Achalasia-Associated Squamous Cell Carcinoma of the Esophagus: Flow-Cytometric and Histological Evaluation

GASTROENTEROLOGY 1995;108:545-549 CASE REPORTS Achalasia-Associated Squamous Cell Carcinoma of the Esophagus: Flow-Cytometric and Histological Evaluation RAINER PORSCHEN, GABRIELE MOLSBERGER, ANDREAS KUHN, MARIO SARBIA, and FRANZ BORCHARD Department of Gastroenterology and Institute of Pathology, Heinrich-Heine-University, D(Jsseldorf, Germany Although the risk of cancer is increased in patients with achalasia, biomarkers of an increased cancer risk have not been evaluated. In an esophagectomy specimen of a patient with achalasia-associated squarnous cell carcinoma, normal mucosal and carcinomatous samples were systematically taken for flow cytometry and histology. The distribution of DNA aneuploidy and dysplasia was mapped within the resected specimen. Four of 10 tumor samples and 4 of 16 normal mucosal samples of the esophagus showed additional aneuploid stem lines. Gastric mucosa only showed diploid DNA histograms. S-phase fraction in normal esophageal samples (7.8% _+ 1.1%) was lower than in dysplastic and carcinomatous samples (8.8% ± 2.4%; P NS). Areas of mild to moderate dysplasia were detected in the esophageal mucosa adjacent to the neoplasm. This report shows the potential applicability of flow cytometry in the surveillance of patients with achalasia. However, prospective endoscopic studies with long followup periods are required before flow cytometric and histological parameters can be used as biomarkers of an increased cancer risk in achalasia. T he most feared complication of achalasia is the development of carcinoma of the esophagus. In the majority of cases, the tumors are squamous cell carcinomas. They are thought to develop as a consequence of chronic inflammation of the squamous epithelium caused by retention and stasis of food and secretions. In some patients, adenocarcinomas are detected as a late sequel of Barrett's esophagus developing after pneumatic dilata- 1-4 tion or surgery. Because patients with long-standing achalasia and dysphagia are likely to attribute any symptomatic change to their original disease, tumors are often detected in advanced stages, resulting in a poor prognosis. 1-4 As a consequence of this fact, it has been suggested that patients with achalasia should be entered into endoscopic surveillance programs) '6 Esophageal squamous cell carcinoma evolves through a series of progressively dysplastic changes/'8 However, in contrast to other precancerous lesions of the gastrointestinal tract (e.g., Barrett's esophagus, ulcerative colitis), 9'1 biomarkers of a malignant transformation have not been evaluated in achalasia. Therefore, we report on a patient with an achalasiaassociated squamous cell carcinoma of the esophagus. In the esophagectomy specimen, systematic histological and flow cytometric analyses were performed to study the appearance of dysplasia and DNA aneuploidy in the normal squamous epithelium as indicators of a neoplastic transformation. Case Report A 56-year-old man with a 21-year history of achalasia that had been treated several times by pneumatic dilatation presented with retrosternal pain and progressive dysphagia at another hospital. Esophagogastroscopy, which was hampered by retained food, showed an ulcer near the cardia. The endoscope could be passed through the cardiac orifice. Biopsy specimens taken from the esophagus did not show carcinomatous changes. Computed tomography of the chest and the upper abdomen showed a dilated esophagus with a thickened wall in the distal third of the esophagus. Because of persistent dysphagia and pain despite bouginage, the patient was admitted to the university hospital. A control endoscopy showed a solid, vulnerable lesion in the distal third of the esophagus. Multiple biopsy specimens showed a squamous epithelium with loss of the normal architecture due to a well to moderately differentiated squamous cell carcinoma. Seven days after admittance, hematemesis developed. Upper endoscopy showed multiple bleeding erosions throughout the esophagus. The patient was transferred to the Department of Surgery. He underwent transthoracic esophagectomy with abdominal and mediastinal lymphadenectomy and a cervical esophagogastrostomy. The postoperative course was uneventful. Immediately after resection, the esophagectomy specimen 1995 by the American Gastroenterological Association 0016-5085/95/S3.00

