the risk of esophageal cancer increased after surgical myotomy. ... achalasia cohort, 20 of 22 esophageal cancers developed in men (SIRs for adenocarcinoma ...
Risk of Esophageal Adenocarcinoma in Achalasia Patients, a Retrospective Cohort Study in Sweden Kazem Zendehdel, MD1,2, Olof Nyrén, MD, PhD1, Annika Edberg, BSc3 and Weimin Ye, MD, PhD1 OBJECTIVES:
Achalasia is a motor disorder of the lower esophageal sphincter, which fails to relax on swallowing. Although a greater risk of esophageal squamous cell carcinoma among achalasia patients is fairly well established, no epidemiological study has evaluated the risk of esophageal adenocarcinoma in these patients.
METHODS:
We compiled a cohort of 2,896 patients recorded with a discharge diagnosis of achalasia between 1965 and 2003 in the Swedish Inpatient Register. The cohort was followed through 2003 via record linkages with essentially complete registers of cancer, causes of death, and migration. Standardized incidence ratios (SIRs) were used to estimate the relative risk of esophageal cancer in achalasia patients compared to the age-, sex-, and calendar period-matched Swedish population. We further estimated SIRs for esophageal cancer among patients treated with esophagomyotomy.
RESULTS:
After excluding the first year of follow-up, we observed excess risks for both squamous cell carcinoma (SIR 11.0, 95% confidence interval [CI] 6.0 –18.4) and adenocarcinoma (SIR 10.4, 95% CI 3.8 – 22.6) of the esophagus. Notwithstanding similar numbers of men and women in our achalasia cohort, 20 of 22 esophageal cancers developed in men (SIRs for adenocarcinoma and squamous cell carcinoma were 8.4 and 13.1, respectively). Increased SIRs among operated patients pertained mainly to esophageal squamous cell carcinoma. We found no evidence that surgical esophagomyotomy increases the risk of esophageal adenocarcinoma.
CONCLUSIONS: Male achalasia patients have substantially greater risks for both squamous cell carcinoma
and adenocarcinoma of the esophagus. Small numbers preclude a firm conclusion about the risk among women. Am J Gastroenterol 2011; 106:57–61; doi:10.1111/j.1572-0241.2007.01258.x
INTRODUCTION Achalasia is characterized by aperistalsis and failure of the lower esophageal sphincter (LES) to relax on swallowing, leading to stagnation of food debris and fluid in a dilated esophagus. While the link with esophageal squamous cell carcinoma (ESCC) seems well established (1–5), the basis for risk evaluation vis-à-vis Barrett’s esophagus and esophageal adenocarcinoma (EAC) rests on case reports (6). A causal relationship has been doubted (3); since the core functional aberration in achalasia is a spasm of LES, gastroesophageal reflux—the main risk factor for Barrett’s esophagus and adenocarcinoma (7)—is expected to be rare. However, as surgical esophagomyotomy is the treatment that provides the best long-term results (5), the possibility of iatrogenic reflux after the myotomy has been a matter of concern (6).
We followed a register-based cohort (2) for a maximum of 38 yr to explore whether achalasia patients also have a higher risk for esophageal adenocarcinoma. We further investigated whether the risk of esophageal cancer increased after surgical myotomy.
METHODS The Swedish Inpatient Register was established in 1964/1965 to document individual hospital discharges. After approval by the Regional Ethics Committee of Karolinska Institutet, all patients in the Inpatient Register with at least one discharge diagnosis of achalasia (ICD7 539,01, ICD8 530,01, ICD9 530A, and ICD10 K22.0) were initially selected to compile the achalasia cohort.
1 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden; 2Cancer Institute Research Center, Medical Sciences/University of Tehran, Tehran, Iran; 3Centre for Epidemiology, The National Board of Health and Welfare, Stockholm, Sweden. Correspondence: Kazem Zendehdel, MD, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Box 281, SE 171 77, Stockholm, Sweden. Received 30 August 2006; accepted 27 February 2007
© 2011 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
57
ESOPHAGUS
ORIGINAL CONTRIBUTIONS
nature publishing group
ESOPHAGUS
58
Zendehdel et al.
