Sorafenib in Advanced Hepatocellular Carcinoma 報告地點：萬芳醫院 報告時間：2009/11/6 報 告 者：和信醫院 李建勳 藥師 指導藥師：

Five A’s Assess：了解臨床的需求（Clinical Problem） Ask ：發現問題的所在（Question） Acquire ：找尋最好的資料（Best Evidence） Appraise ：分析資料（Validity, Importance） Apply ：應用在病人身上（Patient）

Clinical Situation Mr. Cheng is a 60 year old male patient. He was diagnosed as hepatitis B and C virus infection related hepatocellular carcinoma (HCC) on 2003. He has received radiotherapy during 2003/1/24~3/6, transarterial chemoembolization (TACE) during 2006/3/4~2007/12/4, computer tomography-guided radiofrequency ablation (RFA) on 2008/8/5. In 2009 April, abdominal sonography showed that HCC progress to retroperitoneal lymph nodes, which is suspected extrahepatic seeding tumor between abdominal wall and right lobe liver. At that time, the lab data showed elevated alpha-fetoprotein (AFP) 93.46 ng/mL. At that time, the doctor considered sorafenib 400mg PO BID for this patient. Is sorafenib effective for this patient?

Five A’s Assess：了解臨床的需求（Clinical Problem） Ask ：發現問題的所在（Question） Acquire ：找尋最好的資料（Best Evidence） Appraise ：分析資料（Validity, Importance） Apply ：應用在病人身上（Patient）

臨床問題的種類 背景問題（Background Questions）：對疾病的基本認識 5W1H：Who, Where, What, When, How, Why 疾病的某一面向 前景問題（Foreground Questions）：對病人的特定問題 處理某疾病病人的特定問題 每個病人有不同的特性：年齡、性別、伴隨疾病與疾病嚴重程度等

Epidemiology 肝癌是全球排名第五位的癌症，每年約有五十萬至一百萬的新病例。 肝癌更是死亡率第三位的癌症，全球每年約有六十萬人死於肝癌。 全球肝癌的盛行率有地理分佈的差異，低盛行區如北美和中美洲等，其盛行率約每10萬人口3-4 人；而亞太地區屬於肝癌高盛行區，盛行率高達每10萬人口30-40 人。 西元 2000 年全球肝癌患者，亞洲地區高達44 萬人，而北美地區只有1萬 2仟人，全球肝癌患者約 75%集中於亞洲地區。 肝癌為台灣 10 大癌症發生率的第二位，每10萬人口發生率男性約 26人、女性約8人，男女比例約 3:1。 林志陵, 高嘉宏. 肝癌的流行病學. 中華癌醫會誌. 2008; 24(5): 277-281

Risk Factors Liver cirrhosis Hepatitis B Virus Hepatitis C Virus Alcohol Cigarette smoking Betel nut Afaltoxin Obesity Fatty liver Non-alcohol steatohepatitis (NASH) Diabetes Hereditary tyrosinemia Hepatic porphyria Genetic hemochromatosis α1 antitrypsin deficiency 林志陵, 高嘉宏. 肝癌的流行病學. 中華癌醫會誌. 2008; 24(5): 277-281

Clinical Features Common Symptoms Common Physical Sign Abdominal pain Weight loss Weakness Fullness and anorexia Abdominal Swelling Jaundice Vomiting Common Physical Sign Hepatomegaly Hepatic bruit Ascites Splenomegaly Jaundice Wasting Fever DeVita VT, et al. Devita, Hellman & Rosenberg’s Cancer: Principles & Practice of Oncology, 8th Edition

Diagnosis Radiology Biopsy Alpha-fetoprotein (AFP) Serology X-ray Sonography Computed tomography (CT) Magnetic resonance imaging (MRI) Biopsy Alpha-fetoprotein (AFP) Serology DeVita VT, et al. Devita, Hellman & Rosenberg’s Cancer: Principles & Practice of Oncology, 8th Edition

Treatment Option Surgery Local Ablative Therapy Partial hepatectomy Liver transplantation Local Ablative Therapy Cryosurgery Microwave ablation Ethanol injection Acetic acid injection Radiofrequecy ablation (RFA) Regional Therapy (hepatic artery transcatheter treatments) Transarterial chemotherapy Transarterial embolization (TAE) Transarterial chemoembolization (TACE) Transarterial radiotherapy 90Y microspheres 131I lipiodol Conformal external-beam radiation therapy Systemic Therapy DeVita VT, et al. Devita, Hellman & Rosenberg’s Cancer: Principles & Practice of Oncology, 8th Edition

Staging System in HCC 除了腫瘤侵犯程度外，需考慮肝臟存留機能 病理報告取得不易，臨床分期應用廣泛 依應用目的區分預後的預測與治療方法的選擇

Staging System Prognosis Treatment TNM Okuda CLIP (Cancer of the Liver Italian Program) JIS (Japanese Integrated Score) Treatment BCLC (Barcelona Clinic Liver Cancer)

