Appendix R — 附錄 R: Ch13 長期追蹤 — OpenEvidence 一問一答

本附錄為 Ch13 「長期追蹤」章節的 OpenEvidence 實證驗證紀錄(12 題)。每題含:

  • Q:原設計問題(針對手冊特定敘述設計)
  • A:OE 綜合回答(整合多篇原始文獻)
  • Verdict:🟢 支持 / 🟡 修正 / 🟠 部分 / 🔴 推翻
  • Article ID:OpenEvidence session 識別碼
  • Top citations:Top 5 原始文獻(含 DOI 與 chip 代碼;📄 本機 PDF 由 chip drawer 開啟)

驗證日期:2026-04-22 狀態:12 / 12 完成

完整 metadata 見 qa/oe_papers_library/Ch13_long_term/Ch13_citations.json


R.1 驗證記錄

驗證日期:2026-04-22 | 章節:Ch13 長期追蹤 | 題數:12(6 High, 5 Med, 1 Low)| 新引用:109 | 重用既有:8(Marczin 2021 AnesICU、Sharma 2025 AASLD/AST、Xiong 2023 TacCKD、Stein 2011 Meta、Baughman 2012 CHEST、McPheeters 2021 TacTrough、Hannan 2025 BelaIPAH、Penninga 2013 Cochrane)

R.1.1 Q120 [High] Ch13:32 5 年後 CKD 4-5 ~16% 🟢

原 claim(line 32): 「5 年後 ~16% CKD 4-5」

OE 主要發現: - 手冊的 “16% at 5 yr” 與 contemporary 單中心 / cohort 資料精確相符,但非源自大型 registry(ISHLT / UNOS 未專指此 metric)[OE-Doraiswamy-2021-CKDstage]。 - Doraiswamy 2021(單中心 5-yr follow-up):16% on dialysis, 86% 有任一級 CKD(eGFR <60)at 60 months — 手冊數字最直接來源 [OE-Doraiswamy-2021-CKDstage]。 - Hellemons 2011(Dutch cohort, measured GFR):cumulative incidence CKD stage 4 at 5 years = 16.3% — 與 16% 完美一致 [OE-Hellemons-2011-CKDstage]。 - Wiseman 2022 review:CKD 影響肝 / 心 / 肺移植 10-20%,但未分 stage [OE-Wiseman-2022-CKDstage]。 - Kanani 2025 單中心:ESRD 高達 25% at 10 years — 長期惡化趨勢 supports 手冊 “長期累積” [OE-Kanani-2025-CKDstage]。 - KDIGO 2024 CKD guideline 不提供移植特異性 prevalence;ISHLT 2021 anesthetic/ICU consensus(Marczin 2021)強調 perioperative renal protection 但未提 5-yr stage 4-5 比例 [OE-Iatridi-2025-CKDstage, OE-Marczin-2021-AnesICU]。 - Tac long-term exposure 為 CKD 主要 driver [OE-Xiong-2023-TacCKD]。

Verdict: 🟢 SUPPORTED16% at 5 yr 精確對應 Doraiswamy 2021 + Hellemons 2011 數據;宜註記「單中心 / 中型 cohort 數據,非 ISHLT registry specific metric」。

手冊建議修訂: - Ch13:32 保留原文,可補引用:「5 年後 ~16% CKD 4-5(Doraiswamy 2021 單中心 n=? 5-yr:16% on dialysis;Hellemons 2011 Dutch cohort measured GFR:CKD4 cumulative incidence 16.3% at 5 yr)」 - 加引用 [OE-Doraiswamy-2021-CKDstage, OE-Hellemons-2011-CKDstage, OE-Wiseman-2022-CKDstage]

Key citations: [OE-Doraiswamy-2021-CKDstage], [OE-Hellemons-2011-CKDstage], [OE-Wiseman-2022-CKDstage], [OE-Kanani-2025-CKDstage], [OE-Xiong-2023-TacCKD]

R.1.2 Q121 [Med] Ch13:88 PTDM 20-40% 🟢

原 claim(line 88): 「PTDM 20-40%」

OE 主要發現: - 20-40% 區間 broadly supported:1-yr 20-33%、2-3 yr 24-35%;Hackman 2014 prospective OGTT:NODAT 30% at 12 mo, 24% at 24 mo, total DM prevalence 40% at 2 yr(包含 pre-existing) [OE-Hackman-2014-PTDM]。 - UNOS 大型分析(Ye 2010, n=2991):NODAT 33.4% over median 670 d [OE-Ye-2010-PTDM]。 - 單中心:Ollech 2008 LTx NODAT incidence 與 risk factors [OE-Ollech-2008-PTDM];Mizrahi 2020 SOT(含 158 LTx)21.35% at mean 11.3 mo [OE-Mizrahi-2020-PTDM]。 - ISHLT 2021 registry report(Chambers 2021) 未提供 cumulative PTDM incidence specific data [OE-Chambers-2021-PTDM]。 - Tac vs CsA:Tac 顯著增加 PTDM 風險 — Suilik 2024 meta-analysis(4 RCTs, 677 pt)RR 0.33 (CsA vs Tac), I²=0%, p=0.03,即 Tac ~3x 風險;Penninga 2013 Cochrane RR 4.24 (fixed);Fan 2009 meta OR 3.69 [OE-Suilik-2024-PTDM, OE-Penninga-2013-PTDM, OE-Fan-2009-PTDM]。 - 5-yr 特異數據罕見 — general estimate ~32% [OE-Sharif-2016-PTDM];Du 2024 retrospective 補充 risk factor 分析 [OE-Du-2024-PTDM]。

