|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Risk of Second Primary Malignancy After Esophagectomy for Squamous Cell Carcinoma of the Thoracic EsophagusFrom the Department of Surgery, Cancer Institute Hospital, Tokyo, Japan. Address reprint requests to Toshiki Matsubara, MD, PhD, 1-12-7 Kohinata, Bunkyou-Ku, Tokyo 112-0006, Japan; e-mail: matsu-bara{at}umin.ac.jp.
Purpose: To assess the risk of subsequent malignancies after esophagectomy for squamous cell carcinoma of the thoracic esophagus for the establishment of an adequate follow-up program. Patients and Methods: We statistically analyzed clinical factors in 114 cases of second malignancy after esophagectomy that developed in 94 of 679 patients who underwent curative resection. The cancer incidence rates in the general population estimated by the Research Group for Population-Based Cancer Registration in Japan were used as standards for comparison. Results: The 10-year cumulative risk of second malignancy was 34.5%, and the overall relative risk (RR) was 2.98 (95% CI, 2.41 to 3.65). The risk of head and neck cancer was markedly elevated (RR, 34.9; 95% CI, 24.3 to 48.6), followed by the risks of lung cancer (RR, 3.24; 95% CI, 1.89 to 5.19) and stomach cancer (RR, 2.00; 95% CI, 1.17 to 3.21). Multifactor analysis demonstrated that independent factors affecting the risk of subsequent malignancies were presence of other malignancies detected before esophagectomy and any of the following factors: masculine sex, alcohol consumption, and smoking. The 5-year survival rate after detection of subsequent malignancy was 45%. The outcome in patients with subsequent head and neck cancer was significantly less favorable as a result of difficulty in early detection. Conclusion: Patients had a remarkably high risk of subsequent cancer of the upper aerodigestive tract after esophagectomy, in particular, head and neck cancer. Minute postoperative surveillance is strongly recommended, especially in patients with a history of malignancies before esophagectomy. Early detection of second malignancies allowed less invasive treatment with favorable outcome.
IN CANCER of the thoracic esophagus, the outcome of patients subjected to esophagectomy has recently improved as a result of progress in early detection techniques, surgical procedures, and perioperative adjuvant therapies.1,2 As the number of long-term survivors after esophagectomy has increased, it has been noted that the occurrence rate of second malignancies after esophagectomy is markedly high.3–6 The risk of second malignancies has now become one of the major prognostic factors after curative treatment. Because most patients with esophageal cancer were considered as incurable until recently,7 second malignancies after esophageal surgery have seldom been systematically investigated based on records of a representative number of long-term survivors. In the present study, we analyzed frequencies and characteristics of malignancies that were detected after esophagectomy to establish adequate follow-up strategies for such patients.
Of 753 patients who underwent esophagectomy for squamous cell carcinoma of the thoracic esophagus from 1985 through 2001, 679 underwent curative resection (Table 1  ). Curative resection was defined as an operation without macroscopically recognizable cancer lesions left behind at resection. The records of the 679 patients subjected to curative resection were analyzed in terms of risk of developing subsequent second malignancies, which were defined in patients with esophageal cancer as primary malignancies in other organs that were detected after esophagectomy. Independent malignant lesions detected before or at esophagectomy were classified as antecedent second malignancies. Cases with multicentric esophageal cancer in the resected specimens were excluded from the category of second malignancy. However, primary cancer that developed in the residual cervical esophagus after esophagectomy was incorporated in the present study, because it has similar clinical significance in postoperative follow-up observation. Malignant lesions that were not strictly distinguished from recurrent lesions of the original esophageal cancer were excluded from the analysis. Particularly, none of 60 patients who developed squamous cell carcinoma in the lung within 5 years after esophagectomy were included in the group of subsequent second malignancies, because esophageal cancer commonly metastasizes to the lung, and it was difficult to select out primary tumors on rigorous clinicopathologic criteria. Of these 60 patients, 16 underwent pulmonary resection, and seven patients survived more than 3 years after the detection of the pulmonary lesions.
