Abstract
OBJECTIVE To investigate the natural history of fast developing esophageal and cardia precursors.
METHODS Repetitive endoscopic screenings were performed among 40-69-year-olds in the high-incidence areas for esophageal cancer in Shexian.
RESULTS The initial diagnosis and the lag-time for 7 subsequently identified severe dysplasia (SD) subjects were as follows: in one subject 13 months after a baseline diagnosis of normal epithelium, in another subject 7 months after a baseline diagnosis of base cell hyperplasia(BCH), in four subjects 3, 4, 4, and 10.5 months after baseline diagnosis of mild dysplasia (mD), and in one subject 12.5 months after a baseline diagnosis of moderate dysplasia (MD). The initial diagnosis and the lag-time for 6 subsequently identified carcinomas in situ or intramucosal carcinoma cases were: in one case 48 months after a baseline diagnosis of mD, in 2 cases 4 and 13 months after baseline diagnoses of MD, and in the other 3 cases 3.5, 9, and 17.5 months after baseline diagnoses of SD. The initial diagnosis and lagtime for 3 subsequently identified invasive cancer cases were: in one case 50 months after a baseline diagnosis of MD, in 2 cases 14 and 19 months after baseline diagnoses of SD. In addition, during a 4-year-follow-up of 18 subjects after endoscopic mucosa resection, 9 of them were found to have developed precursors again at other sites, and also additional findings were obtained for 11 of the 16 dysplasia cases by repetitive biopsy in less than 2 months after the initial endoscopy.
CONCLUSION A 5-year screening interval for BCH and mD, and a 3-year interval for MD may be too long for the fast developing precursors. Periodic screenings with shorter intervals should be considered to control the number of interval cases due to fast development, multifocal carcinogenesis, and false negative results inherent in one-time endoscopic biopsy sampling.
keywords
- esophageal cancer
- cardiac cancer
- endoscopic screening
- precancerous lesion
- intermittent time
- interval cases
Much attention at present is devoted to the control of esophageal cancer by endoscopic screening and mucosal resection in the high-risk regions in northern China. Cancer screening studies have demonstrated that the length of screening intervals is determined by the intermittent time of precancerous lesions, especially that of fast developing tumors[1,2]. Currently, endoscopic screening intervals for esophageal cancer in the high-risk regions were recommended to be every 5 years for the general population, for base cell hyperplastic (BSH), and the mild dysplastic (mD) subjects; and every 3 years for the moderate dysplastic (MD) subjects[3] based upon the results of a follow-up study performed by Dawsey et al.[4] in Linxian, China. The study was a pioneering work in esophageal cancer research by pinpointing MD and severe dysplasia (SD) as precancerous lesions for esophageal cancer. But the main follow-up results of the study should not be used as a base for screening intervals, because the 682 subjects followed since the end of 1987 by the authors was not a representative sample of the whole dysplastic population. There were another 72 subjects who had already developed into invasive cancer during the time since they were originally screened by balloon cytology from 1983 to 1987, then finally the remaining 682 non-cancerous subjects were chosen as cohort members and the follow-up begun. These 682 latent subjects might be argued as being “sluggish” in their development compared to the 72 cancer cases already developed, and therefore the intermittent time observed might be too long to be taken as screening intervals for the fast developing counterparts. This suspicion was confirmed by Wang et al[5] and our field screening practice. For this reason, we investigated the natural history of 301 subjects by repetitive screening in Shexian County of Hebei province, with the aim of observing the natural history of rapidly developing precursors.
MATERIAS AND METHODS
From March 2001 to June 2005, endoscopic screening was performed among villagers 40-69 years old in ten villages in Shexian County. After informed consent had been obtained, a trained epidemiologist interviewed the individual to be examined to fill out a baseline questionnaire form, and then experienced endoscopists from the Fourth Hospital of Hebei Medical University (i.e. the Hebei Provincial Tumor Hospital) performed the endoscopy with iodine staining to examine the esophagus, cardia, and the stomach until the duodenum was reached. Biopsy specimens were taken from all focal lesions and un-stained or lightlystained places. If no focal lesions were found, two standard sites, one at the mid-esophagus and the other at the lesser curve near the posterior wall of the cardia were sampled.
