甲狀腺素轉運蛋白的介入治療對於遺傳性類澱粉沉積症神經病變之療效和安全性：系統性文獻回顧及網絡統合分析與臨床試驗計畫書

Abstract

背景 近年來遺傳性類澱粉沉積症神經病變的藥物發展快速，但對已上市藥品缺乏頭對頭比較的數據，包括TTR穩定劑和TTR基因沉默。本系統性文獻回顧及網絡統合分析旨在比較這兩種機制在遺傳性類澱粉沉積症神經病變的療效和安全性。 方法 於2022年12月30日對PubMed電子數據庫和ClinicalTrials.gov進行了檢索。檢索標準包含介入措施為TTR穩定劑（tafamidis或diflunisal）和TTR基因沉默（patisiran或inotersen），與安慰劑或未治療的遺傳性類澱粉沉積症神經病變病人比較，研究結果包括療效：神經病變損傷評分（NIS，範圍從0到244，分數越高表明損傷越嚴重）、諾福克生活質量-糖尿病神經病變（Norfolk QOL-DN）問卷（範圍，-4 到136，得分越高表明生活質量較差）或改良式身體質量指數（mBMI，定義為［體重（公斤）除以身高（公尺）的平方］乘以白蛋白（克/升），值越低表明營養狀況越差）和安全性：不良事件（AEs）、嚴重不良事件（SAEs）或死亡。對其隨機對照臨床試驗的資料進行網絡統合分析。 結果 該網絡統合分析共納入四項隨機對照臨床試驗。包括的四項研究是：TTR穩定劑對照安慰劑（共2項研究）和TTR基因沉默對照安慰劑（共2項研究）。在神經病變損傷評分（NIS）的結果中，TTR穩定劑與安慰劑比較，與其基準線的平均差異變化為-16.8（95%信賴區間，-24.80至-8.80）。TTR基因沉默與安慰劑比較，與其基準線的平均差異變化為-12.82（95%置信區間[CI]，-20.54至-5.10）。間接比較時，TTR基因沉默和TTR穩定劑與其基準線的平均差異變化為3.96（95%信賴區間，-1.12至9.08）。根據SUCRA排名，TTR穩定劑得分為0.97，其次是TTR基因沉默得分為0.53。在諾福克生活質量-糖尿病神經病變（Norfolk QOL-DN）問券結果中，TTR穩定劑與安慰劑比較，與其基準線的平均差異變化為-4.50（95%信賴區間，-18.80至9.80）。TTR基因沉默與安慰劑比較，與其基準線的平均差異變化為-16.13（95%信賴區間，-26.18至-6.08）。間接比較時，TTR基因沉默和TTR穩定劑與其基準線的平均差異變化為-11.63（95%信賴區間，-29.11至5.84）。根據SUCRA排名，TTR基因沉默得分為0.95，其次是TTR穩定劑得分為0.42。在改良式身體質量指數（mBMI）結果中，TTR穩定劑與安慰劑比較，與其基準線的平均差異變化為53.82（95%信賴區間，-25.07至132.71）。TTR基因沉默與安慰劑比較，與其基準線的平均差異變化為63.75（95%信賴區間，-13.10至140.60）。間接比較時，TTR基因沉默和TTR穩定劑與其基準線的平均差異變化為9.93（95%信賴區間，-100.21至120.07）。根據SUCRA排名，TTR基因沉默得分為0.76，其次是TTR穩定劑得分為0.67。兩種機制和安慰劑的不良事件，整體的勝算比相似。 結論 從本研究結果可得知，甲狀腺素轉運蛋白的介入治療可能有益於延緩多發性神經病變，並且甲狀腺素轉運蛋白的介入治療與安慰劑組的安全性沒有差異。然而，很難驗證哪種治療機制更好。在神經病變損傷評分（NIS）中，TTR基因沉默與TTR穩定劑與其基準線的平均差異變化為3.96分，雖然沒有達到統計上的意義，但可能代表TTR穩定劑對多發性神經病變更有好處。因此，需要設計一個完整的臨床試驗，來比較TTR穩定劑和TTR基因沉默以獲得驗證。