Staging System – TNM 盧勝男, 顏毅豪, 林芷芸等. 肝細胞癌之臨床分期.中華癌醫會誌. 2008; 24(5): 295-303

Staging System – Okuda

Staging System – CLIP CLIP: Cancer of the Liver Italian Program

Staging System – JIS JIS: Japanese Integrated Score

Staging System – BCLC BCLC: Barcelona Clinic Liver Cancer Bruix J and Sherman M. Hepatology 2005; 42(5): 1208-1236

Treatment Guideline – JSH JSH: Japan Society of Hepatology; TAI: transcatheter arterial infusion chemotherapy，也稱作hepatic arterial infusion (HAI) Kudo M. Oncology 2007; 72(supp1): 2-15

臨床問題的種類 背景問題（Background Questions）：對疾病的基本認識 5W1H：Who, Where, What, When, How, Why 疾病的某一面向 前景問題（Foreground Questions）：對病人的特定問題 處理某疾病病人的特定問題 每個病人有不同的特性：年齡、性別、伴隨疾病與疾病嚴重程度等

P. I. C. O. One-sentenced Question（請用一句話表達你的問題） As to advanced HCC patient, can sorafenib improve survival ? Patient or Problem （描述病患、疾病或病徵的型態） Hepatocellular carcinoma Intervention or Investigation （包括治療、檢驗、或預後因子） Sorafenib Comparison （通常用於比較目前的治療或診斷） Placebo Outcome （對病患有意義的臨床結果） Survival Type of Question ■ Therapy □ Diagnosis □ Prognosis □ Harm □ Others___________ Type of Study Design ■ Meta-analysis ■ RCT □ Cohort Studies □ Case Control Studies □ Case Series □ Case Reports

Five A’s Assess：了解臨床的需求（Clinical Problem） Ask ：發現問題的所在（Question） Acquire ：找尋最好的資料（Best Evidence） Appraise ：分析資料（Validity, Importance） Apply ：應用在病人身上（Patient）

搜尋策略：□ Systems □ Summary □ Synopses □ Synthesis □ Studies 關鍵字Primary or MeSH Term 同義字1 同義字2 P Hepatocellular carcinoma OR AND I Sorafenib C Placebo O Survival 搜尋結果 174 articles: limit to clinical trials (by reviewer) 1.Llovet JM, Ricci S, Mazzaferro V, et al. N Engl J Med 2008; 359(4): 378-90 (Phase III Trial) 2.Cheng AL, Kang YK, Chen Z, et al. Lancet Oncol 2009; 10(1): 25-34 (Phase III Trial) 3.Abou-Alfa GK, Schwart L, Ricci S, et al. J Clin Oncol 2006; 24: 4293-4300 (Phase II Trial) 4.Stumberg D, Clark JW, Awada A, et al. The Oncologist 2007; 12: 426-437 (Phase I Trial) 5.Furuse J, Ishii H, Nakachi K, et al. Cancer Sci 2008; 99: 159-165 (Phase I Trial) 6.Worns MA, Weinmann A, Pfingst K, et al. J Clin Gastroenterol 2009; 43: 489-495 (Small sample size, regardless of liver function, and negative results) 7.11 articles shows the efficacy of combination with other systemic chemotherapy 資料來源： □ UpToDate □Cochrane Library □ ACPJP ■ Pubmed □___ 關鍵性結論

Five A’s Assess：了解臨床的需求（Clinical Problem） Ask ：發現問題的所在（Question） Acquire ：找尋最好的資料（Best Evidence） Appraise ：分析資料（Validity, Importance） Apply ：應用在病人身上（Patient）

Sorafenib in Advanced Hepatocellular Carcinoma N Engl J Med 2008; 359: 378-90 Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, et al.

Study Design Study Design: Randomized, double-blinded, placebo-controlled trial Study Place: 121 centers in 21 countries in Europe, North America, South America, and Australia Intervention: sorafenib 400mg BID or placebo Inclusion Criteria: Advanced HCC, confirmed by pathological analysis ECOG performance status of 2 or less Child-Pugh liver function class A Adequate hematological (PLT, Hb), hepatic (Alb, Bil-T, GOT, GPT), and renal function (sCr) Exclusion Criteria: Previously received molecularly targeted therapy or systemic treatment

Outcome Assessment The Primary Outcome Overall Survival Time to Symptomatic Progression: FHSI8↓4, ECOG progress to 4, or death The Secondary Outcome Time to Radiological Progression: by RECIST The Disease-control Rate: CR+PR+SD Safety FHSI8 Questionnaire：用以評估HCC病人症狀的嚴重程度 CR: complete response; PR: partial response; SD: stable disease

Enrollment & Outcome 902 Patients were screened 300 were excluded 244 had protocol exclusion criteria 24 withdrew consent 15 had an adverse event 11 died 6 were lost to folloe-up 602 underwent randomization 299 were assigned to receive sorafenib (intention-to-treat population) 303 were assigned to receive placebo (intention-to-treat population) 1 had an adverse event 1 had a protocol violation 1 had a protocol violation 297 received sorafenib (safety population) 302 received placebo (safety population) 226 Discontinued sorfenib 86 had an adverse event 61 had radiologic and symptomatic progression 28 withdrew consent 3 died 1 had ECOG score of 4 47 had other reason 242 Discontinued placebo 90 had an adverse event 62 had radiologic and symptomatic progression 25 withdrew consent 7 died 6 had ECOG score of 4 52 had other reason 71 included in the ongoing study 60 included in the ongoing study Llovet JM, et al. New Engl J Med 2008; 359: 378-90

