乾癬與治療的相關致癌危險性探討

Abstract

研究背景 乾癬是一種慢性疾病，目前尚未有根治的治療方式，病人需要終身接受治療。且病情嚴重的病人需要長期使用全身性的藥物或光照療法才能緩解病情。 國外有研究顯示未接受過乾癬全身性藥物或治療的乾癬病人罹癌的危險性略高於一般族群，雖然無法斷言是乾癬本身引起的，但乾癬病人長期使用這些已證實或被懷疑有致癌性的全身性藥物或治療，是否會加成致癌之危險性是個值得深入探討之課題。 研究目的 研究不同疾病嚴重度之乾癬病人發生非黑色素瘤皮膚癌、淋巴癌及黑色素瘤的間接標準化發生率比。並利用Time-dependent Cox 迴歸模型探討乾癬治療型態與非黑色素瘤皮膚癌發生的相關性。 研究材料與方法 以全民健康保險100 萬人承保抽樣歸人資料庫西元1997年至2007年間住院醫療費用清單明細檔和門診處方及治療明細檔之診斷欄位ICD-9碼為696、696.0或696.1之乾癬診斷的病人定義為乾癬世代。進入世代後被診斷為非黑色素瘤皮膚癌、黑色素瘤或淋巴癌的乾癬病人，且申報資料中有與診斷相符的治療者，定義為非黑色素瘤皮膚癌、黑色素瘤或淋巴癌個案。並按照是否曾接受過全身性治療將病人分為嚴重型乾癬及輕微型乾癬。以2005年100萬人承保抽樣歸人檔作為參照族群，計算乾癬世代相較於參照族群之癌症間接標準化發生率比。 校正年齡、性別、察爾森合併症嚴重度指標及砷暴露後，以Time-dependent Cox 迴歸模型分析不同事件發生點之各種治療型態變項 (PUVA療法、紫外線B光光線治療、Methotrexate、Cyclosporin、Retinoids、Azathioprine及Tar )累積治療次數及劑量對發生非黑色素瘤皮膚癌的影響，並針對Retinoids累積治療劑量計算之時間點進行敏感度分析。 結果 本研究定義之乾癬世代病人7061位，嚴重型乾癬的病人共870位，輕微型乾癬治療的病人共6191位。 校正年齡及性別後，大於40歲族群之非黑色素瘤皮膚癌間接標準化發生率比為4.29 (95% CI 2.90-6.35)。輕微型乾癬的非黑色素瘤皮膚癌間接標準化發生率比為3.72 (95% CI 2.34-5.90)。嚴重型乾癬的非黑色素瘤皮膚癌間接標準化發生率比為7.08 (95% CI 3.38-14.85)。黑色素瘤間接標準化發生率比為3.75 (95% CI 0.94-14.99)。輕微型乾癬的黑色素瘤間接標準化發生率比為2.26 (95% CI 0.32-16.04)。嚴重型乾癬的黑色素瘤間接標準化發生率比為11.01 (95% CI 1.55-78.16)。大於20歲的淋巴癌間接標準化發生率比為2.30 (95% CI 1.15-4.60)。輕微型乾癬的淋巴癌間接標準化發生率比為1.75 (95% CI 0.73-4.20)。嚴重乾癬的淋巴癌間接標準化發生率比為4.85 (95% CI 1.56-15.04)。 在探討各種治療型態變項對發生非黑色素瘤皮膚癌的影響的Time-dependent Cox 迴歸模型中，包含6737位新診斷的乾癬病人，出現22個非黑色素瘤皮膚癌的事件。除年齡愈大及曾有砷暴露會增加非黑色素細胞瘤皮膚癌發生的危險性外，Azathioprine之累積治療劑量會增加非黑色素細胞瘤皮膚癌發生的危險性 (Hazard ratio 1.34) 。其他治療型態的累積治療次數及劑量對於非黑色素瘤皮膚癌的發生皆不具統計上的顯著意義。 討論 輕微型乾癬病人罹患非黑色素瘤皮膚癌的危險性高於參照族群，暗示著乾癬疾病本身具有致癌性。病情嚴重的乾癬病人罹患非黑色素瘤皮膚癌、黑色素瘤及淋巴癌的發生率略高，則可能與乾癬疾病嚴重度或治療型態相關。 目前關於Azathioprine的非黑色素瘤皮膚癌致癌性的研究大都是報告在類風溼性關節炎、器官移植或發炎性大腸疾病的病人。尚未有以Azathioprine治療乾癬引起非黑色素瘤皮膚癌的大型研究。 PUVA療法在西方國家的文獻中有顯著的非黑色素瘤皮膚癌致癌性，但在亞洲國家的文獻中相對危險性卻不若歐美國家那麼高。本研究PUVA療法對非黑色素瘤皮膚癌致癌危險性不顯著，除了種族及膚色差異外，累積治療次數偏低應該也是一個重要因素。至於Methotrexate及Cyclosporin治療的非黑色素瘤皮膚癌致癌危險性不顯著，可能與追蹤時間不夠長有關。 結論 接受過乾癬全身性藥物或PUVA療法的乾癬病人罹患非黑色細胞瘤皮膚癌、黑色素瘤及淋巴癌的危險性均高於一般族群。未曾接受過全身性藥物或PUVA療法的乾癬病人罹患非黑色細胞瘤皮膚癌的危險性高於一般族群，因此乾癬疾病本身即具有致非黑色細胞瘤皮膚癌的危險性。醫師在規劃病人治療計畫時須謹慎衡量利弊得失。本研究已證實全身性藥物治療中azathioprine與乾癬病人發生非黑色細胞瘤皮膚癌有關，乾癬病人應謹慎使用並減少紫外線曝曬。

