116 Figure 2 Inclusion process; schematic overview of inclusion and clinical diagnosis of SFN in this study. A total of 79 participants provided informed consent, of whom six were excluded. One participant showed low vitamin B12 levels, four patients were excluded based on low nerve conduction velocity, and one was diagnosed with neurosarcoidosis. Of the 74 patients with sarcoidosis and symptoms of SFN, 48 were diagnosed with probable SFN. Abbreviations: SFN = small fiber neuropathy Data analysis First, the prevalence of patient-reported length-dependent and non-length-dependent pain was determined without distinction between intermittent and continuous pain. Next, the prevalence of patient-reported intermittent and continuous pain was determined by categorizing 4 phenotypes: 1. Intermittent length-dependent pain 2. Intermittent non-length-dependent pain 3. Continuous length-dependent pain 4. Continuous non-length-dependent pain The presence of patient-reported length-dependent and non-length-dependent presentation was compared between intermittent and continuous symptoms. Statistical significance was calculated using odds ratios (OR) and chi-square tests. Patients could have both intermittent and continuous pain at the same time, but for one type of pain (intermittent or continuous), the presentation could be length-dependent or non-length-dependent. Finally, the association between diagnostic methods and patient-reported pain was examined. Therefore, the correlation between the four phenotypes, the SFNSL questionnaire, TTT, IENFD, CCM, Sudoscan and water immersion skin wrinkling (WISW) was assessed. Quantitative Sensory Testing The Medoc Thermal Sensory Analyser 2 (TSA2) was used to assess sensory nerve function. TTT was performed on both feet to assess small nerve fibers. The German Research Network on Neuropathic Pain developed a standard operating file24 that was used to instruct participants. Moreover, their cutoff values were used to define abnormal parameters.25 The number of abnormal TTT parameters (TTT NOAs) was used as parameter for further analysis.22 Skin biopsy Three mm punch biopsies were collected 10 cm above the lateral malleolus of the right leg to determine IENFD. For automatic PGP9.5 staining, 10 µm section were used as previously described.26 According to the guidelines of the European Federation of Neurological Society (EFNS),27 this method shows comparable results to the recommended method based on 50 µm sections. Four consecutive sections separated by 100 µm were analyzed with a light microscope at 400x magnification. Because high-resolution scans are currently available, we counted all nerve fibers in the entire sample. The IntelliSite Image Management System (IMS v.1.8.6824, Philips Healthcare, Amsterdam, the Netherlands) was used to calculate the entire surface of the epidermis. We counted all nerve fiber fractions with a minimal length of 5 µm.28 When fractions were aligned on the same trajectory, they were counted as one nerve, as long as the distance between fractions did not exceed the length of the longest nerve fraction in that trajectory. Corneal Confocal Microscopy The Heidelberg Retina Tomograph III with Cornea Rostock Module (Heidelberg, Germany) was used to image corneal nerve fibers. This allowed a corneal surface of 400 x 400 µm to be photographed with a resolution of 384 x 384 pixels. Corneal imaging was performed according to a publicly available protocol.29 CNFD, cornea nerve branch density (CNBD), and cornea nerve fiber length (CNFL) were 7 122 7
RkJQdWJsaXNoZXIy MjY0ODMw