Verdict: 🟢 SUPPORTED20-40% 區間 精確反映 1-2 yr 時間窗口;可補「Tac → CsA conversion 可降低 PTDM 3-4x 風險」作為 modifiable factor 提醒。

手冊建議修訂: - Ch13:88 保留範圍,補註:「Tac 較 CsA ~3x PTDM 風險(Suilik 2024 meta n=677, RR 0.33);OGTT 診斷更敏感(Hackman 2014 OGTT 1-yr 30%)」 - 加引用 [OE-Hackman-2014-PTDM, OE-Ye-2010-PTDM, OE-Suilik-2024-PTDM, OE-Penninga-2013-PTDM]

Key citations: [OE-Hackman-2014-PTDM], [OE-Ye-2010-PTDM], [OE-Ollech-2008-PTDM], [OE-Suilik-2024-PTDM], [OE-Penninga-2013-PTDM], [OE-Sharif-2016-PTDM]

R.1.3 Q122 [Low] Ch13:188 高血壓 > 80% post-LTx 🔴

原 claim(line 188): 「高血壓 > 80% post-LTx」

OE 主要發現: - “>80% at 1 year” 過時且不正確 — contemporary 1-yr prevalence 45-50%;5-yr 才達 ~85% [OE-Silverborn-2005-HTN, OE-Penninga-2013-Cochrane]。 - Silverborn 2005 JHLT:1-yr 45%, 3-yr 65%, 7-yr 72% — 明確顯示 HTN 為漸進性而非 immediate majority [OE-Silverborn-2005-HTN]。 - Penninga 2013 Cochrane:5-yr 85% — 與手冊 > 80% 數字相符但時間窗口必須為 5 yr 而非 1 yr [OE-Penninga-2013-Cochrane]。 - Savioli 2013 prospective metabolic syndrome:pre-Tx 19.4% → 3-yr 70.1%(HTN 組分) [OE-Savioli-2013-HTN]。 - 歷史數據:Morrison 1993 66% at mean 11 mo(早期 CsA era)[OE-Morrison-1993-HTN];Jenkins 1998 review [OE-Jenkins-1998-HTN]。

Verdict: 🔴 REFUTED(手冊數字與時間窗口不匹配)1-yr HTN 實為 45-50%,非 >80%;“>80%” 僅在 5-yr 後達到。手冊若為 “HTN post-LTx 常見”(不限時)仍 broadly 正確,但具體 “>80% 1-yr” 錯誤。

手冊建議修訂: - Ch13:188 必修:「HTN post-LTx 漸進發展:1-yr 45-50%(Silverborn 2005 JHLT:1-yr 45%),3-yr 65-70%,5-yr ~85%(Penninga 2013 Cochrane)。CNI-induced vasoconstriction + steroid 為主要機制」 - 移除「> 80% post-LTx」若原文暗示 early post-op;保留「長期 > 80%」並註明 5-yr timeline - 加引用 [OE-Silverborn-2005-HTN, OE-Penninga-2013-Cochrane, OE-Savioli-2013-HTN]

Key citations: [OE-Silverborn-2005-HTN], [OE-Penninga-2013-Cochrane], [OE-Savioli-2013-HTN], [OE-Morrison-1993-HTN], [OE-Jenkins-1998-HTN]

R.1.4 Q123 [Med] Ch13:228 1 yr 骨密度 ↓ 2-5% 🟡

原 claim(line 228): 「1 yr 骨密度 ↓ 2-5%」

OE 主要發現: - “2-5%” 的上限(~5%)吻合 landmark Spira 2000 data但此為 bisphosphonate-prophylaxis 前的資料,contemporary 實務下 BMD 可持平或增加 [OE-Spira-2000-BMDloss]。 - Spira 2000 Chest(n=28 landmark, no bisphosphonate):lumbar 4.76%, femoral neck 5.3% BMD loss at 6-12 mo — “5%” 的來源 [OE-Spira-2000-BMDloss]。 - Wang 2013 Transplant Proc 縱貫性 heart/lung 數據補充 [OE-Wang-2013-BMDloss]。 - Modern bisphosphonate prophylaxis 使數字反轉: Ng 2020 J Clin Densitom zoledronic acid prophylaxis:lumbar +8.1% vs untreated -1.2%(p=0.002) [OE-Ng-2020-Bisphos]。 - Caffarelli 2020 JCM(n=128 LTx/HTx):fracture incidence pre-Tx 19.5% → 3-yr 50.4% — 臨床終點顯示 prophylaxis 必要性 [OE-Caffarelli-2020-BMDloss]。 - Tran 2023 J Cyst Fibros CF 專 cohort,modern practice [OE-Tran-2023-BMDloss]。 - Rzepka 2025 JHLT narrative review(最新)+ Anastasilakis 2019 European Calcified Tissue Society recommendations:standard of care 為 universal bisphosphonate prophylaxis 自 Tx 後 [OE-Rzepka-2025-Bisphos, OE-Anastasilakis-2019-Bisphos]。