Background Psoriasis is a kind of chronic diseases which cannot be cured at present. Patients may need lifelong treatment. Moreover, patients in serious conditions may require long-term use of systemic drugs or phototherapy to ease the condition. Several studies have shown that psoriasis patients who do not receive systemic treatment have slightly higher risks of getting cancer than the general population. Although psoriasis is not confirmed to lead to cancer, it is worthwhile to discuss the issue more deeply concerning whether using systemic treatment proved or suspected to be carcinogenic increases the risks of getting cancer for psoriasis patients. Study Purpose The study was aimed at discovering the Indirect Standardized Incidence Ratio (SIR) of nonmelanoma skin cancer (NMSC), lymphoma and melanoma happening on patients with various disease severity of psoriasis. The study would also explore the relevance between different treatment types and the occurrence of NMSC by means of Time-dependent Cox regression mode. Material and Method The cohort of psoriasis was defined based on the administrative claims data of a randomly sampled cohort (n = 1,000,000) of National Health Insurance beneficiaries in Taiwan with the diagnostic column of ICD-9 code being 696, 696.0 or 696.1. Cases of NMSC, melanoma or lymphoma were defined when patients were diagnosed with the above diseases after getting into the cohort and treated with compatible treatment. Psoriasis patients were then divided into two groups: severe or mild in accordance with the experience of receiving systemic treatment. We calculate SIR of cancer for each study group. After adjusting the age, gender, Charlson comorbidity index and arsenic-exposure, Time-dependent Cox regression model was then applied to analyze variables of treatments such as PUVA therapy, UVB therapy, Methotrexate, Cyclosporin, Retinoids, Azathioprine and Tar to discover how the timing of different events and the accumulated times of treatment as well as the amount of doses affected NMSC. The study also conducted sensitivity analysis of the timing of the accumulated doses of Retinoids. Results There were 7061 patients defined as the generation of psoriasis, among which 870 were severe psoriasis patients and 6191 mild were ones. After adjusting age and gender, the SIR of NMSC was 4.29 (95% CI 2.90-6.35) while the SIR of mild psoriasis was 3.72 (95% CI 2.34-5.90) and 7.08 (95% CI 3.38-14.85) of the SIR of severe psoriasis. The SIR of melanoma was 3.75 (95% CI 0.94-14.99) while the SIR of mild psoriasis was 2.26 (95% CI 0.32-16.04) and 11.01 (95% CI 1.55-78.16) of the SIR of severe psoriasis. The SIR of lymphoma was 2.30 (95% CI 1.15-4.60) while the SIR of mild psoriasis was 1.75(95% CI 0.73-4.20) and 4.85 (95% CI 1.56-15.04) of the SIR of severe psoriasis. There were 6737 patients newly diagnosed with psoriasis, among which 22 cases were diagnosed with NMSC in the Time-dependent Cox regression model of discovering the effects of variables of treatments on the occurrence of NMSC. In addition to age and arsenic-exposure, the more the accumulated doses of Azathioprine were, the higher risks of the occurrence of NMSC would be. There was no statistical significance of how the accumulated times and doses of the other treatment affected the occurrence of NMSC. Discussion Mild psoriasis patients had higher risks of getting NMSC than the reference group which implied psoriasis itself was carcinogenic. The occurrence rate of NMSC, melanoma and lymphoma was slightly higher for severe psoriasis patients, which was possibly related to the severity of psoriasis or its treatments. Studies concerning the risks of getting NMSC with Azathioprine were mainly about patients with rheumatoid arthritis, organs transplantation or inflammatory bowel disease. There were still no large scale researches regarding NMSC caused by psoriasis treated with Azathioprine. PUVA therapy was significantly carcinogenic in getting NMSC in the West literatures while its risks were comparatively lower in Asian studies. The risk of getting NMSC with PUVA therapy in our study is not significant. In addition to the difference of races and skin types, the accumulated times of treatment tend to being low may be also a critical factor. As for the treatment of NMSC with Methotrexate and Cyclosporin, the risks of getting cancer were not significant, which may be the result of insufficient tracing time. Conclusion The risks of getting NMSC, melanoma and lymphoma for severe psoriasis patients were higher than the reference group. The risks of getting NMSC for mild psoriasis patients were higher than the reference group. Therefore, psoriasis itself was carcinogenic of getting NMSC. The study had proved that the occurrence of NMSC was related to psoriasis patients’ using Azathioprine. Consequently, psoriasis patients should use the drugs cautiously and avoid being exposed to the ultra light.