Demographic & Baseline Profile Variable Sorafenib (N=299) Placebo (N=303) Age – yr 64.9 ± 11.2 66.3 ± 10.2 Lung 67 (22) 58 (19) Sex – no (%) Macroscopic vascular invasion, extrahepatic spread, or both Male 260 (87) 264 (87) Absent 90 (30) 91 (30) Female 39 (13) Present 209 (70) 212 (70) Region – no (%) Child-Pugh class – no (%) Europe and Australia 263 (88) 263 (87) A 284 (95) 297 (98) North America 27 (9) 29 (10) B 14 (5) 6 (2) Central and South America 9 (3) 11 (4) Biochemical analysis Cause of disease – no (%) Albumin – g/dL Hepatitis C only 87 (29) 82 (27) Median 3.9 4.0 Alcohol only 79 (26) 80 (26) Range 2.7-5.3 2.5-5.1 Hepatitis B only 56 (19) 55 (18) Total bilirubin – mg/dL Unknown 49 (16) 0.7 Other 28 (9) 0.1-16.4 0.2-6.1 ECOG performance status – no (%) Alpha-fetoprotein – ng/mL o 161 (54) 164 (54) 44.3 99.0 1 114 (38) 117 (39) 0-208 x 104 0-5 x 105 2 24 (8) 22 (7) Previous therapy – no (%) BCLC stage – no (%) Surgical resection 57 (19) 62 (20) B (intermediate) 54 (18) 51 (17) Locoregional therapy C (advanced) 244 (82) 252 (83) transarterial chemoembolization 86 (29) Macroscopic vascular invasion – no (%) 108 (36) 123 (41) percutaneous ethanol injection 28(9) 20 (7) Extrahepatic spread – no (%) 159 (53) 150 (50) radiofrequency ablation 17 (6) 12 (4) Lymph nodes 89 (30) 65 (21) radiotherapy 13 (4) 15 (5)

Validity 可信度 本篇研究病人分配是否隨機？是 隨機分配是否對研究人員保密？是 本篇研究隨機分配後的病人是否都被納入分析？是 在接受治療當中，病人與醫療人員是否雙盲？是 除了要比較的治療外，兩組病人是否都被同等對待？是 一開始分配的兩組條件是否相同？是

Summary of Efficacy Measurement Outcome Sorafenib (N=299) Placebo (N=303) Hazard ratio (95% CI) P value Overall survival (mo) 0.69 (0.55-0.87) < 0.001 Median 10.7 7.9 95% CI 9.4–13.3 6.8–9.1 1-yr survival rate (%) 44 33 0.009 Time to symptomatic progression (mo) 1.08 (0.88-1.31) 0.77 4.1 4.9 3.5-4.8 4.2-6.3 Time to radiologic 0.58 (0.45 – 0.74) 5.5 2.8 4.1-6.9 2.7-3.9 Level of response (%) Complete NA Partial 2 1 0.05 Stable disease 71 67 0.17 Disease-control rate (%) 43 32 0.002 Llovet JM, et al. New Engl J Med 2008; 359: 378-90

Llovet JM, et al. New Engl J Med 2008; 359: 378-90

副作用 1.最重要的副作用： Grade 3/4 Hand-foot Skin Reaction Control event rate Experimental event rate Relative risk increase Absolute risk increase Number needed to harm CER EER RRI=(EER-CER)/CER ARI=EER-CER NNH=1/ARI 0/302= 0% 8/297= 2.693602% 37 ARI: 以sorafenib治療100人，會增加約3人三/四級手足症候群的機會。 NNH: 每治療37位病人，會有1人會發生三/四級手足症候群。 NNT (Number Need to Treat) or NNH (Number Need to Harm): 在一段實驗期間內，使一位病人達到實驗組治療之有益結果（或不良反應）所需治療的病人數目。

副作用 1.最重要的副作用： Grade 3/4 diarrhea Control event rate Experimental event rate Relative risk Increase Absolute risk increase Number needed to harm CER EER RRI=(EER-CER)/CER ARI=EER-CER NNH=1/ARI 2/302= 0.662251% 8/297= 2.693602% 3.063263 2.028649% 49 ARI: 以sorafenib治療100人，會增加2人三/四級腹瀉的機會。 NNH: 每治療49位病人，會有1人會發生三/四級腹瀉。 NNT (Number Need to Treat) or NNH (Number Need to Harm): 在一段實驗期間內，使一位病人達到實驗組治療之有益結果（或不良反應）所需治療的病人數目。

Cheng AL, Kang YK, Chen ZD, Tsao CJ, Qin SK, et al. Efficacy and Safety of Sorafenib in Patients in The Asia-Pacific Region With Advanced Hepatocellular Carcinoma: A Phase III Randomized, Double-blind, Placebo-Controlled Trial Lancet Oncol 2009; 10: 25-34 Cheng AL, Kang YK, Chen ZD, Tsao CJ, Qin SK, et al.