The biopsy specimens were fixed in 10% buffered formalin or 95% ethanol, embedded in paraffin, cut into 5-um sections, and stained with hematoxylin and eosin. Three pathologists separately read the biopsy slides. If no satisfactory slides were obtained or if the endoscopic appearance apparently disagreed with the histological diagnosis, a repetitive endoscopy shortly afterwards was offered. Invasive upper gastrointestinal carcinomas identified by the screening were referred to the Hebei Provincial Tumor Hospital for further treatment.
Severe dysplasia, carcinoma in situ or intramucosal cancer cases were endoscopically treated by endoscopic mucosal resection or argon plasma coagulation. For the convenience of repetitive screening, endoscopic working stations were usually set up in the local village. One month after the initial screening, participants were encouraged to undergone endoscopy again on their own choice, resulting in a screening interval ranging from one to 50 months among the 301 repetitively examined subjects.
RESULTS
Initial screening rate and acceptance of repetitive screening
In total, 1,514 participants accepted the initial endoscopy screening, achieving a first-round screening rate over 70% for all the 10 villages. The results of initial screening are shown in Table 1. Of those initially screened subjects, 26.0% (301/1,514) accepted repetitive screenings over 1 to 50 months. A comparison between the initial pathological diagnoses and subsequent diagnoses for the 301 subjects was made. The comparative results are shown in Table 2. For the 40 subjects whose pathological grades progressed as shown in Table 2, their detained development histories are summarized in Table 3.
Results of first-round screening of 1,514 subjects in Shexian County for upper gastrointestinal cancer and precancerous lesions.
Comparison of initial and repetitive screening results for 301 subjects in Shexian County.
Comparison of biopsy diagnoses between the initial and repetitive screening for 40 subjects with progressed precursors.
The initial diagnosis and the lag-time of 16 subsequently identified severe dysplastic (7), carcinoma in situ/intramucosal(6), and invasive cancer(3) cases.
There were cases of severe dysplasia; 6 cases of carcinoma in situ or intramucosal cancer; and 3 cases of invasive cancer identified by repetitive endoscoy after the initial examination. Since they were all detected within a short time of 50 months, they might be regarded as rapidly developing precancerous lesions in contrast with those remaining a long time without significant changes.
The initial diagnosis and the lag-time for the 7 severe dysplastic cases identified by repetitive endoscopies were as follows: in one case 13 months since a baseline diagnosis of “normal”, in another case 7 months since a baseline diagnosis of hyperplasia, in 4 cases 3 to 10.5 months since a baseline diagnosis of mild dysplasia; and in the other case 12.5 months since a baseline diagnosis of moderate dysplasia.
The initial diagnosis and the lag-time for the 6 carcinoma in situ or intramucosal cancer cases were as follows: in one participant in 48 months after an initial diagnosis of mind dysplasia; in 2 participants in 4 and 13 months after initial diagnoses of carcinoma in situ or intramucosal cancer; and in 3 participants in 3.5, 9, and 17.5 months after an initial diagnosis of severe dysplasia.
The initial diagnosis and the lag-time for the 3 cases of invasive cancer were as follows: in one case 50 months after an initial diagnosis of moderate dysplasia; in 2 cases in 14 and 19 months after an initial diagnosis of severe dysplasia.
DISCUSSION
Endoscopic screening with iodine staining can detect a considerable number of dysplasias and early cancers in the first round of screening: for instance, the detection rate in the present study was 2.1% and 1.1% for early cancer and severe dysplasia in the esophagus, and 0.6% and 0.3% for early cancer and severe dysplasia in the cardia respectively. However, there was still a large number of interval cases which occurred in practice. Interval cases may occur due to three reasons: 1) Some of the precancerous lesions may progress more rapidly than others. They are not easily detected by screenings, and tend to occur during the intervals. 2) A number of esophageal and cardia cancers result from multiple-site carcinogenesis, so local mucosal resections should not be expected to prevent cancer development in other sites. 3) The sensitivity of one time endoscopy and biopsy is limited by factors such as biopsy accuracy and the size of a specimen. Follow-up examination is needed to correct false-negative results.