The degree of disease progression was evaluated by the depth of wall invasion (International Union Against Cancer T factor) and the number of involved lymph nodes. The extent of a patient’s smoking and drinking habits was retrospectively assessed from the records in the routine questionnaire patients had to answer at the time of admission for esophagectomy. The degree of smoking exposure was defined as the product of the number of cigarettes consumed per day and the smoking duration in years. Alcohol exposure was assessed simply by the grams of alcohol consumed per day calculated from the records, regardless of the kind of drink. Before esophagectomy, we routinely performed gastroesophageal endoscopy, gastroesophagography, computed tomography of the chest and upper abdomen, and surveillance of the head and neck region not only to assess the extension of esophageal cancer but also to search for other malignancies. Patients were routinely followed up by physical examination, blood analysis, and computed tomography every 6 months and by endoscopy every year. As regards statistical procedures, the status of all patients was examined between March and September 2002. No patients were lost at this follow-up time. Detection of subsequent malignancies, death irrespective of its causes, and the latest survival certification were regarded as events. The cumulative risks and their SEs were calculated by the Kaplan-Meier method. In the analysis of risks of second malignancies after esophagectomy, the end point was defined as the time when the second malignancy was detected. Patients who were alive or dead without developing other malignancies after esophagectomy were treated as censored at the time of follow-up examination and death, respectively. The follow-up of the 679 patients for subsequent second malignancy was based on a total of 2,591 person-years. In the survival analysis, the time of death was considered as the end point, irrespective of the cause of death. Tests for significance of the difference in cumulative risks were performed by the log-rank method. The cumulative incidence rates of malignancies in a group of general Japanese population with the same calendar-period, age, and sex as the object groups of patients were calculated from the age- and sex-specific cancer incidence rates estimated by the Research Group for Population-Based Cancer Registration in Japan (a study supported by a Grant-in-Aid for Cancer Research from the Japanese Ministry of Health and Welfare).8 Because registration data after 1997 were not available in this database, the data from 1997 were used as substitutes for incident rates in 1998 and later years. These registration data were also used as standards to calculate relative risks of second malignancies. The relative risk ratio was defined as the ratio of the percentage of patients with an event in the object group to the expected percentage of persons with an event among the general population group with the same person-year distribution. Relative risk rates and their 95% CIs were computed using PAMCOMP (http://medweb.uni-muenster.de/institute/epi/pamcomp/pamcomp.html; Institute of Epidemiology and Social Medicine, University of Muenster, Muenster, Germany) software.9 The Cox proportional hazards model was used to estimate the effect of suspected factors on the risk of subsequent second malignancies.
Second malignancies were found in 254 of 679 patients subjected to curative esophagectomy. Among them, 94 patients had a total of 114 subsequent primary malignancies, and 194 patients had antecedent primary malignancies. In 33 of them, multiple second malignancies were found in both preoperative and postoperative periods. Subsequent malignancies were most frequently found in the head and neck region, followed by the stomach, lung, residual esophagus, and colorectal region (Table 2 ). The subsequent lung cancer was adenocarcinoma in nine patients, squamous cell carcinoma in six, and small cell carcinoma in two. Among antecedent malignancies, head and neck cancer was the most common, followed by stomach cancer and colorectal cancer.
For the total of 679 patients, the cumulative risk of developing second primary malignancy 5 and 10 years after esophagectomy was 16.1% (SE, 2.0%) and 34.9% (SE, 3.6%), respectively, whereas the expected cumulative risk of malignancy for the general Japanese population matched for calendar-age and sex was 5.8% and 12.9% at 5 and 10 years, respectively (Fig 1 ). The overall relative risk was 2.98 (95% CI, 2.41 to 3.65).
The risks of head and neck cancer, stomach cancer, and lung cancer were all significantly higher in patients subjected to esophagectomy as compared with those in the general population (Table 3 ). The incidence rate of head and neck cancer was outstandingly high. The 5-year cumulative risk of head and neck cancer after esophagectomy was 7.3%, whereas the 5-year cumulative risks of stomach and lung cancers were less than 2% (Fig 2).
In the single factor analysis for cumulative risk of subsequent second malignancies, sex, history of other malignancies detected before surgery for esophageal cancer, and drinking and smoking habits were significant factors (Table 4 ). Although the risk of developing subsequent malignancies was significantly low in patients who did not smoke or drink, the amount of cigarettes or alcohol was not correlated with the cumulative risk of developing a second malignancy in smokers and drinkers. The presence of multicentric lesions in the resected esophagus demonstrated borderline significance.
There were exceptionally significant mutual correlations among three binary categorizations: sex, drinking habits (yes, no), and smoking habits (yes, no; P < .0001 in each pair;  2 test). Therefore, analysis by the Cox proportional hazards model with factors including two or all of them demonstrated that only history of other previous malignancy was a significant factor that affected the risk of developing other malignancies after esophagectomy. However, analysis with factors including one of these three categorizations showed that any of them was another significant factor (Table 5).
