Shiu Ming-Neng, Chen Tony Hsiu-Hsi
Institute of Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan.
Oncol Rep. 2003 Nov-Dec;10(6):1683-92.
Elimination of leukoplakia is one of the strategies for prevention of oral cancer. However, the efficacy of reducing malignant transformation for the treated leukoplakia has rarely been reported despite a number of studies addressing its malignant transformation to oral cancer after intervention. The obstacle to do so is partially due to different lengths of follow-up and partially due to lack of information pertaining to the disease natural history that can be taken as a standard group for comparison. This work aimed to quantify the progression from leukoplakia to carcinoma with and without intervention after systematic literature review. The overall comparison of the efficacy of intervention across studies was therefore made. The literature between 1934 and 2001 was first reviewed, including two studies addressing the disease natural history model and 15 studies pertaining to intervention model in reducing malignant transformation. The simple constant-incidence exponential model and non-constant incidence models (Weibull and Quadratic models) were therefore applied to estimate annual malignant transformation after intervention in various studies. Annual transition rates pertaining to the disease natural history were also estimated by Markov models. Intervention efficacy index using 10-year cumulative incidence from both models was also developed to assess intervention efficacy across studies. For the disease natural history, the estimates of annual transition rate of leukoplakia, annual transition rate from leukoplakia to oral cancer in the PCDP, and annual transition rate from the PCDP to clinical phase were 0.00121 (0.00019-0.00150), 0.0605 (0.0436-0.0755), and 1.8797 (0.13242-2.4352), respectively. Similar findings were observed in another retrospective study. Annual malignant transformation rates after intervention range from 0.0003 to 0.062 assuming constant incidence. Studies assuming non-constant incidence show different patterns of increasing or decreasing risk with time. The estimates regarding the efficacy of intervention for each study with different follow-up periods ranged from 42.9 to 99.3%. The heterogeneity of intervention efficacy due to a wide range of annual rates of malignant transformation was demonstrated in this review. This implies that prevention program for reducing malignant transformation may vary with different areas. Factors need to be considered including different diagnosis criteria, different histological type, distributions of risk factor, different patient resources, and different compliance rates or insufficient medical intervention.
消除白斑是预防口腔癌的策略之一。然而，尽管有多项研究探讨了治疗后的白斑发生恶性转化为口腔癌的情况，但关于降低治疗后白斑恶性转化的疗效却鲜有报道。这样做的障碍部分是由于随访时间长短不同，部分是由于缺乏可作为对照标准组的疾病自然史相关信息。这项工作旨在通过系统的文献综述，量化有干预和无干预情况下白斑向癌的进展情况。因此，对各研究中干预疗效进行了总体比较。首先回顾了1934年至2001年的文献，包括两项关于疾病自然史模型的研究和15项关于降低恶性转化的干预模型的研究。因此，应用简单恒定发病率指数模型和非恒定发病率模型（威布尔模型和二次模型）来估计各研究中干预后的年度恶性转化率。还通过马尔可夫模型估计了与疾病自然史相关的年度转化率。还制定了使用两个模型的10年累积发病率的干预疗效指数，以评估各研究中的干预疗效。对于疾病自然史，白斑的年度转化率、PCDP中白斑向口腔癌的年度转化率以及PCDP向临床阶段的年度转化率估计分别为0.00121（0.00019 - 0.00150）、0.0605（0.0436 - 0.0755）和1.8797（0.13242 - 2.4352）。另一项回顾性研究也观察到了类似结果。假设发病率恒定，干预后的年度恶性转化率在0.0003至0.062之间。假设发病率不恒定的研究显示风险随时间增加或减少的模式不同。不同随访期的各研究干预疗效估计值在42.9%至99.3%之间。本综述表明，由于年度恶性转化率范围广泛，干预疗效存在异质性。这意味着降低恶性转化的预防计划可能因不同地区而异。需要考虑的因素包括不同的诊断标准、不同的组织学类型、危险因素分布、不同的患者资源以及不同的依从率或医疗干预不足。