The national registration numbers (NRNs), unique personal identifiers assigned to all residents in Sweden, permitted unambiguous record linkages with essentially complete national registers of the total population, cancer (8), death, and migration, as described previously (9). After exclusion of erroneous data (N = 179, 5.3%), and all persons with emigration or a cancer diagnosis preceding index hospitalization (N = 266, 8%), the final achalasia cohort included 2,896 patients (1,573 men). Individual person-time was calculated from the first recorded (index) hospitalization for achalasia until the first cancer diagnosis, death, emigration, esophageal resection, or December 31, 2003, whichever came first. Person-time accrued and cancer events observed during the first year of follow-up were not counted in the analyses, since prevalent cancers that were initially misdiagnosed as achalasia will dominate among the cases diagnosed in the first months after the index hospitalization. Cancers found incidentally at autopsy were also disregarded to avoid possible ascertainment bias related to differential autopsy rates among achalasia patients and the general population. The number of observed cancer cases, divided by the expected number, produced standardized incidence ratios (SIRs) as measures of relative risks. The expected numbers were calculated by multiplying age-, sex-, and calendar yearspecific person-time accrued in the cohort by the corresponding incidence rates derived from the corresponding strata in the general Swedish population (10). We calculated 95% confidence intervals (CIs) of SIRs assuming that observed cancer occurrences were Poisson distributed. In total, 688 patients (402 men) underwent surgical esophagomyotomy during follow-up. Unoperated patients contributed person-time to the nonsurgery stratum and were switched to the surgery stratum on the day of the myotomy. If a cancer developed in the first year after surgery, the operation was ignored and the patient remained in the nonoperated cohort to avoid bias due to reverse causation. To compare risks among subcategories within the cohort, we estimated the relative risk using time-dependent Cox proportional hazards modeling with attained age as the time scale and sex as well as calendar year at entry as covariates. Validation Study
To shed light on the validity of the achalasia diagnoses in the Swedish Inpatient Register, staff at the National Board of Health and Welfare (that keeps the register) randomly selected three achalasia patients from each department of general surgery and otorhinolaryngology at all county and regional referral hospitals in Sweden. According to the register, these departments managed the overwhelming majority of the recorded achalasia patients. In order to facilitate the field-work, the sample was restricted to patients seen in the past 10 yr. A questionnaire was sent out to the chairman or the specialist in charge of these patients at the respective units. Questions were asked about diagnostic and therapeutic practices, criteria for diagnosis, when hospitalization was considered, and the percentage of achalasia patients who were managed exclusively as outpatients. Moreover, these local experts were also asked to review the case records of the selected patients in order to answer questions about the time sequence of The American Journal of GASTROENTEROLOGY
onset of symptoms—first diagnosis—treatments given, diagnostic tests used, and results of dilatations and esophagomyotomies. They were also asked to grade the certainty of the diagnosis, from unambiguously correct to doubtlessly incorrect. The responses were sent to the National Board of Health and Welfare, where the responses were registered, de-identified, and forwarded to the investigators.
RESULTS Men constituted 54.3% of the cohort and the mean age at entry was 54.4 and 59.6 yr among men and women, respectively. Average length of follow-up was 9.9 yr, and 25,766 person-years at risk were accrued after excluding the first year of follow-up (Table 1). We observed 238 incident cancers (any site), corresponding to a SIR close to unity (SIR 1.1, 95% CI 0.9–1.2). During year 2–38, 22 incident cases of esophageal cancer represented a more than 10-fold greater risk (SIR 10.5, 95% CI 7.0–15.9) (Table 2). The excess risks were equally evident for both adenocarcinoma (SIR 10.4, 95% CI 3.8–22.6) and squamous cell carcinoma (SIR 11.0, 95% CI 6.0–18.4). Although women contributed 46% of the observed person-time in our achalasia cohort, only 2 of 22 (9%) esophageal cancers occurred in women, both adenocarcinomas (Table 3).