Enrollment & Outcome 271 Patients were screened 45 Not randomized 3 had an adverse event 4 withdrew 1 lost to follow-up 37 protocol exclusion criteria 226 underwent randomization 150 were assigned to receive sorafenib (intention-to-treat population) 76 were assigned to receive placebo (intention-to-treat population) 1 had a protocol violation 1 excluded due to adverse event 149 received sorafenib (safety population) 75 received placebo (safety population) 129 Discontinued sorfenib 69 had disease progresson 22 had adverse events 23 withdrew consent 12 died 2 lost to follow-up 1 non-compliant to treatment 72 Discontinued placebo 48 had disease progression 7 had adverse events 11 withdrew consent 2 died 3 lost to follow-up 1 protocol violation 20 included in the ongoing study 3 included in the ongoing study Cheng AL, et al. Lancet Oncol 2009; 10: 25-34

Demographic & Baseline Pofile

Validity 可信度 本篇研究病人分配是否隨機？是 隨機分配是否對研究人員保密？是 本篇研究隨機分配後的病人是否都被納入分析？是 在接受治療當中，病人與醫療人員是否雙盲？是 除了要比較的治療外，兩組病人是否都被同等對待？是 一開始分配的兩組條件是否相同？是

Summary of Efficacy Measurement Outcome Sorafenib (N=299) Placebo (N=303) Hazard ratio (95% CI) P value Overall survival (mo) 0.69 (0.55-0.87) < 0.001 Median 107 7.9 95% CI 9.4–13.3 6.8–9.1 1-yr survival rate (%) 44 33 0.009 Time to symptomatic progression (mo) 1.08 (0.88-1.31) 0.77 4.1 4.9 3.5-4.8 4.2-6.3 Time to radiologic 0.58 (0.45 – 0.74) 5.5 2.8 4.1-6.9 2.7-3.9 Level of response (%) Complete NA Partial 2 1 0.05 Stable disease 71 67 0.17 Disease-control rate (%) 43 32 0.002

Summary of Efficacy Measurement Outcome Sorafenib (N=150) Placebo (N=76) Hazard ratio (95% CI) P value Overall survival (mo) 0.68 (0.50-0.93) 0.014 Median 6.5 4.2 95% CI 5.56-7.56 3.75-5.46 6-mo survival rate (%) 53.3 36.7 NDA Time to symptomatic progression (mo) 0.90 (0.67-1.22) 0.50 3.5 3.4 2.80-4.24 2.40-4.08 Time to progression (mo) (by RECIST criteria) 0.57 (0.42-0.79) 0.005 2.8 1.4 2.63-3.58 1.35-1.55 Level of response (%) Complete Partial 5 (3.3) 1 (1.3) Stable disease 81 (54.0) 21 (27.6) Disease-control rate (%; 95% CI) 53 (35.3; 27.7-43.6) 12 (15.8; 8.4-26.0) 0.0019

副作用 1.最重要的副作用： Grade 3/4 Hand-foot Skin Reaction Control event rate Experimental event rate Relative risk increase Absolute risk increase Number needed to harm CER EER RRI=(EER-CER)/CER ARI=EER-CER NNH=1/ARI 0% 16/149= 0.107382 10.7% 9.3125 ARI: 以sorafenib治療100人，會增加約11人三/四級手足症候群的機會。 NNH: 每治療9個病人，會有1人會發生三/四級手足症候群。 NNT (Number Need to Treat) or NNH (Number Need to Harm): 在一段實驗期間內，使一位病人達到實驗組治療之有益結果（或不良反應）所需治療的病人數目。

副作用 1.最重要的副作用： Grade 3/4 diarrhea Control event rate Experimental event rate Relative risk Increase Absolute risk increase Number needed to harm CER EER RRI=(EER-CER)/CER ARI=EER-CER NNH=1/ARI 0% 9/146= 0.061643 6.143% 24.83 ARI: 以sorafenib治療100人，會增加6人三/四級腹瀉的機會。 NNH: 每治療24.83位病人，會有1人會發生三/四級腹瀉。 NNT (Number Need to Treat) or NNH (Number Need to Harm): 在一段實驗期間內，使一位病人達到實驗組治療之有益結果（或不良反應）所需治療的病人數目。

Comparison Between 2 RCTS SHARP Trial Cheng AL. Region of Study Europe. American Asia-Pacific HCV/Alcohol/HBV-related HCC(%) 28%/26%/18% 8%/?/73% ECOG PS: 0/1/2 (%) 54%/38%/7.6% 26%/68%/5% Extrahepatic spread (%) 51.3% 68.6% Clinical Outcome Sorafenib Placebo Overall Survival (mo) 10.7 7.9 6.5 4.2 Survival Rate (%) 44(1yr) 33(1yr) 53(6mo) 37(6mo) Time To Progression (mo) 4.1 4.9 3.5 3.4 Time To Radiologic Progression 5.5 2.8 1.4 Disease Control Rate(%) 43 32 53 12 HHR of Grade III/V HSR 37 9 HHR of Grade III/V Diarrhea 49 24 Comment by Cheng AL. et al: More advanced disease (extrahepatic spread, number of tumor, ECOG PS, alpha-feroprotein (AFP) level)