Verdict: 🟡 PARTIALLY SUPPORTED“2-5%” 是 historical(no prophylaxis)數字上限 accurate;現代 bisphosphonate 常規使用下實際 BMD loss 常為 0 或正增加,手冊宜標註 prophylaxis 背景

手冊建議修訂: - Ch13:228 改為:「1 yr 骨密度 loss ~4-5%(Spira 2000 landmark, no prophylaxis; lumbar 4.76%, femoral neck 5.3%);當代常規 bisphosphonate prophylaxis 下 BMD 可持平或增加(Ng 2020 zoledronic acid:lumbar +8.1% vs -1.2% untreated)」 - 加 fracture 數據:「pre-Tx 19.5% → 3-yr 50.4%(Caffarelli 2020)」說明 prophylaxis 重要性 - 加引用 [OE-Spira-2000-BMDloss, OE-Ng-2020-Bisphos, OE-Caffarelli-2020-BMDloss, OE-Rzepka-2025-Bisphos]

Key citations: [OE-Spira-2000-BMDloss], [OE-Ng-2020-Bisphos], [OE-Caffarelli-2020-BMDloss], [OE-Rzepka-2025-Bisphos], [OE-Anastasilakis-2019-Bisphos]

R.1.5 Q124 [Med] Ch13:239 Alendronate 70/wk 或 Zoledronic acid 5 mg/yr 🟡

原 claim(line 239): 「Alendronate 70 mg/wk 或 Zoledronic acid 5 mg/yr」

OE 主要發現: - 唯一 head-to-head RCT:Shane 2012 JCEM(n=84 heart + liver Tx):單次 zoledronic acid 5 mg IV vs alendronate 70 mg/wk × 12 mo:lumbar spine BMD 差異 2.2%(95% CI 0.6-3.9, p=0.009)favor zoledronic acid;heart subgroup 差異更大(4.2%, p<0.001) [OE-Shane-2012-Bisphos]。 - 無 LTx-specific head-to-head RCT — evidence 為 extrapolation from heart/liver Tx + glucocorticoid-induced osteoporosis [OE-Shane-2012-Bisphos, OE-Rzepka-2025-Bisphos]。 - HORIZON-GIOP(Reid 2009 Lancet, 非 Tx):zoledronic acid 優於 risedronate for lumbar spine BMD(prevention 2.6% vs 0.6%, p<0.0001;treatment 4.1% vs 2.7%, p=0.0001) [OE-Reid-2009-Bisphos]。 - LTx-specific retrospective:Ng 2020/2021 prophylactic zoledronic acid within 6 mo of LTx:median lumbar BMD +8.1% vs -1.2% untreated (p=0.002) [OE-Ng-2020-Bisphos]。 - January 2025 retrospective comparing bisphosphonates vs denosumab vs PTH analogs in LTx:all three classes effective at improving lumbar BMD [OE-January-2025-Bisphos]。 - Stein 2011 meta SOT(所有器官, fracture endpoint):bisphosphonates 顯著降低 fracture risk [OE-Stein-2011-Meta]。 - Guidelines:2022 ACR GIOP guideline(Humphrey 2023)+ Anastasilakis 2019 ECTS 均推薦 bisphosphonate first-line for post-Tx osteoporosis;無 ISHLT specific guideline [OE-Humphrey-2023-Bisphos, OE-Anastasilakis-2019-Bisphos]。 - AASLD/AST 2026 liver Tx guideline 涵蓋 osteoporosis management parallels [OE-Sharma-2025-InfectISreduc]。

Verdict: 🟡 PARTIALLY SUPPORTED兩個選項均為合理一線但 zoledronic acid 在 heart/liver Tx RCT 為優選(Shane 2012 lumbar 2.2% advantage)。手冊兩選項並列 acceptable,但未註明 IV form 實證較佳。

手冊建議修訂: - Ch13:239 改為:「一線選項:Alendronate 70 mg/wk PO 或 Zoledronic acid 5 mg/yr IV(優選,Shane 2012 JCEM 顯示 lumbar BMD 差異 +2.2%, p=0.009 優於 alendronate,尤其 heart Tx subgroup +4.2%, p<0.001);LTx-specific 資料 Ng 2020 zoledronic prophylaxis lumbar BMD +8.1%」 - 加引用 [OE-Shane-2012-Bisphos, OE-Ng-2020-Bisphos, OE-January-2025-Bisphos, OE-Rzepka-2025-Bisphos]

Key citations: [OE-Shane-2012-Bisphos], [OE-Ng-2020-Bisphos], [OE-January-2025-Bisphos], [OE-Rzepka-2025-Bisphos], [OE-Humphrey-2023-Bisphos], [OE-Reid-2009-Bisphos]