The cumulative risks of developing subsequent second malignancy within 5 years after esophagectomy in male and female groups were 18.0% (SE, 2.3%) and 8.3% (SE, 4.2%), respectively, whereas the risks of malignancy within 5 years in general male and female populations matched for calendar-age were 6.2% and 3.4%, respectively (Fig 3 ). The difference in cumulative risk in males was statistically significant. The cumulative risks in patients with both drinking and smoking habits are shown in Fig 4. The long-term risk in patients without drinking or smoking habits was similar to the general population.
Among patients who had had other malignancies preceding esophagectomy, the risk of subsequent malignancy after esophagectomy was 5.0 times higher than the general population (95% CI, 3.47 to 7.08). The cumulative risk at 10 years after esophagectomy was 51.0% (SE, 9.0%; Fig 5 ). The subsequent malignancy risk was still high in patients without antecedent second malignancies, the overall relative risk of subsequent malignancy being 2.44 (95% CI, 1.87 to 3.14). The relative risks of developing head and neck, stomach, and lung cancer after esophagectomy in patients with and without antecedent malignancies are shown in Table 6. The risk of head and neck cancer after esophagectomy was markedly high, even in patients without antecedent malignancy.
The survival outcome after detection of a subsequent second malignancy was not so unfavorable (Fig 6 ). The overall 5-year survival rate was 45%. The outcome of patients who developed head and neck cancer was significantly less favorable than the outcome of patients who developed other malignancies (P = .02).
Evidence shows that the association between esophageal cancer and malignancies developing in other organs is not rare.3,4,10 Although second primary esophageal cancer developing after or simultaneously with malignancy of other organs has been occasionally reported, the nature of second malignancy after esophageal cancer has not been minutely analyzed because the outcome of esophagectomy used to be unfavorable. However, a considerable proportion of patients with esophageal cancer are now cured and survive long thanks to progress in cancer screening, preoperative diagnosis, and treatment. The 5-year survival rates after esophagectomy are approximately 50% in most prominent institutions in Japan.3,11 Technical progress in early detection of other cancers in preoperative studies may also have contributed to the improved survival after esophagectomy. Even in 103 patients who had simultaneous other malignancies at esophagectomy, the 5-year survival rate was 46% in our series. We now commonly encounter patients who develop other malignancies after esophagectomy. Knowledge concerning second malignancies is essential to identify populations and target organs at a high risk of developing a second malignancy that require inclusion in a screening program for subsequent malignancies. There have been few studies that have assessed the exact risk ratio of a second malignancy after esophageal cancer based on a representative number of patients with precise and accurate clinical records and long follow-up periods. In the present series, the risk of second malignancy after esophageal cancer was three-fold compared with the risk of malignancy in the general population. The cumulative risk at 10 years after esophagectomy was 35%. Hoar et al5 reported a relative risk of 1.78 for any second cancer after esophageal cancer, the average follow-up period being less than 1 year. Although it is possible that the markedly higher second cancer risk in the present series has some correlation with factors specific to the Japanese population, it is difficult to evaluate the difference because of the short follow-up periods in the other series. The difference in the histologic spectrum of esophageal cancer might contribute to this difference. Although adenocarcinoma has become more common in Western countries, we have still only a limited number of patients with adenocarcinoma of the thoracic esophagus. It is well known that the most common site for synchronous and metachronous second primary malignancy associated with esophageal cancer is the upper aerodigestive organs that include the head and neck, the stomach, and the lung.3–6,10 In particular, an exceptionally strong association between esophageal cancer and head and neck cancer has been reported.3,10,12 However, most previous reports have included a limited number of patients with metachronous primary cancer subsequent to esophageal cancer. The present study demonstrated that sites of second primary malignancies after esophagectomy also presented a similar distribution pattern as sites of malignancies preceding esophageal cancer. Head and neck cancer was the most common second primary malignancy. The relative risk of head and neck cancer after esophagectomy was 34.9. A 7% 5-year cumulative risk of developing head and neck cancer was noteworthy. It was supposed that the increased number of patients with head and neck cancer before esophagectomy was enhanced by the singularity of this institute as a specialized central cancer hospital. However, head and neck cancer was still the most frequently found second primary malignancy among patients who did not have malignancies before esophageal cancer. The stomach and the lung were the next most common sites for second cancer after esophagectomy. We routinely resect the cardiac and lesser-curvature portions of the stomach and form a stomach roll to use as an esophageal substitute.13 Because stomach cancer more commonly occurs in the lesser curvature portion, it is supposed that stomach-roll formation decreases the area of the stomach mucosa at risk of cancer occurrence. However, a significantly high incidence of stomach cancer was still observed in patients who had undergone esophagectomy (relative risk, 2.0). Although only four patients