Sorafenib in HCC Treatment

Five A’s Assess：了解臨床的需求（Clinical Problem） Ask ：發現問題的所在（Question） Acquire ：找尋最好的資料（Best Evidence） Appraise ：分析資料（Validity, Importance） Apply ：應用在病人身上（Patient）

Practicability 目前Sorafenib (Nexavar®)在本院為臨採品項，健保給付規範如下：1. 晚期腎細胞癌且已接受interferon-alpha或 interleukin-2治療失敗，或不適合以上兩種藥物治療之病患。2.無效後則不給付temsirolimus及其他酪胺酸激酶阻斷劑（tyrosine kinase inhibitor, TKI）3.需經事前審查核准後使用，每次申請之療程以 3個月為限，送審時需檢送影像資料，每3個月評估一次 (98/10/1) 。病人需自費購買，自費價為一錠1387元，一天5547元。 研究結論應用在此病人時無已知可能的限制：根據病歷所記載資料，Child-Pugh A-B(未知腹水與腦病變程度)，BCLC Stage C (advanced stage)，亞洲人種，HBV/HCV-related HCC，大致符合臨床試驗收案條件。 依據鄭安理教授等人之亞太地區大型第三期臨床試驗的結果，我們預期此病人透過治療存活時間可延長2.3個月，但發生Grade 3/4的手足症候群、腹瀉的NNH分別為9和24。

Clinical Effectiveness & Safety 3/27~ 4/24~ 5/1~ 5/8~ 5/15~ 5/22~ 5/29~ 6/5~ 6/12~ 6/26~ 7/3~ 7/17~ 8/14~ Sorafenib 2# BID Sorafenib 3# QD AFP 93.46 104.7 74.37 49.74 43.04 37.94 46.03 48.38 56.51 100.9 126.6 148.1 242.2 5/1：watery diarrhea 6~7 times on day 1, muscle soreness, arthritis, chest wall pain after cough, loss of appetite, hand-foot skin reaction 5/8：hand-foot skin reaction & Gr 2-3 bullae at bilateral feet 5/15：hand-foot skin reaction & Gr 2-3 bullae at bilateral feet, skin rash at face and leg 5/22：Gr 2 hand-foot skin reaction, skin rash at face and leg↓ 5/29：Gr 2 hand-foot skin reaction, skin rash at face and leg↓, hot flash, hypertension 6/5：Gr 2 hand-foot skin reaction, skin rash at face and leg↓, hot flash, hypertension, general malaise,

Thank You For Your Listening

Reference 教科書 DeVita VT, et al. Devita, Hellman & Rosenberg's Cancer: Principles & Practice of Oncology, 8th Edition 臨床準則 AASLD Practice Guideline: Bruix J and Sherman M. Management of hepatocellular carcinoma. Hepatology 2005; 42(5): 1208-1236 JSH Guideline: Kudo M, Okunoue T, and Clinical Practice Manual of HCC Expert Panel. Management of hepatocellular carcinoma in Japan: consensus-based clinical practice manual proposed by the Japan Society of Hepatology. Oncology 2007; 72(supp1): 2-15 英文期刊 Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, et al. Sorafenib in Advanced Hepatocellular Carcinoma. N Engl J Med 2008; 359: 378-90 Cheng AL, Kang YK, Chen ZD, Tsao CJ, Qin SK, et al. Efficacy and Safety of Sorafenib in Patients in The Asia-Pacific Region With Advanced Hepatocellular Carcinoma: A Phase III Randomized, Double-blind, Placebo-Controlled Trial. Lancet Oncol 2009; 10: 25-34 中文期刊 林志陵, 高嘉宏. 肝癌的流行病學. 中華癌醫會誌. 2008; 24(5): 277-281 盧勝男, 顏毅豪, 林芷芸等. 肝細胞癌之臨床分期.中華癌醫會誌. 2008; 24(5): 295-303

Sorafenib in Advanced Hepatocellular Carcinoma N Engl J Med 2008; 359: 378-90 Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, et al.

Cheng AL, Kang YK, Chen ZD, Tsao CJ, Qin SK, et al. Efficacy and Safety of Sorafenib in Patients in The Asia-Pacific Region With Advanced Hepatocellular Carcinoma: A Phase III Randomized, Double-blind, Placebo-Controlled Trial Lancet Oncol 2009; 10: 25-34 Cheng AL, Kang YK, Chen ZD, Tsao CJ, Qin SK, et al.

Outline 5 Steps in EBM (5As) Assess Ask Background or foreground question, PICO Acquire Level of evidence, Searching skills Appraise Validity, importance, and practicability Apply

What is Evidence-based Medicine ? Evidence-based medicine is the integration of best research evidence with clinical expertise and patient values ~ Dr. Sackett, et al 2001 Patient Concerns Clinical Expertise Best research evidence EBM

臨床問題的種類 背景問題（Background Questions）：對疾病的基本認識 5W1H：Who, Where, What, When, How, Why 疾病的某一面向 前景問題（Foreground Questions）：對疾病的基本認識 處理某疾病病人的特定問題 每個病人有不同的特性：年齡、性別、伴隨疾病與疾病嚴重程度等