Table 1. Characteristics of the Cohort of Swedish Achalasia Patients Hospitalized at Least Once Between 1965 and 2003 Men
Women
Overall
All achalasia cohort No. of patients (%)
1,573 (54.3) 1,323 (45.7)
2,896 (100)
Age at entry into the cohort < 30 yr
254 (61.2)
161 (38.8)
415 (100)
30 – 59 yr
594 (58.0)
430 (42.0)
1,024 (100)
≥60 yr
725 (49.8)
732 (50.2)
1,457 (100)
54.4 (21.6)
59.6 (21.6)
56.8 (21.8)
Mean follow-up duration (yr, SD)
9.8 (8.5)
10.0 (8.8)
9.9 (8.6)
Total person-years under surveillance
15,378
13,267
28,645
Ditto, after exclusion of first year
13,816
11,950
25,766
No. of patients (%)
402 (58.5)
286 (41.5)
688 (100)
Mean age at entry into cohort (yr, SD)
41.3 (17.0)
43.8 (17.7)
42.3 (17.3)
Mean follow-up duration (yr, SD)
13.8 (9.4)
15.7 (9.7)
14.6 (9.6)
Mean interval until surgery (yr, SD)
1.4 (3.53)
1.5 (3.52)
1.4 (3.5)
4,605
3,788
8,393
Mean age at entry into cohort (yr, SD)
Operated patients
Person-years after exclusion of first year
VOLUME 106 | JANUARY 2011 www.amjgastro.com
Achalasia and Risk of Esophageal Adenocarcinoma
59
Cancer Type
2–9 yr
10–38 yr
2–38 yr *
SIR (95% CI)
No. of Cases
Incidence Rate (1/100,000 Person-Years)
10
12.7 (6.1– 23.3)
22
85.4
10.5 (7.0–15.9)
4
16.3 (4.4 – 41.8)
6
23.3
10.4 (3.8 – 22.6)
5
10.8 (3.5 – 25.1)
14
54.3
11.0 (6.0 –18.4)
No. of Cases
SIR (95% CI)
No. of Cases
Esophageal cancer *
12
9.2 (4.8 – 16.1)
EAC
2
6.1 (0.7 – 21.8)
ESCC
9
11.1 (5.1 – 21.1)
SIR (95% CI)
*There were, totally, 2 cases for which histological subtypes were not specified, i.e., undifferentiated cancers. Therefore, the sum of adenocarcinoma and squamous cell carcinoma cases will not add up to the total number of esophageal cancers. EAC = esophageal adenocarcinoma; ESCC = esophageal squamous cell carcinoma.
Table 3. Standardized Incidence Ratios (SIRs) and 95% Confidence Intervals (CIs) for Esophageal Cancer in Patients Hospitalized for Achalasia, Stratified by Gender and Surgery (Sweden, 1965 – 2003)
Male EAC ESCC Female EAC Achalasia without esophagomyotomy
Person-Years *
No. of Cases *
Incidence Rate (1/100,000 Person-Years)
SIR (95% CI) *
13,816
20
144.8
13.1 (8.1 – 20.4)
—
4
29.0
8.4 (2.3 – 21.6)
—
14
101.3
16.1 (8.8 – 26.9)
11,950
2
16.7
3.5 (0.4 – 12.6)
—
2
16.7
19.8 (2.4 – 71.6)
17,373
15
86.3
9.1 (5.1 – 15.0)
Follow-up year 2 – 9
11,669
9
77.1
8.0 (3.7 – 15.2)
Follow-up year 10 +
5,704
6
105.2
11.6 (4.2 – 25.2)
Achalasia with esophagomyotomy
8,393
7
83.4
16.0 (6.4 – 33.1)
Follow-up year 2 – 9
4,505
4
88.8
20.4 (5.6 – 52.2)
Follow-up year 10 +
3,888
3
77.2
12.5 (2.6 – 36.5)
*The first year of observation was excluded. Because of 2 cancer cases with unspecified histological subtypes among men, the sum of adenocarcinoma and squamous cell carcinoma cases will not add up to the total number of esophageal cancer. EAC = esophageal adenocarcinoma; ESCC = esophageal squamous cell carcinoma.