BACKGROUND Drug development for hereditary amyloid transthyretin (hATTR) amyloidosis with polyneuropathy has progressed rapidly in recent years, but there is no head-to-head trial data to provide the efficacy comparative information for these marketed drugs, including TTR stabilizers and TTR gene silencers. This systematic review (SR) with network meta-analysis (NMA) aims to compare the efficacy and safety of these two mechanisms in the treatment of hATTR amyloidosis with polyneuropathy (FAP). METHODS A search of the PubMed electronic database and Clinicaltrials.gov was conducted on December 30th, 2022. The search criteria included the interventions in patients with FAP using TTR stabilizers (tafamidis or diflunisal) and TTR gene silencers (patisiran or inotersen), compared with the treatment naïve patients and those with placebo treatment. The study outcomes included 1) three efficacy evaluations: neuropathy impairment score (NIS, range of 0 to 244, with higher scores representing more neurological impairment), Norfolk QOL-DN questionnaire (range of -4 to 136, with higher scores indicating worse quality of life) and modified body mass index (mBMI, kg/m2 × albumin [g/L], with lower values, indicate poorer nutritional status), and 2) three safety assessments: adverse events (AEs), serious adverse events (SAEs), or deaths. An NMA was conducted employing the data from the randomized controlled trials. RESULTS A total of four RCTs were included in this NMA. The four studies were two trials with TTR stabilizer vs placebo and two with TTR gene silencers vs placebo. The mean difference change of NIS in FAP patients with treatment of TTR stabilizer vs those with placebo was -16.8 (95% CI, -24.80 to -8.80). The mean difference change in patients with TTR gene silencers vs those with placebo was -12.82 (95% CI, -20.54 to -5.10). The mean difference change from baseline for TTR gene silencer vs TTR stabilizer was 3.96 (95% c CI, -1.12 to 9.08) while compared indirectly. The SUCRA ranking demonstrated that the score for TTR stabilizer was 0.97, while 0.53 for the TTR gene silencer. The Norfolk QOL-DN results showed that the mean difference change from baseline for TTR stabilizer vs placebo was -4.50 (95% CI, -18.80 to 9.80), whereas -16.13 (95% CI, -26.18 to -6.08) for the groups with TTR gene silencer. The indirect comparison between TTR gene silencer vs TTR stabilizer was -11.63 (95% CI, -29.11 to 5.84). The SUCRA ranking scores were 0.95 for TTR gene silencer group and 0.42 for TTR stabilizer group. The mean difference change from baseline of mBMI for those patients with TTR stabilizer vs placebo was 53.82 (95% CI, -25.07 to 132.71), whereas 63.75 (95% CI, -13.10 to 140.60) for those with TTR gene silencer treatment. By comparison indirectly, the mean difference change for TTR gene silencer vs TTR stabilizer was 9.93 (95% CI, -100.21 to 120.07). The scores for SUCRA ranking were 0.76 for TTR gene silencer group and 0.67 for TTR stabilizer group. The overall odds ratios of adverse events were similar between the trials using the treatment of TTR gene stabilizer and TTR gene silencer. CONCLUSION According to the analysis, transthyretin interventions may be beneficial in delaying the progress of polyneuropathy, and there is no difference in safety concerns. However, it remains difficult to verify which one is better, based on the meta-analysis. The indirect comparison of the neuropathy impairment score (NIS) for the group of TTR gene silencer vs those with TTR stabilizer was 3.96 points which may represent a more favorable outcome of TTR stabilizer on the treatment of FAP although not statistically significant. Therefore, a clinical trial comparing TTR stabilizer and TTR gene silencer head-to-head needs to be designed to validate the observation.