had lung cancer preceding or simultaneous with esophageal cancer in the present series, the risk of subsequent primary lung cancer was significantly high (relative risk, 3.2). Such dissociation was probably an artifact owing to selection bias. Esophageal cancer after lung cancer has often been treated nonsurgically. The change in smoking habits after lung cancer is possibly another factor for dissociation.14 The risk of subsequent primary lung cancer is probably higher than observed, because squamous cell carcinoma detected in the lung within 5 years after esophagectomy was not considered as primary lesion in the present study. Six patients developed nonsquamous cell carcinoma of the lung within 5 years after esophagectomy, whereas six of 11 lung cancers that developed after 5 years were squamous cell carcinomas. Therefore, it was supposed that several patients who developed lung cancer within 5 years had been excluded. Because the object population of the present study had markedly biased characteristics compared with the general population, it is beyond this study to rigorously assess the causes of the increased carcinogenic risk. However, the present study still provides some information on the dynamics of sequential carcinogenesis, because most of the synchronous malignancies in other organs were filtered off by minute systemic examinations before esophagectomy.
Notably, the artifact caused by the difference in the degree of medical surveillance deserves special mention. Most of postesophagectomy patients had received more frequent and cautious health screening after operation than the general population, resulting in earlier detection of malignant lesions or detection of cancers that would not have been found during their lives. On the other hand, it was supposed that intensive preoperative examinations had a contrary effect on the cancer-detection rate after operation. Although it is impossible to properly evaluate the bias caused by difference in surveillance by retrospective analysis, it seemed unlikely that the significantly high second cancer risk after esophagectomy demonstrated in the present study was totally caused by this factor. The approximately linear cumulative risk curve indicated that the hazard of second malignancy varied little for a term of 10 years after esophagectomy (Fig 1 Possible other factors of the increased risk are classified in extrinsic and intrinsic factors that were or were not correlated with treatment for esophageal cancer. The factors that were correlated with esophageal cancer treatment include the effects of radiation and chemotherapy and impaired host susceptibility caused by tumor invasion and treatment. In the present series, the risk of second cancer was not significantly correlated with perioperative therapy nor with the degree of tumor extension. The approximately linear cumulative risk curve also indicated a limited influence of these factors. Among various extrinsic factors, the predominant carcinogenic effect of smoking and drinking habits has been commonly recognized.10,15–17 The results of the present analysis were consistent with such reports. The distribution of organs that developed second cancers also supported the effect of these agents. However, these two factors and sex were strongly correlated with each other in the object population, so it was impossible to determine which of them were the true principal factors. Because information on smoking and drinking habits in the present series of patients were obtained through retrospective assessment, further prospective assessment is required to precisely identify principal factors. Although other extrinsic factors such as dietary and socioeconomic status possibly influence the risk of second cancer, variation in these factors is limited among the Japanese population. The observation that patients with a history of other malignancies before or synchronous with esophageal cancer had significantly increased risk, independent of major extrinsic factors, deserves special attention. It indicates that patients with subsequent malignancies share some intrinsic factors in common. Further genetic investigation on these patients is required.16,18,19
Although the prognosis of patients with a second malignancy after esophageal cancer has been considered unfavorable,4 the present series presented opposite outcomes. The overall 5-year survival rate after detection of the second cancer was 45%. The principal causes of such improved outcome are improvement in treatment for esophageal cancer, intensive screening study for other-organ malignancies immediately before esophageal operation, and periodic follow-up surveillance after esophagectomy. It was supposed that preoperative screening examinations filtered off most other advanced malignancies, as mentioned earlier. Stomach and lung cancers were frequently detected in earlier stages by follow-up endoscopy and computed tomography, respectively. Of 17 patients who developed cancer of the stomach roll, 14 could be successfully treated (seven with endoscopic mucosal resection and seven with partial or distal gastrectomy). Gastrectomy was performed without difficulty because the stomach had been brought up through a subcutaneous route in all cases. Of 17 patients with second lung cancer, 10 were cured by resection. However, we occasionally failed to detect head and neck cancer in the earlier stages. This is possibly one of the causes of early occurrence of head and neck cancer after esophagectomy. Detection of early hypopharyngeal cancer was particularly difficult. Ten of 12 hypopharyngeal cancer cases were detected in the advanced stage, although tumors were in the earlier stage in 15 of 20 cases with cancer of the larynx or the buccal cavity. Because occurrence of head and neck cancer is the most common within 5 years after esophagectomy (Fig 5
The authors indicated no potential conflicts of interest.
