Clinico-pathological variability and proposal of new biomarkers in central nervous system and muscular sarcoidosis

Sarcoidosis is a multi-system granulomatous disease of unknown etiology characterized by the infiltration of various organs by non-necrotizing granulomas. Disease morbidity is strictly related to the mechanisms that govern granulomatous inflammation, including the release of pro-inflammatory cytokines (mainly IFNγ, IL-2 and TNFα) by immune cells. An important and debated role in sarcoidosis diagnosis is played by serum biomarkers such as angiotensin-converting enzyme (ACE), lysozyme (LSZ), chitotriosidase (CTO) and Krebs von den lungen-6 (KL-6). Neurosarcoidosis is namely the involvement of central and/or peripheral nervous system by sarcoid granulomas. Symptomatic nervous system involvement is reported in 3-16% of patients with sarcoidosis, but post-mortem studies report higher prevalence. Moreover, neurosarcoidosis can present isolated (with no other organ involvement) in about 10% of cases and neurologic manifestations can be the presenting syndrome of sarcoidosis in more than 50% of cases. Finally, neurosarcoidosis can virtually involve every part of CNS and PNS or skeletal muscle with polymorphic clinical presentations. Three main sets of diagnostic criteria have been proposed in the literature to define neurosarcoidosis, all attributing the highest grade of diagnostic certainty to cases with CNS or PNS/muscle pathology demonstrating sarcoid granulomas. Moreover, brain and spinal cord MRI and cerebrospinal fluid (CSF) analysis are the main diagnostic tools taken into account by the aforementioned criteria, but the specificity of their finding is often poor. Therefore, CNS and neuromuscular presentations of sarcoidosis represent a diagnostic challenge for neurologists. Thanks to the cooperation with the Sarcoidosis Regional Referral Center, managed by Respiratory Diseases and Lung Transplantation Unit of the University Hospital of Siena, we were able to evaluate and follow a large number of sarcoidosis patients with suspected or confirmed neurologic involvement. Therefore, we decided to critically analyze the data derived from consultant activity and to perform focused clinico-pathologic studies, in order to gain more insight on the disease and its complications, and to propose new diagnostic tools. Specifically, our study was developed in three main parts: (i) A retrospective cohort study on a population of CNS neurosarcoidosis patients, performed with the aims of identifying recurrent or peculiar clinical patterns and evaluating the possible use of established diagnostic criteria in clinical practice; (ii) A controlled study on serum and CSF biomarkers of CNS neurosarcoidosis including KL-6 (a mucin-like glycoprotein secreted by type II pneumocytes, increased in serum and BAL of patients with ILD and sarcoidosis), which was never systematically investigated in this localization of the disease; (iii) A retrospective/prospective cohort study on sarcoidosis patients undergoing muscular biopsy, in order to define every clinico-pathological presentation and investigate, for the first time, the role of TNFα in sarcoid myopathy. (i). From the retrospective review of the charts from Sarcoidosis Regional Referral Center we identified 22 patients who fulfilled criteria for possible neurosarcoidosis or higher in at least one of the proposed diagnostic criteria. Neurologic symptoms were the presenting feature of sarcoidosis in 59,1% of patients and pathological confirmation of sarcoidosis was available in 17 of them (in one patient it was obtained in CNS biopsy, allowing diagnosis of defined neurosarcoidosis). Lesions involved brain parenchyma, meninges, spinal cord, cranial nerves, hypothalamus/hypophysis and intracranial arteries, and 8 patients had multiple localizations. Most useful investigations for the diagnosis were MRI (positive in 20 patients and showing post-contrast enhancement in 16) and CSF analysis (showing inflammatory findings in 11 out of 13 patients), while classic serum biomarkers such as ACE and lysozyme were increased only in a small proportion of patients; instead, serum chitotriosidase (CTO) was significantly increased in the majority of the tested patients. (ii). From this cohort, we retrospectively enrolled patients for whom samples of serum and CSF had been simultaneously collected and stored at -20°C and were available for further analysis. We enrolled 9 patients with a chronic inflammatory disease, namely multiple sclerosis (MS), and 9 patients with a chronic neurodegenerative disease, namely amyotrophic lateral sclerosis (ALS), as control groups. Serum samples from 9 healthy controls were also collected for KL-6 assay. The four groups were matched for sex and age. Measurable CSF concentrations of KL-6 were detected in 7/9 NS patients but in no ALS or MS patients. In NS patients, CSF concentrations of KL-6 were directly correlated with CSF albumin index, albumin, IgG and total protein concentrations. As KL-6 is a high molecular weight (200 kDa) protein produced outside the central nervous system, it is unable to diffuse in CSF when blood-brain barrier is intact. Instead, when the barrier is damaged as in NS, KL-6 could cross it. This interpretation is confirmed by the direct correlation between KL-6 concentration in CSF and albumin index. Our finding of KL-6 in CSF of NS patients but not in ALS and MS patients suggests that this protein may discriminate this specific phenotype of sarcoidosis with 81,2% sensitivity and 100% specificity. (iii). Regarding skeletal muscle involvement, we performed a retrospective/prospective cohort study, enrolling 29 sarcoidosis patient who underwent muscular biopsy due to suspect muscular involvement. General myopathic changes, with different degrees of inflammation, were detected in almost all patients, while granulomas were detected only in 4/29 patients. Inflammatory changes ranged from massive endomysial infiltrates to minimal perivasal deposits. Upregulation of tissue immunity markers was observed. Electron microscopy confirmed endothelial involvement in most cases. Associated neurogenic changes were also detected in a subgroup. Our findings support the hypothesis of symptomatic muscle involvement in sarcoidosis besides granulomatous sarcoid myositis. Local and circulating cytokines result crucial in Th1 inflammatory response of sarcoidosis and increased circulating TNFα is considered a possible cause of sarcoidosis associated small fiber neuropathy. Therefore, in order to investigate the pathophysiology of non-granulomatous sarcoid related myositic changes and of the common muscle pain complaint in these patients, we investigated by immunocytochemical localization tissue expression of TNFα and its receptors (TNFR1-2) in 9 patients from our cohort and in healthy controls. An endothelial increase and sarcolemmal/sarcoplasmic deposition of TNFα/TNFR1-2 was detected in all cases, with or without granulomas, including samples with minimal pathological or immunohistologic changes. Our findings suggest that TNFα/TNFR1-2 may represent a sensitive diagnostic marker, especially in milder presentations of sarcoid myopathy, and underline the relevance of capillary endothelium as the first site of the pathological process.