By repetitive endoscopy and biopsy, we identified some subjects with fast developing precursors of esophageal/cardia cancer. For instance, 2 subjects initially diagnosed as normal and BCH were discovered to have developed into SD in 7 and 13 months respectively; and another 2 SD subjects were rediagnosed as carcinoma in situ and intramucosal cancer in 4 and 13 months. Wang et al.[5] followed 578 non-cancerous cases for 11 years by endoscopy. They observed an average time since entering into the study to cancer diagnosis of about 5.0±2.9 years for the 18 cases of male esophageal cancer, and 4.7 +/-3.2 years for the 7 cases of female esophageal cancer. Among the totally observed 25 cases, 11 had developed esophageal cancer from normal epithelium confirmed at the beginning. Although the interval of endoscopic screening in our report did not necessarily equal the lag time of the precursors, and there also existed the possibility of false negative results in the initial diagnoses, but both the present results and those reported by Wang et al.[5] support the idea that intermittent time of esophageal/cardia cancer is greatly varied among individual cases. For those with fast developing precursors, a screening interval of every five-years for the general population in the high risk region is likely to leave out a considerable number of interval cases.
The idea of multiple carcinogenic sites for esophageal/cardia cancer was supported by pathological, clinical and epidemiological observations. Wang et al.[6] reported that 4% (2/55) of the dysplasia cases and 47% (26/55) of the BCH cases were found to have precursors originating from more than one place. Song Shaoqian[7] examined 100 surgically resected specimens and found 94% to have unconnected BCH or carcinoma in situ. Recently, data from a population-based cancer registry[8] show that the formerly well-known esophageal cancer epidemic region in north China is actually a high risk region for both esophageal and stomach cardia carcinomas. Esophageal and cardia cancer are similarly distributed in the areas. The facts suggest that the high-risk places contain strong carcinogens for the whole upper gastrointestinal tract, and that this is the basis for multiple-site carcinogenesis.
In total, 64 subjects accepted repetitive endoscopy and biopsy within 2 months after the initial screening. Among them, 16 were initially identified as having precursors and were recommended by the endoscopist to accept re-examination within 2 months. Among the latter, additional findings were obtained for 11 by re-examination: one subject was initially diagnosed as “having esophagitis at 30 cm, mD at 34 cm,” but rediagnosed as “having MD or SD at the 32 cm, SD or carcinoma in situ at the 30 and 34 cm”. Wang et al.[9] once compared the pre- and 10-day-later endoscopic biopsy diagnoses and found that a consistent diagnosis was reached for only 52% of the samples. A first high but second lower diagnosis was found for 32% of the samples, and a first low but second higher diagnosis was found for 16% of the samples. These figures suggest that it is difficult to reach an accurate histological diagnosis based upon one-time endoscopy and biopsy alone, and the need of re-examination shortly afterwards is warranted.
Due to the limited number, it is unsafe to suggest any screening interval at present, but we believe that the 5-year interval for the general population, the BCH, the mD, and the 3-year interval for MD is too long according to our and other’s observations[5]. We suggest therefore that shorter intervals, for instance, every 2 to 3 year for BCH and mDs, and every half a year for MD be tried in practice. We also recommend that repetitive screenings take the place of massive screenings performed mostly once, conducted for all in the past, and that the screened population be put under a strict registry system for good evaluation of the effect by endoscopic screening in the future.
Footnotes
* This work was partially supported by Grants from the Hebei Provincial Natural Scientific Foundation (No.C2005000797) and from Funds for the Potentially Distinguished Scientific Project Construction in Hebei Universities.
- Received February 5, 2007.
- Accepted February 3, 2007.
- Copyright © 2007 by Tianjin Medical University Cancer Institute & Hospital and Springer
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.