We observed excess risks of esophageal cancer among both nonoperated (SIR 9.1, 95% CI 5.1–15.0) and operated (SIR 16.0, 95% CI 6.4–33.1) achalasia patients (Table 3). However, this difference was not statistically significant in overall (RR 1.2, 95% CI 0.5–3.2) and also in a model restricted to men only (RR 1.4, 95% CI 0.5–3.6). Further stratification by follow-up time, albeit hampered by small numbers, gave no indication of increasing risk with increasing time after the esophagomyotomy (Table 3). Only 1 of 7 esophageal cancers observed among operated patients was an adenocarcinoma (SIR 8.0, 95% CI 0.2–44.4). In the validation study answers were received from 31 clinics (60%). Out of 93 selected patients at these units, case records could be traced for 83 (89%) (Table 4). Based on the records of 50 patients, the time between onset of achalasia symptoms and diagnosis was 2.6 (SD 4.2) yr. Of 16 (19%) patients with conditions that were incorrectly classified as achalasia, 4 (5%) had postreflux stricture, 1 (1%) had status postsurgery for obesity, while the other had diagnoses that were unrelated to esophageal cancer risk, e.g., © 2011 by the American College of Gastroenterology
cricopharyngeal achalasia. The overwhelming majority of patients (81%) were diagnosed with more than two diagnostic procedures, typically barium swallow and endoscopy (59%). The estimated percentage of achalasia in patients who were managed as outpatients, and therefore not registered in the Inpatient Register, varied between 0% and 100% among reporting departments (median 0). The majority of responding departments (80%) believed that only few patients are managed exclusively as outpatients.
DISCUSSION The link between achalasia and esophageal squamous cell carcinoma is fairly well established (2,5,11), but this is the first epidemiological study that demonstrates an increased risk of esophageal adenocarcinoma. Barrett’s esophagus, a presumably obligatory precursor lesion in the development of esophageal adenocarcinoma, typically develops following longstanding gastroesophageal reflux disease (GERD). The hypertensive LES in The American Journal of GASTROENTEROLOGY
ESOPHAGUS
Table 2. Standardized Incidence Ratios (SIRs) and 95% Confidence Intervals (CIs) for Esophageal Cancer in Patients Hospitalized for Achalasia, Stratified by Period of Follow-up (Sweden, 1965–2003)
60
Zendehdel et al.
ESOPHAGUS
Table 4. Diagnostic and Treatment Information for Achalasia Patients in Validation Study (N = 83) Frequency Achalasia diagnosis unambiguously correct
67 (100%)
Diagnostic approach Single procedure
13 (19%)
Combination of two or more procedures
54 (81%)
M, BS, and E
24 (36%)
BS and E
15 (23%)
M and E
10 (15%)
M and BS Excellent and long-lasting response to treatment
5 (7%) †
Myotomy Forceful dilation
18/27 (67%) 22/53 (41.5%)
*Since information about diagnostic approach was not meaningful in patients who were incorrectly recorded as achalasia patients (N =16, 19%), they were not included in the analyses. M = manometry; BS = barium swallow; E = endoscopy. † Based only on patients who were received such treatment and for whom an assessment was given. There were 16 (30%) patients who treated with both forceful dilation and myotomy. Information for 3 (7.5%) patients was missing.
