口腔癌前病變之光動力及冷凍治療

Abstract

背景：口腔白斑（oral leukoplakia, OL）、紅白斑（oral erythroleukoplakia, OEL）及疣狀增生（oral verrucous hyperplasia, OVH）是三種常見的口腔癌前病變。本研究中，我們採用兩種保守性的治療方法，即以局部塗抹5–氨基酮戊酸為導引之光動力療法（5-aminolevulinic acid-mediated photodynamic therapy, ALA-PDT）與棉棒冷凍療法（cotton-swab cryotherapy, CSC），來治療這些病變。此外，我們也利用免疫組織化學染色的方法，來評估細胞凋亡相關蛋白在接受PDT前活體切片組織中的表現量，是否可以作為預測局部ALA-PDT療效的生物標記。 方法：本研究收集36例OVH、65例OL及20例OEL病變，以局部ALA-PDT治療，每週治療一次，另32例OL病變，以相同的局部ALA-PDT方式治療，每週治療兩次。我們以卡方檢定（chi-square test）來比較任兩組之臨床療效差異。同時我們也評估那些臨床及病理參數，會影響局部ALA-PDT對OVH病變的療效。我們將病變對治療的反應分成三類：完全反應（complete response, CR）、部分反應（partial response, PR）以及無反應（no response, NR）。本研究利用對抗Bak、Mcl-1、caspase-3、caspase-8、caspase-9、p53、p21及PCNA等蛋白的抗體與免疫組織化學染色法，來觀察18例OVH與40例OL病變，在接受PDT前活體切片組織中這些蛋白的表現量。我們分別記錄這些蛋白在細胞質與細胞核的標記指數（labeling indices, LIs）及染色強度（staining intensity, SI），並利用統計方法來比較CR與PR或NR兩組之間，在表面角質層厚度與標記分數（labeling score, LS，定義為 LI × SI）平均值的差異。本研究同時也使用CSC方法治療47位病人共60例OL病變，患者每兩週治療一次，直到病變完全消除為止。 結果：所有36例OVH病變，於平均3.8次的局部ALA-PDT治療之後，皆得到CR。65例OL病變，以局部ALA-PDT每週治療一次，發現有5例為CR，33例為PR，27例為NR。32例OL病變，以相同的局部ALA-PDT方式，每週治療兩次，發現有11例達到CR，21例呈現PR。另20例OEL病變，以局部ALA-PDT每週治療一次，發現有17例為CR，3例呈現PR。比較各組間之療效發現，32例OL病變以局部ALA-PDT每週治療兩次的療效，比65例OL病變以局部ALA-PDT每週治療一次的療效顯著較佳（P < 0.001）。另外，20例OEL病變以局部ALA-PDT每週治療一次的療效，比65例OL病變以局部ALA-PDT每週治療一次的療效也顯著較佳（P = 0.000）。此外，當OVH病變的外形為腫塊、最大徑小於1.5公分、外觀為粉紅色、上皮細胞發生變異，以及表面角質層厚度小於或等於40微米時，比起其外形呈現斑狀或合併中間為腫塊但周圍為斑狀（P = 0.000）、最大徑大於或等於1.5公分（P = 0.011）、外觀為白色（P = 0.000）、上皮細胞無變異（P = 0.043）或表面角質層厚度大於40微米（P = 0.003）者，需明顯較少的治療次數就能達到CR。以多變項統計分析顯示，OVH病變的外觀（P = 0.0069），為可用來預測局部ALA-PDT療效的唯一獨立因子。免疫組織化學染色的結果顯示，CR組別的Bak標記分數遠高於PR或NR組別（P = 0.003）。同時我們也發現CR組別的Bak/Mcl-1比值遠高於PR或NR組別（P = 0.02）。我們也觀察到CR組別的表面角質層厚度，遠低於PR或NR組別（P = 0.0036）。所有60例OL病變，在經過平均6.3次CSC治療之後，均呈現CR。當OL病變位於舌以外的口腔黏膜上、面積小於2平方公分、上皮發生變異、或表面角質層厚度小於55微米時，其達到CR所需要的治療次數，明顯少於其位於舌（P = 0.003）、面積大於或等於2平方公分（P = 0.024）、上皮無變異（P = 0.033）、或表面角質層厚度大於或等於55微米（P = 0.045）者。以多變項統計分析顯示，OL病變的位置（P = 0.000176）與面積（P = 0.021280）為影響CSC達到CR療效的獨立預測因子。 結論：局部ALA-PDT對治療OVH病變非常有效。當OVH病變最大徑小於或等於3.1公分，經由每週一次局部ALA-PDT治療，可在少於7次的療程時達到CR。局部ALA-PDT對OVH病變的療效，取決於病變的外形、大小、顏色、上皮是否變異，以及表面角質層的厚度。OVH病變的外觀是影響ALA-PDT療效的唯一獨立因子。OL病變以局部ALA-PDT每週治療兩次，比每週治療一次，有顯著較佳的療效。OEL病變以局部ALA-PDT每週治療一次，比OL病變以局部ALA-PDT每週治療一次的療效顯著較佳。OVH及OL病變在PDT之前的活體切片組織中，其Bak的表現量與Bak/Mcl-1表現量的比值，可當作預測PDT療效的生物標記。CSC對OL病變而言是一種簡單、安全、方便與保守的療法。當OL病變的表面積介於0.1至6.5（平均1.8）平方公分之間，以平均少於7次的CSC療程，即可達到CR。當OL病變位於舌以外的口腔黏膜上、上皮發生變異、以及具較薄的表面角質層厚度，達到CR所需要的治療次數，明顯少於病變位於舌頭、上皮無變異、或具較厚的表面角質層厚度者。