1. Altorki N, Kent M, Ferrara C, et al: Three-field lymph node dissection for squamous cell and adenocarcinoma of the esophagus. Ann Surg 236:177–183, 2002[CrossRef][Medline] 2. Peracchia A, Bonavina L, Ruol A, et al: Esophageal cancer: A European perspective. Recent Results Cancer Res 155:119–122, 2000[Medline] 3. Shibuya H, Wakita T, Nakagawa T, et al: The relation between an esophageal cancer and associated cancers in adjacent organs. Cancer 76:101–105, 1995[CrossRef][Medline]
4. Poon RTP, Law SYK, Chu KM, et al: Multiple primary cancers in esophageal squamous cell carcinoma: Incidence and implications. Ann Thorac Surg 65:1529–1534, 1998 5. Hoar SK, Wilson J, Blot WJ, et al: Second cancer following cancer of the digestive system in Connecticut, 1935–82. Natl Cancer Inst Monogr 68:49–82, 1985[Medline] 6. Fogel TD, Harrison LB, Son YH: Subsequent upper aerodigestive malignancies following treatment of esophageal cancer. Cancer 55:1882–1885, 1985[CrossRef][Medline] 7. Orringer MB, Sloan H: Esophagectomy without thoracotomy. J Thorac Cardiovasc Surg 76:643–654, 1978[Abstract]
8. The Research Group for Population-based Cancer Registration in Japan: Cancer incidence and incidence rates in Japan in 1997: Estimates based on data from 12 population-based cancer registries . Jpn J Clin Oncol 32: 318–322, 2002 9. Taeger D, Sun Y, Keil U, et al: A stand-alone Windows application for computing exact person-years, standardized mortality ratios and confidence intervals in epidemiological studies. Epidemiology 11:607–608, 2000[CrossRef][Medline]
10. Begg CB, Zhang ZF, Sun M, et al: Methodology for evaluating the incidence of second primary cancers with application to smoking-related cancers from the surveillance, epidemiology, and end results (SEER) program. Am J Epidemiol 142:653–665, 1995 11. Matsubara T, Ueda M, Nagao N, et al: Cervicothoracic approach for total mesoesophageal dissection in cancer of the thoracic esophagus. J Am Coll Surg 187:238–245, 1998[CrossRef][Medline]
12. Petit T, Georges C, Jung GM, et al: Systematic esophageal endoscopy screening in patients previously treated for head and neck squamous-cell carcinoma. Ann Oncol 12:643–646, 2001 13. Matsubara T, Ueda M, Uchida C, et al: Modified stomach roll for safer reconstruction after subtotal esophagectomy. J Surg Oncol 75:214–216, 2000[CrossRef][Medline]
14. Day GL, Blot WJ, Shore RE, et al: Second cancers following oral and pharyngeal cancers: Role of tobacco and alcohol. J Natl Cancer Inst 86:131–137, 1994
15. Siemiatycki J, Krewski D, Franco E, et al: Association between cigarette smoking and each of 21 types of cancer: A multi-site case-control study. Int J Epidemiol 24:504–514, 1995
16. Yokoyama A, Muramatsu T, Ohmori T, et al: Alcohol-related cancers and aldehyde dehydrogenase-2 in Japanese alcoholics. Carcinogenesis 19:1383–1387, 1998 17. Chyou PH, Nomura AM, Stemmermann GN: Diet, alcohol, smoking and cancer of the upper aerodigestive tract: A prospective study among Hawaii Japanese men. Int J Cancer 60:616–621, 1995[Medline]
18. Grandis JR, Falkner DM, Melhem MF, et al: Human leukocyte antigen class I allelic and haplotype loss in squamous cell carcinoma of the head and neck: Clinical and immunogenetic consequences. Clin Cancer Res 6:2794–2802, 2000
19. Muto M, Nakane M, Hitomi Y, et al: Association between aldehyde dehydrogenase gene polymorphisms and the phenomenon of field cancerization in patients with head and neck cancer. Carcinogenesis 23:1759–1766, 2002 Submitted December 11, 2002; accepted September 10, 2003.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