Ask Questions Background Question How does HBV infection progress to HCC？ How could we diagnose HCC？ How could we determine the stage of HCC？ When should we consider liver transplantation？ Foreground Question Is ultrasound more accurate than computed tomography？ Is sorafenib more effective than doxorubicin-based chemotherapy？

形成臨床問題常遇到的困難 不知道要如何開始 問題很多，時間很少 問不到想要的結果 先問背景問題，對疾病的基本認識 考慮問題的優先順序 問題結構的不清楚，使用PICO模式產生前景問題

Components of Clinical Questions 形成前景問題 - PICO Components of Clinical Questions Patient/ Problem Intervention Comparison Outcome 病人特性 檢查 觀察、安慰劑 住院天數 病人疾病 藥物治療 另一種藥物 生活品質 過去病史 外科治療 另一種手術 死亡率 … … … …

搜尋實證常見的障礙 資源太多，不知道從何下手 不知道怎麼定義關鍵字 不熟悉搜尋方式 只找到（看到）片面現象，而非貼近事實。 PICO，plus 同義字，MeSH Terms 不熟悉搜尋方式 只找到（看到）片面現象，而非貼近事實。

Levels of Evidence Oxford Centre for EBM Levels of Evidence http://www.cebm.net/index.aspx?o=1001

如何下手？ 實證醫學資源金字塔 越上層 資訊精粹 簡單的關鍵字 省時搜尋與評讀 由下層累積 越下層 文獻雜多 完整的關鍵字 費時搜尋與評讀 Systems 連結個別病歷的臨床知識與支援決策系統 越上層 資訊精粹 簡單的關鍵字 省時搜尋與評讀 由下層累積 越下層 文獻雜多 完整的關鍵字 費時搜尋與評讀 注重檢索技巧 資訊新穎 Summaries 整合証據提供特定臨床問題之概述與建議 UpToDate DynaMed BMJ Clinical Evidence FirstConsult ACP PIER Synopses 對單篇研究或回顧性文獻作摘要與評述 ACP Journal Club, Evidece-Based Medicine (PubMed, Ovid Medline) Syntheses 特定臨床問題的系統性評論文獻 Cochrane Database of Systematic Reviews Database of Abstracts of Reviews of Effects (PubMed, Ovid Medline): Systematic Reviews Studies 原始文獻 (PubMed, Ovid Medline, CINAHL, EMBASE Cochrane CENTRAL, Google scholar CEPS中文電子期刊, 中文期刊篇目索引)

Summaries型資料庫特性比較 資料庫 證據標示 特色 限制 更新 UpToDate [疾病/藥物] 70,000 Level: 1-2 Grade: A-C 查詢容易 很少標示證據等級 書目未直接連到PubMed 季 DynaMed 3,000 Level: 1-3 查詢容易，日更新 每頁首詳細註明新近更新 日 BMJ Clinical Evidence [疾病] 2,500 Type of evidence: 1-6 Grade: High,Medium,Low 評論過程透明 列出Clinical Question 表格呈現療效 強調盡可能說明Harm 有提供excluded reference 未提供etiology, diagnosis, cost 更新速度慢 採用英式英文 年 FirstConsult [疾病] 800 提供test或therapy順序 證據來源未逐項列明 月 ACP PIER [疾病] 400 呈現推薦、理由、證據 每頁首註明新近更新 主題數很少

資訊檢索 - 關鍵字 選用MeSH term （並避免雙引號、切截字元） 列舉同義字 避免使用縮寫 使用明確詞彙 調整廣狹義詞 COX-2 selective inhibitor >> COX-2 inhibitor 列舉同義字 Conventional >> conventional OR traditional 避免使用縮寫 CVD >> cardiovascular diseases 使用明確詞彙 Cardiovascular risk >> cardiovascular diseases AND risk 調整廣狹義詞 Cardiovascular diseases <<>> myocardial infarction

Cardiovascular disease AND risk 自由詞彙 或 控制詞彙 author Articles Cyclo-oxygenase-2 selective inhibitors and nonsteroidal anti-inflammatory drugs: balancing gastrointestinal and cardio-vascular risk. BMC Musculoskelet Disord 2007 Aug 3; 8: 73. Translation 1. Cardiovascular disease/ chemically induced* 2. Risk Standard (MeSH) Database Translation 1. Cardiovascular disease 2. Risk user Queries Cardiovascular disease AND risk

資訊檢索概念 單字 或 片語 Cardiovascular diseases OR "cardiovascular diseases" 切截字 Therap*: therapy, therapeutic, therapeutics 同義字（用MeSH可避免大部分的困擾） 單複數：inhibitor/inhibitors；英美拼法：edema/oedema；全名縮寫：cardiovascular disease/CVD；同義異形字：carotid ultrasonography/carotid ultrasound/carotid Doppler 廣狹義詞 Cardiovascular disease > myocardial infarction

資訊檢索 – 搜尋技巧 彈性組合檢索： 透過Detail檢驗是否選用MeSH term 透過Display: Citation線索修正關鍵字 P+I+C+O 透過Detail檢驗是否選用MeSH term 透過Display: Citation線索修正關鍵字 限定條件： Limit (年齡、性別、人類、研究設計) 滾出更多文章：Related Articles