R.1.6 Q125 [Med] Ch13:272-279 惡性 2-4x; SCC ~65x 🟡

原 claim(line 272-279): 「惡性 2-4x; SCC ~65x」

OE 主要發現: - “Overall cancer 2-4x” 強 supported:Finnish 30-yr cohort(n=6548)overall SIR 3.6(heart 5.0 / liver 2.7);excluding NMSC SIR 2.2 [OE-Friman-2022-CancerSIR]。UNOS LTx(n=18,093)non-skin cancer SIR 3.26 [OE-Magruder-2017-CancerSIR]。Ge 2020 meta(21 cohorts, 65,265 LTx)pooled SIR 4.28 [OE-Ge-2020-CancerSIR]。 - “SCC ~65x” 是合理中間值但可能低估 LTx-specific 風險: - Jin 2024 Lancet Oncology meta:pooled cSCC meta-SIR 45.9(I²=99%),heart/lung 最高 [OE-Jin-2024-CancerSIR]。 - Krynitz 2013 Swedish registry:overall SCC SIR 121(95% CI 116-127);heart/lung specifically SIR 198(95% CI 174-224) [OE-Krynitz-2013-CancerSIR]。 - Rizvi 2017 Norwegian:overall SIR 51.9 (95% CI 48.4-55.5) [OE-Rizvi-2017-CancerSIR]。 - Swedish LTx-specific(Stenman 2023)n=614:NMSC SIR 76.5 (95% CI 61.7-94.8) [OE-Stenman-2023-CancerSIR]。 - Hortlund 2017 Scandinavian combined NMSC SIR 44.7 (Sweden) / 41.5 (Denmark) [OE-Hortlund-2017-CancerSIR]。 - Wysong 2023 NEJM review: LTx cSCC incidence factor 5-113 — 巨大 heterogeneity [OE-Wysong-2023-CancerSIR]。 - Dusendang 2022 cohort/nested case-control 補充 LTx cSCC risk factors [OE-Dusendang-2022-CancerSIR]。 - Engels 2026 US population-level trends + Dillibabu 2025 Quebec + Wilken 2021 review 皆 supports 高 SCC burden [OE-Engels-2026-CancerSIR, OE-Dillibabu-2025-CancerSIR, OE-Wilken-2021-CancerSIR]。

Verdict: 🟡 PARTIALLY SUPPORTED2-4x overall precisely correct“~65x SCC” 是合理中位點但 Scandinavian LTx-specific data 高達 76-198x,手冊可能 低估 LTx 特異性 SCC 風險

手冊建議修訂: - Ch13:272-279 refine:「整體癌症 SIR 2-4x(Finnish cohort n=6548 SIR 3.6;Ge 2020 meta LTx SIR 4.28);cSCC SIR 範圍 45-200x 視 registry 而定(Jin 2024 meta pooled 45.9;Krynitz 2013 Swedish heart/lung 198;Stenman 2023 Swedish LTx 76.5)。~65x 為合理中值但對 LTx 可能低估」 - 加引用 [OE-Friman-2022-CancerSIR, OE-Krynitz-2013-CancerSIR, OE-Stenman-2023-CancerSIR, OE-Jin-2024-CancerSIR, OE-Ge-2020-CancerSIR]

Key citations: [OE-Krynitz-2013-CancerSIR], [OE-Friman-2022-CancerSIR], [OE-Stenman-2023-CancerSIR], [OE-Jin-2024-CancerSIR], [OE-Ge-2020-CancerSIR], [OE-Magruder-2017-CancerSIR], [OE-Wysong-2023-CancerSIR]

R.1.7 Q126 [Med] Ch13:300 PTLD 3-10% post-LTx 🟡

原 claim(line 300): 「PTLD 3-10% post-LTx」

OE 主要發現: - “3-10%” 為 NEJM review 引用的 wide range;contemporary large registry 數據位於 lower end [OE-Dierickx-2018-PTLD, OE-Fishman-2007-PTLD]。 - ISHLT Registry 分析(Zaffiri 2020, n=19,309 LTx 2006-2016)1.1% at 1 yr, 4.1% at 10 yr(47.4% cases 發生於第一年)— 最大最 rigorous 資料 [OE-Zaffiri-2020-PTLD]。 - 單中心 contemporary cohorts:4-5%:Iasella 2020(n=611, 7 yr)4.6%;Kremer 2012(n=705)4.8%;Leyssens 2017 nested case-control 3.9%;Kumarasinghe 2015 3.1%;Aris 1996 老期序列 6.2-9.4%(IS 較強) [OE-Iasella-2020-PTLD, OE-Kremer-2012-PTLD, OE-Leyssens-2017-PTLD, OE-Kumarasinghe-2015-PTLD, OE-Aris-1996-PTLD]。 - EBV D+/R- stratification 是關鍵:SRTR(Courtwright 2018, n=11,633):D+/R- 6.2% vs 其他 1.4%(adjOR 4.0, 95% CI 2.8-5.9) [OE-Courtwright-2018-PTLD]。單中心 Pittsburgh LTx:EBV mismatch HR 34.4 (95% CI 15.6-76.1) for PTLD [OE-Iasella-2020-PTLD]。 - Wong 2004 早期 LTx EBV-naïve 組高達 31-33% PTLD incidence(primary EBV infection 42%) [OE-Wong-2004-PTLD]。 - Allen & Preiksaitis 2019 AST-IDCOP guideline:pre-Tx EBV seronegativity + primary donor-transmitted EBV 為最重要 risk factors [OE-Allen-2019-PTLD]。 - Peters 2018 + Rampersad 2025 + Kim 2024:epidemiology 漸進變化 / time-varying risk [OE-Peters-2018-PTLD, OE-Rampersad-2025-PTLD, OE-Kim-2024-PTLD]。