achalasia may lessen the probability of reflux, but the possibility of iatrogenic reflux after esophagomyotomy has been a matter of concern among surgeons, and routine prophylactic fundoplication has been advocated (12,13). As no more than 1 out of 6 adenocarcinomas (17%) developed in an operated patient while operated patients accounted for 33% of the total persontime observed in the cohort, our findings did not provide support for implicating surgically induced reflux as a quantitatively important mechanism. Concomitant GERD, existing already before the diagnosis of achalasia, may be one alternative mechanism. Several reports have suggested that esophageal motility disorders may progress from one type to another (14–16). Furthermore, notwithstanding the hypertensive LES in achalasia, complete and prolonged relaxation of the LES was demonstrated in the majority of 11 such patients when monitored for 24 h (17). Given the importance of acid clearance (18–20), it may be hypothesized that even minor reflux in a dysmotility-stricken esophagus may increase the risk of Barrett’s esophagus and esophageal adenocarcinoma. Finally, it has been hypothesized that chronic inflammation caused by fermentation of retained food to lactic acid within the esophagus (2,21) may lead to metaplastic transformation of the squamous mucosa to Barrett’s esophagus. The possibility of adenocarcinogenetic mechanisms that do not involve metaplastic transformation must also be entertained (22), conceivably proceeding from the submucosal glands or from islands of heterotopic columnar epithelium. Surgically treated patients had a higher SIR for esophageal cancer than had the unoperated ones, but it is reasonable to assume that operated patients had more severe and/or long-standing disease. Moreover the difference between operated and unoperated patients did not attain statistical significance. Stratification The American Journal of GASTROENTEROLOGY
by time since surgery demonstrated a tendency towards falling SIRs with time, contrary to the tendency for increasing SIRs in the unoperated group. Thus, our overall data give us little reason for suspecting that the surgical manipulation per se may contribute to cancer development in achalasia patients. If anything, it appears that esophagomyotomy might somewhat reduce the risk of squamous cell carcinoma in the long term. Our validation study showed that specificity of the achalasia diagnosis in the Swedish Inpatient Register was satisfactory. Among the 19% misclassified patient with a false-positive achalasia diagnosis in the validation sample, few (6%) of the entire examined sample had diseases that in themselves are established risk factors for esophageal cancer. Because we excluded all patients with a cancer diagnosis before entry into the cohort and disregarded the first year of follow-up, most cancers that were initially misclassified as achalasia should have been cleared from the cohort when our incidence calculations commenced. Accordingly, exposure misclassification is highly unlikely explanation for the increased esophageal cancer risks observed in our study. Although our study has some unique strength, we also faced certain limitations. Most importantly, we did not have information about other risk factors for esophageal cancer, such as smoking, alcohol consumption, a diet low in fruit and vegetables, or presence of obesity. However, a relative risk equal to unity for lung cancer among the cohort members (N = 15, SIR 1.0, 95% CI 0.6–1.6) as well as low relative risks for alcohol-related cancers, i.e., liver cancer (N = 0) and oral cancer (N = 4, SIR 0.7, 95% CI 0.2–2.1), allay concerns about positive confounding by smoking or alcohol consumption. Further, obesity is not a prominent feature of achalasia patients (5). Moreover, the strength of the observed associations in this study makes confounding by any of the factors discussed in the foregoing unlikely as the sole explanation. In conclusion, while the risk assessment among women with achalasia was inconclusive due to small numbers, our study provided evidence that men with achalasia are at greater risk not only of esophageal squamous cell carcinoma but also of esophageal adenocarcinoma through mechanisms that are yet to be determined. Investigation into the nature of these mechanisms may offer some additional general clues to the etiology of this rapidly increasing cancer. ACKNOWLEDGMENTS
Kazem Zendehdel was supported by a scholarship from the Ministry of Health and Medical Education of Iran. CONFLICT OF INTEREST
Guarantor of the article: Olof Nyrén, MD, PhD. Specific author contributions: Kazem Zendehdel: Study design, data analysis, interpretation of results, preparation of manuscript, and conducting the validation study. Olof Nyrén: Study design, data acquisition, interpretation of results, editing of manuscript, and conducting the validation study. Annika Edberg: Sampling of patients and data acquisition for the validation study, and editing the manuscript. Weimin Ye: Study design, data analysis, interpretation of results, editing of manuscript. VOLUME 106 | JANUARY 2011 www.amjgastro.com
Financial support: The study was made possible by a grant from Karolinska Institutet, Sweden. Potential competing interests: None.
Study Highlights WHAT IS CURRENT KNOWLEDGE
3A high risk of esophageal squamous cell carcinoma in achalasia has been reported. 3Little information is available about esophageal adenocarcinoma risk in achalasia. 3It has been hypothesized that iatrogenic reflux following esophagomyotomy may increase risk of esophageal adenocarcinoma.
WHAT IS NEW HERE
3In addition to esophageal squamous cell carcinoma, the risk of esophageal adenocarcinoma is also substantially elevated.
3We did not find any indication that myotomy is responsible for the observed risk increase.
