Would you use immunotherapy in a patient with metastatic lung adencoarcinoma, high PDL1, but had polymyositis as a presenting paraneoplastic syndrome?

I personally would not start with a PD-1/L1 inhibitor in this population. Many trials with this class of immune checkpoint inhibitors allowed patients with autoimmune histories to be included if they had not required treatment for a prolonged period of time. Given the potential debilitating nature of polymyositis, I would fear exacerbating this condition. In small cell lung cancer, patients with paraneoplastic syndromes are noted to typically have better prognosis, likely due to enhanced activation of the immune system. It is not clear if PD-1/L1 check point inhibition could safely improve upon this prognosis. In early trials with PD(L)-1 blockade in small cell lung cancer, paraneoplastic-like autoimmune toxicities were seen, and we have seen these in our clinic as well. 

Additionally, recent data presented at ASCO by Arbour et al. suggests inferior response rates to PD(L)-1 blockade in patients who required steroids prior to initiation of thearpy (J Clin ONcol 36, 2018(suppl abstr 9003)). It may not be worth the risk in a patient presenting with a paraneoplastic polymyositis that required steroids. 

Santini et al. (JCO 2017) report on the safety of retreatment after immune-related adverse events and approximately half of patients will develop further toxicities and two patients in their series died from these toxicities though little is known about patients with active paraneoplastic syndromes as these patients would have been excluded from trials.