68-2

35/110

Table 6　Results of comparisons of clinical backgrounds and primary clinical outcomes between patients developing PHN and not developing PHN (＊), and univariate logistic analyses performed in an attempt to identify predictors of PHN (#)PHN or non-PHNClinical characteristicsAge (years)PHN (n=29)70.4±12.7M, 9 (31.0)F, 20 (69.0)AMNV, 14 (48.3)FCV, 15 (51.7)Early, 17 (58.6)Late, 12 (41.4)2 (6.9)7 (24.1)6 (20.7)1 (3.4)V, 8 (27.6)C, 4 (13.8)TU, 6 (20.7)TL, 7 (24.1)L, 2 (6.9)S, 2 (6.9)7 (24.1)2 (1, 3)2 (1, 6)3 (2, 6)4 (2, 6)non-PHN (n=245)64.7±16.2M, 111 (45.3)F, 134 (54.7)AMNV, 129 (52.7)FCV, 116 (47.3)Early, 133 (54.3)Late, 112 (45.7)25 (10.2)31 (12.7)43 (17.6)14 (5.7)V, 67 (27.3)C, 48 (19.6)TU, 53 (21.6)TL, 36 (14.7)L, 23 (9.4)S, 18 (7.3)63 (25.7)2 (1, 3)2 (1, 4)1 (0, 3)0 (0, 1)As a result of the univariate logistic regression analysis, none of clinical characteristics examined, including antiherpetic agents used (i.e. AMNV vs. FCV) and the time to the initiations of antiherpetic agent therapy (i.e. early visitors vs. late visitors), were not associated with the development of PHN, although older age and steroid users tended to be associated with a higher incidence of PHN (p = 0.0714 and p = 0.0973, respectively) (Table 6). On the other hand, the development of PHN was significantly associated with some primary clinical outcomes; although the development of PHN was not associated with the pain score at Day 0 or Day 3-4, it was significantly associated with the pain score at Day 7 (r2 = 0.088, p = 0.00006), and even more significantly associated with the pain score at Sex; Males and FemalesAntiherpetic agents,AMNV and FCVEarly and Late visitorsAntibiotic therapySteroid userImmunocompromised stateAntidepressant userRash locationsPrimary clinical outcomesRequirements of NSAIDsNRS pain score at Day 0NRS pain score at Day 3-4NRS pain score at Day 7NRS pain score at Week 2-3Data are shown as Mean ± SD, Median (Interquartile Range), or Number (%), and compared between patients with unpaired t test, Mann-Whitney U test, or Pearson’s chi-square test (＊). Univariate logistic regression analysis was performed to identify predictors of PHN (#).PHN, postherpetic neuralgia; AMNV, amenamevir; FCV, famciclovir; M, males; F, females; V, trigeminal; C, cervical; TU, upper thoracic (T1-8); TL, lower thoracic (T9-12); L, lumbar; S, sacral; NSAIDs, non-steroidal anti-inflammatory drugs; NRS, numerical rating scaleWeek 2-3 (r2 = 0.180, p < 0.00001) (Table 6). The ROC analysis revealed that a cutoff point of the NRS pain score at Week 2-3 for predicting an increased risk for PHN was 0.9 (Figure 6). Based on these results, we dichotomized patients into 88 patients with any pain (rated as ≥ 1 on the NRS) at Week 2-3 and 186 patients with no pain (rated as 0 on the NRS) at Week 2-3, and performed the univariate logistic regression analysis, which revealed the odds ratio (OR), 25.6 (95% confidential interval [CI], 7.5-87.5); area under the curve (AUC), 0.822; r2, 0.257; and a p value < 0.00001. The actual inci-dences of PHN in our patients was 18.3 times higher in patients with any pain at Week 2-3, compared with those with no pain at Week 2-3 (29.5 % [26/88] vs. 1.6 % [3/186], p < 0.00001).p values ＊0.0688 0.1430 0.6554 0.6574 0.5720 0.09060.6766 0.6120 0.8312 0.8540 0.7130 0.3460 0.00003<0.00001p values #0.07140.14760.65570.65770.57470.09730.67710.61590.54060.8540 0.2740 0.1010 0.00006<0.00001127