropriate patients with close observation.