Llovet JM, et al. Sorafenib in Advanced Hepatocellular Carcinoma Llovet JM, et al. Sorafenib in Advanced Hepatocellular Carcinoma. New Engl J Med 2008; 359: 378-90 研究設計 人/動物、年齡、性別

資訊檢索 - 關鍵字 選用MeSH term （並避免雙引號、切截字元） 列舉同義字 避免使用縮寫 使用明確詞彙 調整廣狹義詞 COX-2 selective inhibitor >> COX-2 inhibitor 列舉同義字 Conventional >> conventional OR traditional 避免使用縮寫 CVD >> cardiovascular diseases 使用明確詞彙 Cardiovascular risk >> cardiovascular diseases AND risk 調整廣狹義詞 Cardiovascular diseases <<>> myocardial infarction

資訊檢索概念 單字 或 片語 Cardiovascular diseases OR "cardiovascular diseases" 切截字 Therap*: therapy, therapeutic, therapeutics 同義字（用MeSH可避免大部分的困擾） 單複數：inhibitor/inhibitors；英美拼法：edema/oedema；全名縮寫：cardiovascular disease/CVD；同義異形字：carotid ultrasonography/carotid ultrasound/carotid Doppler 廣狹義詞 Cardiovascular disease > myocardial infarction

資訊檢索 - 關鍵字 選用MeSH term （並避免雙引號、切截字元） 列舉同義字 避免使用縮寫 使用明確詞彙 調整廣狹義詞 COX-2 selective inhibitor >> COX-2 inhibitor 列舉同義字 Conventional >> conventional OR traditional 避免使用縮寫 CVD >> cardiovascular diseases 使用明確詞彙 Cardiovascular risk >> cardiovascular diseases AND risk 調整廣狹義詞 Cardiovascular diseases <<>> myocardial infarction

資訊檢索 – 搜尋技巧 彈性組合檢索： 透過Detail檢驗是否選用MeSH term 透過Display: Citation線索修正關鍵字 P+I+C+O 透過Detail檢驗是否選用MeSH term 透過Display: Citation線索修正關鍵字 限定條件： Limit (年齡、性別、人類、研究設計) 滾出更多文章：Related Articles

彈性組合檢索

謹慎：使用limits後，in process、supplied by publisher、非Medline的文獻將被排除，因為此機制是針對已被Medline做過索引程序的書目所設計

Llovet JM, et al. Sorafenib in Advanced Hepatocellular Carcinoma Llovet JM, et al. Sorafenib in Advanced Hepatocellular Carcinoma. New Engl J Med 2008; 359: 378-90 研究設計 人/動物、年齡、性別

Appraise The Evidence Validity 可信度 Importance 重要性 Practicability 應用性

Validity Internal Validity: Is there a causal relationship between two variables? Criteria to establishing a cause-effect relationship: Temporal Precedence: the cause must happen before the effect. Co-variation: If X then Y, and if not X then not Y. Or a dose-response relationship: if more of the X then more of Y, and if less of X then less of Y. Without other Plausible Explanations: the problem with the “third variable” or “missing variable”. Other variable(s) causes the outcome. It is important but not always possible to “rule out” plausible explanations.

Validity 可信度 本篇研究病人分配是否隨機？ 隨機分配是否對研究人員保密？ 在接受治療當中，病人與醫療人員是否雙盲？ 除了要比較的治療外，兩組病人是否都被同等對待？ 一開始分配的兩組條件是否相同？ 本篇研究隨機分配後的病人是否都被納入分析？

Intention-to-treat Analysis Montori VM and Guyatt GH. CMAJ 2001; 165(10): 1339-1341

Validity 可信度 病人是否追蹤時間夠長且完整？ 所有研究對象追蹤的時間是否夠久來觀察到測量結果的發生？ 失去追蹤比率 < 5%，理想，偏差少 > 20%，嚴重影響實驗結果，通常結果不被接受。 假定一實驗有100人，4人死亡，如果有20人失去追蹤，出死亡率=4/80=5% 如果失蹤的20人皆存活，4/100=4% 如果失蹤的20人皆死亡，24/100=24%

Appraise The Evidence Validity 可信度 Importance 重要性 Practicability 應用性

Critically Appraisal 本篇文獻研究的結果在臨床上重要嗎？ 強度：本實驗是否有效或有害 效度：本實驗是否準確？ 有效： RRR: relative risk reduction ARR: absolute risk reduction NNT: number need to treat 有害： RR: relative risk OR: odd ratio 效度：本實驗是否準確？ 95% confidence interval (95%CI)

Appraise The Evidence Validity 可信度 Importance 重要性 Practicability 應用性

Practicability 如何將本文獻資料應用在我們的病人身上？ 該治療方式在我們醫療系統是否可行？ 該研究結論是否可應用在我們的患者？ 我們的患者透過該治療可能得到的好處與傷害（benefits and harms）？ 我們患者的價值觀及對結果的期待？

Practicability 目前Sorafenib (Nexavar®)在本院為臨採品項，無健保價，病人需自費購買，自費價為一錠1387元，一天5547元。 本研究結論應用在此病人時可能的限制為： SHARP Trial收納的研究對象為歐洲與美國的病人，發生HCC的原因以HCV與alcohol為主，而此病人為HBV-related HCC，研究結果可能不適用。 SHARP Trial的研究結果應用於亞洲國家人種時，會有基因上的差異，可能不適用。 依據SHARP Trial的研究結果，我們預期此病人透過治療存活時間可延長2.8個月，但發生Grade 3/4的手足症候群、腹瀉的NNH分別為37和49。