Verdict: 🟡 PARTIALLY SUPPORTED3-10% 作為教科書範圍 acceptable,但 contemporary ISHLT registry 數字 4% at 10 yr 位於下緣;更重要的是 EBV D+/R- 風險 ~6% 而 EBV-seropositive <2% — 手冊應加 EBV stratification 註記。

手冊建議修訂: - Ch13:300 refine:「PTLD incidence:ISHLT Registry 大型數據(Zaffiri 2020 n=19,309)1.1% at 1 yr, 4.1% at 10 yr;單中心當代 cohorts 3-5%。老期序列 6-9%(高 IS era)EBV D+/R- 風險 ~6%(adjOR 4.0)vs seropositive <2%(Courtwright 2018 SRTR n=11,633);LTx EBV-naïve primary infection 可達 31-42%(Wong 2004 歷史 cohort)」 - 加引用 [OE-Zaffiri-2020-PTLD, OE-Courtwright-2018-PTLD, OE-Iasella-2020-PTLD, OE-Allen-2019-PTLD]

Key citations: [OE-Zaffiri-2020-PTLD], [OE-Courtwright-2018-PTLD], [OE-Iasella-2020-PTLD], [OE-Dierickx-2018-PTLD], [OE-Kremer-2012-PTLD], [OE-Allen-2019-PTLD], [OE-Wong-2004-PTLD]

R.1.8 Q127 [High] Ch13:318 多次 SCC → switch to mTOR (evidence B) 🟢

原 claim(line 318): 「多次 SCC → switch to mTOR (evidence B)」

OE 主要發現: - CNI→mTORi conversion 顯著降低 cSCC 復發 in kidney Tx:2026 systematic review + meta-analysis 4 RCTs(n=393):IRR 0.51(95% CI 0.39-0.67),high certainty evidence [OE-Foerster-2026-mTORconv]。 - TUMORAPA 為 landmark RCT(Euvrard 2012, n=120 kidney Tx with ≥1 prior cSCC):2-yr new cSCC 22% sirolimus vs 39% CNI (RR 0.56, p=0.02);HR for SCC-free survival 0.37 (95% CI 0.16-0.85)。Benefit 最顯著於 single prior SCC(HR 0.03)、non-significant for multiple prior SCCs(HR 0.67) [OE-Euvrard-2012-mTORconv]。 - Dantal 2018 TUMORAPA 5-yr extension:22% vs 59% new SCCs (p<0.001),benefit 持續 [OE-Dantal-2018-mTORconv]。 - CONVERT Trial(n=830 kidney Tx):total malignancy 2.1 vs 6.0 per 100 PY (p<0.001),NMSC 1.2 vs 4.3 per 100 PY [OE-Alberú-2011-mTORconv, OE-Alberú-2010-mTORconv]。 - Hoogendijk-van den Akker 2013(n=155 kidney + ≥1 SCC):primary endpoint SCC-free survival at 2 yr not significant;但 after year 1 50% risk reduction;39% discontinuation rate 副作用 [OE-Hoogendijkvan-2013-mTORconv]。 - Campbell 2012(n=86 kidney):yearly SCC rate lower (p=0.038),但 46% discontinued for AE [OE-Campbell-2012-mTORconv]。 - Knoll 2014 meta:overall malignancy reduction + mortality concern(early sirolimus 可能 increase mortality)[OE-Knoll-2014-mTORconv]。 - LTx / heart Tx 資料 limited 且多為 observational:Asleh 2019 heart Tx retrospective sirolimus 降低 malignancy;Abulail 2026 meta NMSC in LTx;Gómez-Tomás 2026 SOT skin cancer review;Préterre 2021 mTORi strategies [OE-Asleh-2019-mTORconv, OE-Abulail-2026-mTORconv, OE-GómezTomás-2026-mTORconv, OE-Préterre-2021-mTORconv]。 - Massey 2021 SOT consensus-based recommendations on SCC prevention + de Fijter 2017 cancer mTORi review:early conversion 優於多次 SCC 後才 convert [OE-Massey-2021-mTORconv, OE-de-2017-mTORconv]。 - Adverse events 為主要限制:TUMORAPA sirolimus SAE 0.94 vs 0.25 per patient;23% 停藥;progressive conversion 較 rapid 好 [OE-Euvrard-2012-mTORconv]。

Verdict: 🟢 SUPPORTED“evidence B” 為恰當 strength rating(>1 RCT 但 LTx-specific 缺乏 direct RCT);核心 caveat:benefit 在 single SCC 最強、multiple SCCs 減弱,與手冊 “多次 SCC” 推薦恰恰相反——宜改為 “≥1 SCC(尤其第一次 SCC 後)”

手冊建議修訂: - Ch13:318 修訂時機:「發生第一次或 cSCC 再發 → 考慮 CNI → mTORi conversion(sirolimus/everolimus),evidence level B。TUMORAPA RCT(Euvrard 2012 NEJM, kidney Tx n=120)顯示 single prior SCC benefit HR 0.03(p<0.001),但 multiple prior SCCs benefit 消失(HR 0.67, NS)——early conversion 比多次 SCC 後才 switch 更有效。2026 meta IRR 0.51 (95% CI 0.39-0.67, 4 RCTs n=393)。權衡 23-46% 副作用停藥率、progressive(非 rapid)conversion 較 tolerated」 - 加 LTx-specific caveat:「LTx 資料 observational(Asleh 2019 heart, Abulail 2026 LTx meta),direct RCT 缺乏」 - 加引用 [OE-Euvrard-2012-mTORconv, OE-Foerster-2026-mTORconv, OE-Dantal-2018-mTORconv, OE-Massey-2021-mTORconv]