Background: Oral leukoplakia (OL), oral erythroleukoplakia (OEL), and oral verrucous hyperplasia (OVH) are three common oral precancerous lesions. In this study, we used two conservative treatment modalities, topical 5-aminolevulinic acid-mediated photodynamic therapy (ALA-PDT) and cotton-swab cryotherapy (CSC), for treatment of these lesions. In addition, we also used immunohistochemistry (IHC) to study whether the expression of apoptosis-associated proteins in oral precancerous lesions before PDT could be a biomarker to predict the treatment outcomes of topical ALA-PDT. Methods: Thirty-six OVH lesions, 65 OL lesions, and 20 OEL lesions were treated with topical ALA-PDT once a week and 32 OL lesions were treated with the same topical ALA-PDT twice a week. Their clinical outcomes between 2 different groups were compared by chi-square test. We also assessed what clinicopathological parameters of OVH lesions could influence PDT treatment outcomes. Lesion response was characterized into three categories: complete response (CR), partial response (PR), and no response (NR). IHC was performed with antibodies against Bak, Mcl-1, caspase-3, caspase-8, caspase-9, p53, p21, or PCNA protein in 18 OVH and 40 OL biopsy specimens taken before PDT. Both the labeling indices (LIs) and staining intensity (SI) of cytoplasmic or nuclear staining by each antibody were recorded. The means of surface keratin thickness or labeling score (LS, defined as LI × SI) were statistically compared between the CR and PR or NR group. Sixty OL lesions from 47 patients were treated with CSC once 2 weeks until CR of the lesion. Results: All 36 OVH lesions showed CR after an average of 3.8 treatments of topical ALA-PDT. The 65 OL lesions treated with topical ALA-PDT once a week showed CR in 5, PR in 33, and NR in 27. The 32 OL lesions treated with the same topical ALA-PDT twice a week demonstrated CR in 11 and PR in 21. The 32 OL lesions treated twice a week had a significantly better clinical outcome than the 65 OL lesions treated once a week (P < 0.001). The 20 OEL lesions treated with topical ALA-PDT once a week showed CR in 17 and PR in 3. The 20 OEL lesions treated once a week had a significantly better clinical outcome than the 65 OL lesions treated once a week (P = 0.000). In addition, OVH lesions with the clinical appearance of a mass, with the greatest diameter < 1.5 cm, with the pink color, with epithelial dysplasia, or with the surface keratin layer ≦ 40 μm needed significantly less mean treatment numbers of PDT to achieve a CR than OVH lesions with the clinical appearance of a plaque or a combination type of peripheral plaque and central mass (P = 0.000), with the greatest diameter ≧ 1.5 cm (P = 0.011), with the white color (P = 0.000), without epithelial dysplasia (P = 0.043), or with the surface keratin layer > 40 μm (P = 0.003), respectively. Multivariate analysis showed that only the clinical appearance of OVH lesions was the independent factor (P = 0.0069) to predict the PDT treatment outcome. IHC results revealed that the Bak LS was significant higher in the CR group than in the PR or NR groups (P = 0.003). A significant difference in the Bak/Mcl-1 LS ratio was also found between the CR and PR or NR groups (P = 0.02). We also showed a significant difference in the surface keratin thickness between the CR and PR or NR groups (P = 0.036). All 60 OL lesions treated with CSC showed CR after an average of 6.3 treatments. OL lesions on the oral mucosal sites other than the tongue, < 2 cm2, with epithelial dysplasia, or with the surface keratin thickness < 55 μm needed significantly less treatment number of CSC to achieve a CR than OL lesions on the tongue (P = 0.003), ≧ 2 cm2 (P = 0.024), without epithelial dysplasia (P = 0.033), or with the surface keratin thickness ≧ 55 μm (P = 0.045), respectively. Multivariate analyses showed that only the location (P = 0.000176) and area (P = 0.021280) of OL lesions were independent factors to influence the treatment number of cryotherapy to achieve a CR. Conclusion: Topical ALA-PDT is a very effective treatment modality for OVH lesions. For OVH lesions less than or equal to 3.1 cm in greatest diameter, CR of the lesions can be achieved by less than 7 treatments of topical ALA-PDT once a week. The PDT treatment outcome for OVH depends on the clinical appearance, size, color, epithelial dysplasia, and surface keratin thickness of the lesion. The clinical appearance of OVH lesions is the only independent factor affecting the PDT treatment outcome. OL lesions treated with topical ALA-PDT twice a week have a significantly better clinical outcome than OL lesions treated with the same PDT protocol once a week. OEL lesions treated with topical ALA-PDT once a week have a significantly better clinical outcome than OL lesions treated with the same PDT protocol once a week. The Bak LS and the Bak/Mcl-1 LS ratio in tissue sections of OVH and OL lesions before PDT can be used as biomarkers to predict the PDT treatment outcomes. CSC technique is a simple, safe, easy, and conservative treatment modality for OL lesions. For OL lesions with the surface area ranging from 0.1 to 6.5 (mean, 1.8) cm2, CR of the lesion can be achieved by less than 7 CSC treatments in average. OL lesions on oral mucosal sites other than the tongue, with dysplasia, and with thinner surface keratin layer needed significantly less treatment number of CSC to achieve a CR than OL lesions on the tongue, without dysplasia, and with thicker surface keratin layer, respectively.