REFERENCES 1. Meijssen MA, Tilanus HW, van Blankenstein M et al. Achalasia complicated by oesophageal squamous cell carcinoma: a prospective study in 195 patients. Gut 1992;33:155–8. 2. Sandler RS, Nyren O, Ekbom A et al. The risk of esophageal cancer in patients with achalasia. A population-based study. JAMA 1995;274:1359–62. 3. Streitz JM Jr, Ellis FH Jr, Gibb SP et al. Achalasia and squamous cell carcinoma of the esophagus: analysis of 241 patients. Ann Thorac Surg 1995;59:1604–9. 4. Brucher BL, Stein HJ, Bartels H et al. Achalasia and esophageal cancer: incidence, prevalence, and prognosis. World J Surg 2001;25:745–9. 5. St Peter SD, Swain JM. Achalasia: a comprehensive review. Surg Laparosc Endosc Percutan Tech 2003;13:227–40. 6. Guo JP, Gilman PB, Thomas RM et al. Barrett’s esophagus and achalasia. J Clin Gastroenterol 2002;34:439–43.
© 2011 by the American College of Gastroenterology
7. Nyren O, Adami HO. Esophageal Cancer. In: Adami HO, Hunter D, Trichopoulos D (eds). Textbook of Cancer Epidemiology. Oxford University Press: New York, 2002, 137–61. 8. Mattsson B, Wallgren A. Completeness of the Swedish Cancer Register. Non-notified cancer cases recorded on death certificates in 1978. Acta Radiol Oncol 1984;23:305–13. 9. McLaughlin JK, Lipworth L, Fryzek JP et al. Long-term cancer risk among Swedish women with cosmetic breast implants: An update of a nationwide study. J Natl Cancer Inst 2006;98:557–60. 10. Nordenvall C, Nyren O, Ye W. Elevated anal squamous cell carcinoma risk associated with benign inflammatory anal lesions. Gut 2006; 55:703–7. 11. Dent TL, Kukora JS, Buinewicz BR. Endoscopic screening and surveillance for gastrointestinal malignancy. Surg Clin North Am 1989;69:1205–25. 12. Donahue PE, Schlesinger PK, Sluss KF et al. Esophagocardiomyotomy-floppy Nissen fundoplication effectively treats achalasia without causing esophageal obstruction. Surgery 1994;116:719–24; discussion 724-5. 13. Gaissert HA, Lin N, Wain JC et al. Transthoracic Heller myotomy for esophageal achalasia: analysis of long-term results. Ann Thorac Surg 2006;81:2044–9. 14. Usai Satta P, Oppia F, Piras R et al. Extrinsic autonomic neuropathy in a case of transition from diffuse esophageal spasm to achalasia. Clin Auton Res 2004;14:270–2. 15. Griniatsos J, Vlavianos P, Karvounis E et al. Diffuse oesophageal spasm masking achalasia. Int Surg 2004;89:32–4. 16. Robson K, Rosenberg S, Lembo T. GERD progressing to diffuse esophageal spasm and then to achalasia. Dig Dis Sci 2000;45:110–3. 17. van Herwaarden MA, Samsom M, Smout AJ. Prolonged manometric recordings of oesophagus and lower oesophageal sphincter in achalasia patients. Gut 2001;49:813–21. 18. Holloway RH. Esophageal body motor response to reflux events: secondary peristalsis. Am J Med 2000;108 (Suppl 4a): 20S–6S. 19. Achem AC, Achem SR, Stark ME et al. Failure of esophageal peristalsis in older patients: association with esophageal acid exposure. Am J Gastroenterol 2003;98:35–9. 20. Iwakiri K, Hayashi Y, Kotoyori M et al. Transient lower esophageal sphincter relaxations (TLESRs) are the major mechanism of gastroesophageal reflux but are not the cause of reflux disease. Dig Dis Sci 2005;50:1072–7. 21. Smart HL, Foster PN, Evans DF et al. Twenty-four-hour oesophageal acidity in achalasia before and after pneumatic dilatation. Gut 1987;28:883–7. 22. Lagergren J, Bergstrom R, Lindgren A et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999;340:825–31.
The American Journal of GASTROENTEROLOGY
61
ESOPHAGUS
Achalasia and Risk of Esophageal Adenocarcinoma