Key citations: [OE-Euvrard-2012-mTORconv], [OE-Foerster-2026-mTORconv], [OE-Dantal-2018-mTORconv], [OE-Hoogendijkvan-2013-mTORconv], [OE-Campbell-2012-mTORconv], [OE-Knoll-2014-mTORconv], [OE-Massey-2021-mTORconv]

R.1.9 Q128 [High] Ch13:348 GERD 高達 70%; fundoplication 改善 CLAD 🟡

原 claim(line 348): 「GERD 高達 70%; fundoplication improves CLAD」

OE 主要發現: - Evidence 為 entirely observational cohort studies — 無 RCT;signal moderately strong 但 substantially limited by confounding(selection bias, immortal time bias) [OE-Green-2023-Fundo]。 - Duke seminal(Davis 2003, n=43):fundoplication 後 FEV1 +24%(1.87 → 2.19 L, p<0.0002)at 6 mo;13/26 BOS 患者 不再 meet BOS criteria [OE-Davis-2003-Fundo]。Hartwig 2011 Duke follow-up:abnormal pH 接受 fundoplication 者 peak FEV1 84% vs 75% (p=0.004),1-yr FEV1 77% vs 68% (p=0.003) [OE-Hartwig-2011-Fundo]。 - Toronto(Biswas Roy 2018):早期 fundoplication (<6 mo) vs 晚 associated with 5.7% 更高 %predicted FEV1,5-yr FEV1 40.7% higher (p=0.019) [OE-Biswas-2018-Fundo]。 - Multicenter 2023(Green, Toronto/Duke/WashU/其他 n=542, 136 fundo)standard analysis NOT statistically significant(BOS HR 0.93, 95% CI 0.58-1.49;death HR 0.97, 95% CI 0.47-1.99);time-dependent trends favorable but NS(BOS HR 0.68, p=0.126;death HR 0.59, p=0.157) [OE-Green-2023-Fundo]。 - Columbia propensity-matched(Leiva-Juarez 2022, n=971, 128 pre-CLAD fundo):pre-CLAD fundoplication 改善 survival HR 0.57 (95% CI 0.4-0.8, p=0.001);established CLAD 後 fundoplication post-CLAD survival HR 0.27 (95% CI 0.12-0.61, p=0.002) [OE-LeivaJuarez-2022-Fundo]。 - 2025 meta-analysis(Krahelski, 20 studies):anti-reflux surgery FEV1 rate of change improvement -1.153 mL/day(95% CI -12.117 to -0.188),survival HR 0.39 (95% CI 0.19-0.60);Nissen fundoplication 效果最大(-2.353 mL/day) [OE-Krahelski-2025-Fundo]。 - Razia 2023 antireflux surgery vs medical + Kowalski 2023 partial fundoplication:supporting evidence [OE-Razia-2023-Fundo, OE-Kowalski-2023-Fundo]。 - Confounding 嚴重:selection bias(接受手術者系統性較健康)、immortal time bias、timing confounding(early fundoplication 患者可能 baseline 較好) — Green 2023 multicenter 明確 acknowledge “RCT warranted”。

Verdict: 🟡 PARTIALLY SUPPORTEDObservational 訊號強(meta HR 0.39 for survival),但無 RCTGreen 2023 multicenter 未達 statistical significance in standard analysis。“改善 CLAD” 宜表述為 “associated with 較慢 FEV1 decline 與 improved survival in observational data, 無 RCT 證據”。

手冊建議修訂: - Ch13:348 refine:「GERD post-LTx 30-70% 常見(依診斷標準而定)。Fundoplication 在 observational cohorts 顯示 FEV1 decline 減緩 + 改善 survival(Leiva-Juarez 2022 Columbia propensity-matched pre-CLAD HR 0.57;Krahelski 2025 meta 20 studies HR 0.39),但 2023 multicenter(Green, n=542)standard analysis 未達統計顯著性(BOS HR 0.93, NS);無 RCT,selection/immortal time bias 嚴重。⚠️ Early fundoplication(<6 mo, Biswas 2018 Toronto)benefit 最大」 - 加引用 [OE-Green-2023-Fundo, OE-LeivaJuarez-2022-Fundo, OE-Davis-2003-Fundo, OE-Biswas-2018-Fundo, OE-Krahelski-2025-Fundo]

Key citations: [OE-Green-2023-Fundo], [OE-LeivaJuarez-2022-Fundo], [OE-Davis-2003-Fundo], [OE-Hartwig-2011-Fundo], [OE-Biswas-2018-Fundo], [OE-Krahelski-2025-Fundo]