History of Evidence-based Medicine 1972年英國臨床流行病學者Archie Cochrane提出實證醫學 的概念。強調Randomized controlled trials 的重要性，並 認為所有醫療行為都應有嚴謹研究及證實為有效的根據， 才能將醫療資源做最有效的運用。 在近代，實証醫學這個名詞於1992年由加拿大McMaster大 學的Gordon Guyatt博士正式提出，隨後引起國際醫學界廣 泛關注。結合網路並強調以隨機對照實驗結果改進各種醫 療計劃，David L. Sackett於1993年成立Cochrane Collaboration 和 Cochrane Library。 摘自馬偕紀念醫院時正醫學中心 http://www.mmh.org.tw/taitam/medical_edu/www/default.asp?contentID=598

Randomized Controlled Trial (RCT) SUNY Downstate EBM Tutorial http://library.downstate.edu/EBM2/research.htm

Systemic Review SUNY Downstate EBM Tutorial http://library.downstate.edu/EBM2/research.htm

Cohort Study SUNY Downstate EBM Tutorial http://library.downstate.edu/EBM2/research.htm

Case Control Study SUNY Downstate EBM Tutorial http://library.downstate.edu/EBM2/research.htm

Case Series and Case Report SUNY Downstate EBM Tutorial http://library.downstate.edu/EBM2/research.htm

Practice 實用性 (1) 這個治療是否可行？是病患可否負擔？可 這個治療對你的病人有幫助嗎？有 評估我們的患者透過該治療可能得到的好處與傷害。 估計本國此疾病的發生率（PEER, patient’s expected event rate） PEER = 16.2% 估計方法 PEER = 論文中的CER

Practice 實用性 (2) 根據疾病的發生率（PEER），計算本治療在國內的NNT、NNH。 PEER method，假設RRR，RRI不受地區影響改變。 NNT = 1/(PEER x RRR) = 1/(0.162 x 0.13) = 47.48 NNH = 1/(PEER x RRI) = 1/(0.07 x 0.142) = 100.60 根據個人經驗粗估本疾病治療好處及副作用的危險性是論文中的 f 倍 ft = 1 NNTpatient = NNTreported/ft = 47.48 fh =1 NNHpateint = NNHreported/fh = 100.60

Practice 實用性 (3) 根據病人的期望，比較患者接受此治療的好處與壞處。 Sdisease（得到疾病的嚴重度）= 0.1 Sharm（進行治療的副作用）= 0.9 S(severity factor) = Sharm/Sdisease = 9 患者接受此治療得到的利害比（LHH, likelihood of being helped or harmed） 原始LHH = ARR: ARI = (1/NNT) : (1/NNH) = 2.12 Adjusted LHH = [(1/NNT) x ft x S] : [(1/NNH) x fh] = 19.08

Validity (1) Definition: the best available approximation to the truth or falsity of a given proposition, inference, or conclusion. Typology of validity Conclusion Validity (statistical conclusion validity): Is there a relationship between the two variables? The confidence we have on the conclusions we reach about relationships in our data (i.e. there is a significant correlation between two variables.) It concerns only whether such a relationship exists, but not whether such a relationship is a causal one. Internal Validity: Assuming that there is a relationship in this study, is the relationship a causal one? External Validity: Assuming that there is a causal relationship between the constructs of the cause and the effect, can it be generalized to other persons, places or times.

Validity (2) Typology of Validity (~ continued) Content Validity is evaluated by showing how well the content of the test samples the class of situations or subject matter about which conclusions are to be drawn. Criterion-related Validity is evaluated by comparing the test scores with one or more external variables (called criteria) considered to provide a direct measure of the characteristic or behavior in question. Predictive validity indicates the extent to which an individual’s future level on the criterion is predicated from prior test performance. Concurrent Validity indicates the extent to which the test scores estimate an individual’s present standing on the criterion. Construct Validity is evaluated by investigating what qualities a test measures, that is, by determining the degree to which certain explanatory concepts or constructs account for performance on the test.

Survival Analysis 生存時間是指從某個標準時間（如發病、確診、開始治療或手術的時間）算起至死亡為止的存活時間。 廣義的生存時間為研究開始至事件發生的時間。 生存研究的資料可分為完全資料（complete data）或截尾資料（censored data）。 因某種原因未能觀察到研究對象明確的研究結果，則不能確定該研究對象的確切生存時間，這類資料稱為截尾資料（censored data）。 雖然截尾資料提供的訊息是不完全的，但它告訴我們此研究對象歷經了一段時間仍然存活。

Interpretation of Survival Analysis Is the number of censoring small enough? (226+242)/(299+303)=77.74% Is the reason of censoring associated to the primary event? Adverse events, radiologic and symptomatic progression, withdrew consent, died of other reason, ECOG score of 4, and other reason

如何下手？ 實證醫學資源金字塔

前景問題 Foreground Questions 背景問題 Background Questions 經驗的累積