151 Diagnostic criteria Current international2 and national diagnostic criteria7 emphasize the use of IENFD and/or TTT for a definite diagnosis of SFN. However, experts in the field of SFN and the Dutch guidelines for neuropathy assign superiority of abnormal IENFD instead of equivalence with TTT.7 This is mostly based on the fact that IENFD is considered an objective test, while the interpretation of TTT is more subjective. These guidelines also claim that a normal IENFD does not exclude SFN, emphasizing the lack of specificity for this method and, most importantly, the lack of a good alternative diagnostic method for SFN. Multiple reasons exist why assigning superiority of abnormal IENFD seems unreasonable in our opinion. First, the guidelines were based on studies which included incorporation bias.8 Second, multiple studies including patients with sarcoidosis-associated SFN showed poor sensitivity of IENFD between 28-38%,4–6 while the same studies showed a sensitivity up to 83% for TTT.6 Third, for the general population diagnosed with SFN at a referral center for SFN in a period of 16 years, 90% of patients showed abnormal TTT of which 40% also showed decreased IENFD. Therefore, the added value of IENFD was only for 10% in which no abnormal TTT could be established.9 Finally, IENFD is very challenging to implement because of a complex and labor intensive protocol, which ultimately makes it a very expensive procedure. In our opinion, this does not contribute to cost-effective health-care considering that it only adds value in 10% of all referred patients.9 Alternatives for standard IENFD and TTT which are studied in this thesis are discussed below. Intra-epidermal nerve fiber density: 10 µm sections As alternative for the 50 µm section IENFD according to the EFNS guidelines,3 we investigated a method based on 10 µm sections based on the methods of Koskinen et al.10 According to the EFNS guidelines3 this method showed comparable diagnostic accuracy as 50 µm sections. No significant decreased IENFD in patients with sarcoidosis and SFN compared to patients with sarcoidosis without SFN could be established, see Appendix 1 – Table 1. Therefore, we had to conclude that the 10 µm method was insufficient and could not add value to establish SFN in our cohort. Additionally, in a subanalysis of 5 patients with sarcoidosis and 5 healthy controls, the skin biopsies were analysed in another center using 50 µm sections according to the EFNS guidelines.3 This data was presented in Appendix 1 – Figure 1 and showed no association with our 10 µm IENFD results in the same participants, confirming the lack of diagnostic accuracy for this method. Thermal threshold testing: number of abnormalities This thesis introduced the parameter TTT number of abnormalities (TTT NOAs) as result for TTT measurements.11 The cutoff value for this parameter takes into account the amount of included parameters and measuring sites, which should improve its diagnostic accuracy and uniformity in measurements around the world. Our study revealed that a cutoff value of at least 2 abnormal parameters, based on measuring 6 parameters at each foot and measuring both feet, shows the best diagnostic accuracy for TTT NOAs. However, validation of TTT NOAs should be confirmed with additional studies. Ideally, these validation studies should include a larger sample size and 50 µm skin biopsy according to the EFNS guidelines.3 Contrary to the diagnostic criteria,2 we found that the method of levels in TTT measurements resulted in lower diagnostic accuracy. Moreover, based on the aforementioned results, we believe that adding measurements with pain thresholds improves diagnostic accuracy, which is in line with the DFNS guidelines.1 Table 1 shows that pain thresholds measure other small nerve fiber modalities than detection thresholds. The fact that TTT measures a large spectrum of small nerve fiber functions showed beneficial effects on diagnostic accuracy in comparison with other diagnostic methods.11 10 158 10
RkJQdWJsaXNoZXIy MjY0ODMw