R.1.10 Q129 [Med] Ch13:376 Airway: stenosis 10-15%, dehiscence < 2% 🟢

原 claim(line 376): 「Airway: stenosis 10-15%, dehiscence < 2%」

OE 主要發現: - “stenosis 10-15%, dehiscence <2%” 與 contemporary data broadly consistent — 但 overall airway complication rate 範圍 2-22% 視 series 而定 [OE-Crespo-2021-Airway, OE-Yin-2025-Airway]。 - Huang 2023 meta-analysis(38 studies, n=52,116 LTx):pooled overall airway complication 12.4% (95% CI 9.5-15.8%);stenosis dominant [OE-Huang-2023-Airway]。 - Stenosis 具體 rates:Kim 2023 韓國 cohort 16.8%;U.S. 2020-2023 cohort 12.7%(Dhanani 2025);Dutch Van Pel 2023 30-yr experience 67% of airway complications = stenosis (~4.3% of anastomoses) [OE-Kim-2023-Airway, OE-Dhanani-2025-Airway, OE-van-2023-Airway]。 - Dehiscence < 2% 強 supported:Muñoz-Fos 2023 Spanish (n=400) 1% (4/400);dehiscence 為最致命 complication:30-day mortality 60.9% (95% CI 20.6-95.2%)(Huang 2023 meta) [OE-MuñozFos-2023-Airway, OE-Huang-2023-Airway]。 - Temporal decline:Schweiger 2020 Austrian (single running suture) n=2941:2.4% (1999-2003) → 0.8% (2014-2017) — 技術進步降低 incidence [OE-Schweiger-2020-Airway]。Overall morbidity historical 2-33% → contemporary 2-18% [OE-Crespo-2021-Airway]。 - 2018 ISHLT 共識 grading system 改善定義 standardization [OE-Crespo-2021-Airway]。 - Mahajan 2017 review:diagnosis + management standard [OE-Mahajan-2017-Airway]。

Verdict: 🟢 SUPPORTED兩數字均與 contemporary evidence 符合;可補「technique-dependent:single running suture 可降至 <1%(Schweiger 2020)」。

手冊建議修訂: - Ch13:376 保留原文,可補:「Stenosis 10-15%(range 7-17%)+ dehiscence <2%(Muñoz-Fos 2023 1%)Huang 2023 meta n=52,116 overall airway complications 12.4%2018 ISHLT consensus grading system 提供 standardized definitions;single running suture technique 可降 overall 至 <1% (Schweiger 2020 Austrian n=2941)」 - 加 mortality 警示:「dehiscence 30-day mortality 60.9%(最致命 complication)」 - 加引用 [OE-Huang-2023-Airway, OE-Schweiger-2020-Airway, OE-MuñozFos-2023-Airway, OE-Crespo-2021-Airway]

Key citations: [OE-Huang-2023-Airway], [OE-Crespo-2021-Airway], [OE-Schweiger-2020-Airway], [OE-MuñozFos-2023-Airway], [OE-Kim-2023-Airway], [OE-Mahajan-2017-Airway]

R.1.11 Q130 [High] Ch13:430 IPF 25-30% carry STS variants 🟡

原 claim(line 430): 「IPF 25-30% carry STS variants」

OE 主要發現: - “25-30%” 反映 familial PF 的 telomere gene variant 比例,不直接適用所有 IPF — 重要 clarification [OE-Ryerson-2025-STSgene, OE-Newton-2022-STSgene]。 - ACCP consensus 2025 (Ryerson / Newton):telomere-related variants 在unselected familial PF kindreds 佔 20-30%;selected strong family history cohorts 可達 80%;TERT 最常見(10-20%),其次 RTEL1, PARN, TERC(2-5%) [OE-Newton-2022-STSgene, OE-Ryerson-2025-STSgene]。 - LTx 受者 specificallyAlder 2022 發現 24% IPF LTx recipients 帶有 telomere-associated gene variants — 與手冊 25-30% 吻合但為 LTx-enriched population [OE-Alder-2022-STSgene]。 - Sporadic IPF(非 familial)telomere variant prevalence 較低:Petrovski 2017 exome sequencing sporadic IPF;Dressen 2018 protein-altering variants analysis — 一般人群 sporadic IPF 可能 <10% [OE-Petrovski-2017-STSgene, OE-Dressen-2018-STSgene]。 - Zhang 2022 whole genome sequencing utility + Newton 2022 ACCP perspective [OE-Zhang-2022-STSgene, OE-Newton-2022-STSgene]。 - Lung transplant outcomes:Swaminathan 2019 shorter telomere → worse LTx outcomes;Phillips-Houlbracq 2022 多中心 telomerase mutation LTx determinants of survival;Mackintosh 2025 comprehensive pretransplant/posttransplant review [OE-Swaminathan-2019-STSgene, OE-PhillipsHoulbracq-2022-STSgene, OE-Mackintosh-2025-STSgene]。 - ACCP 建議 panel-based gene sequencing for all familial PF,考慮於 sporadic PF with age <50 y 或 strong family history不建議 routine telomere testing 決定 LTx candidacy in sporadic PF [OE-Ryerson-2025-STSgene]。

Verdict: 🟡 PARTIALLY SUPPORTED25-30% 適用於 familial PF 或 LTx-enriched IPF cohorts,對所有 IPF 會 overestimate;手冊宜 clarify denominator。

手冊建議修訂: - Ch13:430 refine:「IPF 中 telomere-related gene variant prevalencefamilial PF kindreds 20-30%(unselected),selected 強家族史可達 80%(Newton 2022 ACCP);LTx recipients 特化 cohort 24%(Alder 2022 IPF LTx)。Sporadic IPF 一般 <10%(Petrovski 2017 exome)。最常見 TERT(10-20%),次 RTEL1/PARN/TERC (2-5%)」 - 加 LTx implication:「ACCP 2025 建議 panel sequencing for all familial PF;sporadic PF 不 routine test for LTx candidacy(Ryerson 2025)」 - 加引用 [OE-Alder-2022-STSgene, OE-Ryerson-2025-STSgene, OE-Newton-2022-STSgene, OE-Swaminathan-2019-STSgene, OE-PhillipsHoulbracq-2022-STSgene]

