81 Table 2 Overview of thermal threshold testing protocol. Order Methods Contralateral side Test Round Contralateral side Full protocol Test side Full protocol 1 CDT & WDT MLi 1 3 3 2 CDT & WDT MLe 1 0 Variable* 3 TSL & PHS 2 6 3 4 CPT & HPT 1 3 3 The first column shows the order in which the measurements were programmed. A test round measured each parameter once and was performed at the forearm of the contralateral site. Next, a full test protocol measured each parameter three times. The full protocol was performed at the contralateral side first, followed by the test side. MLe was not performed for the contralateral side. MLe varies in the number of levels required to determine the detection threshold. Moreover, the hands were always measured first, before the feet. *see Figure 3. CDT = cold detection threshold; WTD = warm detection threshold; MLi = method of limits; MLe = method of levels; TSL = thermal sensory limen; PHS = paroxysmal heat sensations; CPT = cold pain threshold; HPT = heat pain threshold. The Medoc Thermal Sensory Analyzer II (Medoc Ltd., Ramat Yishai, Israel) was used. Thermal stimulation was provided by a 3.0 x 3.0 cm thermode, based on the Peltier effect. Baseline temperature was set at 32 C, with a range of 0-50 C. For detection and pain thresholds, temperature changed at a rate of 1 C/s, while the return to baseline was set at a rate of 5 C/s. A variable time window of 4-6s was set between stimuli, to prevent an automated reaction. For TSL, temperature changed at a constant rate of 0.7 C/s. The SOP file described in the Investigator’s Brochure 20 published by the DFNS was translated into Dutch and used to instruct the participants. Results were transformed from temperature units into z-scores, corrected for age, sex and measuring site, according to the DFNS protocol.13 A positive abnormal z-score indicates increased sensitivity to stimuli (hyperesthesia, hyperalgesia and allodynia), while a negative abnormal z-score indicates a decreased sensitivity to stimuli (hypoesthesia and hypoalgesia). Normative values from the DFNS13 were used to determine abnormality for each parameter. PCA analysis First, principal component analysis (PCA) was applied, to identify the correlation between and discriminative power of TTT parameters. To this end, patients with sarcoidosis were divided into two groups: probable SFN and no SFN, based on the new Besta criteria.6 The group without SFN was clustered with healthy controls. Method of Levels The number of MLe measurements depends on the consistency of the participants’ performance and can vary considerably. For example, when a participant pressed “yes” the first time and “no” the second time while the same temperature level was applied, additional measurements were required to establish the detection threshold. Therefore, the number of levels required to establish the detection threshold was first assessed to gain insight into the variation between participants. Next, since MLe corrects for the reaction time artifact of MLi, MLi was expected to show abnormal findings more often than MLe. Whether this is true was assessed by plotting the distribution of participants classified as normal or abnormal using both methods. Four groups were defined as “both normal”, “MLi abnormal”, “MLe abnormal” and “both abnormal”, respectively, for CDT and WDT. In order to investigate the discriminative power of individual parameters, the percentage of participants with at least one abnormal result for the same parameter measured at both feet was calculated. Results were displayed for three groups: patients with sarcoidosis with probable SFN, patients with sarcoidosis without SFN and healthy controls. The same calculations were performed for four definitions of “abnormal QST”: 5 85 5
RkJQdWJsaXNoZXIy MjY0ODMw