Key citations: [OE-Alder-2022-STSgene], [OE-Ryerson-2025-STSgene], [OE-Newton-2022-STSgene], [OE-Swaminathan-2019-STSgene], [OE-PhillipsHoulbracq-2022-STSgene], [OE-Mackintosh-2025-STSgene]

R.1.12 Q131 [High] Ch13:456 STS: Tac 6-8, MMF 減半, avoid ATG 🟡

原 claim(line 456): 「STS: Tac 6-8, MMF 減半, avoid ATG」

OE 主要發現: - 所有 3 項 modifications 均為 expert consensus / observational data,缺乏 STS-specific RCTs — 重要限制 [OE-Courtwright-2025-STSimmuno, OE-Carlier-2025-STSimmuno]。 - ISHLT 2025 consensus statement on STS and lung transplantation(Courtwright 2025, Nov 2025):acknowledge STS 需 multidisciplinary + modified IS regimens;但 specific numeric targets 未在 abstract 提供 [OE-Courtwright-2025-STSimmuno]。 - Tac 6-8 ng/mL target 無 direct evidence:一般 LTx Tac standard 5-15 ng/mL(early 10-15, later 8-12);McPheeters 2021 CNI maintenance data;Sohn 2023 continuous IV PK study [OE-McPheeters-2021-TacTrough, OE-Sohn-2023-STSimmuno]。Baughman 2012 ACCP CHEST 指引 CNI monitoring [OE-Baughman-2012-CHEST]。無 cohort 比較 6-8 vs 標準 targets in STS;建議基於 expert opinion balance against 骨髓毒性 [OE-Carlier-2025-STSimmuno, OE-Courtwright-2019-STSimmuno]。 - MMF 減半 supported by STS 骨髓不耐受 observational data:Carlier 2025 hematologic complications in telomerase mutation LTx;Borie 2015 severe hematologic complications post-LTx in short telomere [OE-Carlier-2025-STSimmuno, OE-Borie-2015-STSimmuno]。French multicenter(Borie 2015):6/9 telomerase mutation 患者發展 MDS/骨髓衰竭 post-LTx,median survival 214 d — 嚴重警訊 [OE-Borie-2015-STSimmuno]。但無 study 直接比較 halved(500 BID)vs full(1000 BID)dose in STS;alternative agents belatacept / everolimus 被使用(Hannan 2025 belatacept for bone-marrow spared IS) [OE-Hannan-2025-BelaIPAH]。 - ATG avoidance 有強 mechanistic 基礎但缺 comparative data:ATG 導致 thrombocytopenia / leukopenia / neutropenia in standard populations;STS 骨髓 reserve 低,理論上 devastating。George 2019 LTx for IPF review;Newton 2017 telomere length in PF; Trindade 2019 alemtuzumab for CLAD [OE-George-2019-STSimmuno, OE-Newton-2017-STSimmuno, OE-Trindade-2019-STSimmuno]。 - Courtwright 2019 review “Telomeres in ILD: short and long of it” + Mackintosh 2025:STS 需 careful IS reduction + 考慮 bone marrow toxicity [OE-Courtwright-2019-STSimmuno, OE-Mackintosh-2025-STSgene]。

Verdict: 🟡 PARTIALLY SUPPORTED三項修改均為合理 expert consensus,但無 specific cohort 驗證 6-8 ng/mL、500 BID MMF 等精確 parameters;ISHLT 2025 consensus 支持「modified IS」概念但未給具體數字。手冊保留原文可,宜明確標示 evidence grade = expert consensus / observational(非 RCT)

手冊建議修訂: - Ch13:456 refine:「STS post-LTx IS 調整(ISHLT 2025 consensus + 多中心 observational, expert consensus level,無 RCT): - Tac target 6-8 ng/mL(vs 一般 8-12 ng/mL)—— balance IS 與 bone marrow toxicity,無 direct STS cohort 驗證 - MMF 減量(可減半至 500 BID 或停用)—— Borie 2015 French 多中心 6/9 telomerase mutation 發展 MDS/骨髓衰竭、median survival 214 d;MMF 最常被報告不耐受 - Avoid ATG induction(強 mechanistic 基礎:已知 hematologic 毒性 + STS 骨髓 reserve 低) - 考慮 belatacept / everolimus 作為 bone marrow-sparing alternative(Hannan 2025)」 - 加引用 [OE-Courtwright-2025-STSimmuno, OE-Borie-2015-STSimmuno, OE-Carlier-2025-STSimmuno, OE-Hannan-2025-BelaIPAH, OE-PhillipsHoulbracq-2022-STSgene]

Key citations: [OE-Courtwright-2025-STSimmuno], [OE-Borie-2015-STSimmuno], [OE-Carlier-2025-STSimmuno], [OE-Hannan-2025-BelaIPAH], [OE-PhillipsHoulbracq-2022-STSgene], [OE-Mackintosh-2025-STSgene], [OE-George-2019-STSimmuno]