546 PORSCHEN ET AL. GASTROENTEROLOGY Vo1.108, No. 2 was opened longitudinally. Samples of fresh mucosa were taken in a gridlike pattern from the resected esophagus and the resected upper part of the stomach. At each level, three samples were obtained circumferentially. The interval between levels was 2 centimeters. These samples were divided for histological and flow cytometric analysis. For histology, samples were fixed in formalin and embedded in paraffin. Light microscopic analysis was independently performed by two of the authors (M.S. and F.B.) without knowledge of the flow cytometric analysis. Dysplasia was classified as mild, moderate, or severe, v'11 The other half of each sample to be processed for flow cytometry was stored at 20 C in dimethyl sulfoxide-citrate buffer. Before analysis, the specimens were thawed in a water bath. The tissue was dissociated by gently grinding the sample in a steel mesh with a glass pestle. The resulting cell suspension was flushed with dimethyl sulfoxide-citrate buffer into a test tube. Cell nuclei were isolated by a detergent-trypsin method 12 and stained with propidium iodide. Chicken erythrocytes added to each test tube were used as an internal DNA standard. 10,000-15,000 cell nuclei per sample were measured in a FACScan flow cytometer (Becton Dickinson, Heidelberg, Germany). The mean coefficient of variation of the G1 peaks was 2.3% + 0.4% (range, 1.9%-3.3%). Samples were either classified as diploid or aneuploid. Samples with more than one G0/G1 peak in the DNA histogram and a corresponding second G2/M peak were judged as aneuploid. 13 These samples were only considered aneuploid if a DNA aneuploidy was confirmed by a second run of the same sample. For DNA aneuploid samples, a DNA index was calculated as the ratio of the abnormal G0/G1 mean peak channel number to the diploid G0/G1 mean peak channel number. DNA indices were considered identical if the deviation between the indices was lower than 3%. A shoulder was defined as a sample with cell nuclei with an excess DNA content to the right of the G0/G1 peak, not showing bimodality. Histogram classification was performed by one of the authors (R.P.) without previous knowledge of the pathological data. Cell cycle parameters were analyzed with the commercially available soft- ware program RFIT (Becton Dickinson). Although a significant cell aggregation could be excluded, cell cycle analysis was performed after doublet discrimination by generating a dot plot of fluorescence area vs. fluorescence width. Values are expressed as arithmetic means + 1 SD. Results Histology Pathological study showed a well to moderately differentiated squamous cell carcinoma with invasion of the adventitia (Figure 1). The tumor was located in the distal and middle thirds of the esophagus. It extended beyond the esophagogastric junction and undermined the gastric epithelium. No metastatic carcinoma was found in the resected mediastinal and perigastric lymph nodes (pt3 pn0 cm0). The normal squamous epithelium partially contained erosions and ulcerations. Dysplastic changes (Figure 2) were only noted in the esophageal epithelium but not in the adjacent gastric mucosa. Eight of 24 samples taken from the esophagus in a gridlike pattern showed carcinoma (with 2 additional tumor samples in the stomach), 3 samples showed moderate dysplasia, and 1 sample showed mild dysplasia. Flow Cytometry Of the 33 samples analyzed by flow cytometry, 8 samples had additional aneuploid stem lines (Figure 3). Four of these samples were taken from the carcinoma and 4 from the esophageal mucosa. The fraction of aneuploid cells amounted to 10%-35% of the total cells analyzed in the carcinomatous samples and to 5%-8% in the noncarcinomatous samples. The DNA index was 1.06-1.10 in the tumor samples and 1.18-1.24 in the mucosal samples. Gastric mucosa only showed diploid DNA histograms. Figure 1. Well-differentiated squamous cell carcinoma of the esophagus with keratinization (H&E; original magnification 526x). Figure 2. Moderate dysplasia adjacent to the carcinoma with irregular papillae, anisonucleosis, and mature epithelium at the surface (H&E; original magnification 526x).

February 1995 DNA PLOIDY AND DYSPLASIA IN ACHALASIA 547 Minor DNA abnormalities in form of a shoulder (Figure 3) were present in 4 other tumor samples and in 1 sample of normal esophageal epithelium. S-phase fractions in the dysplastic esophageal mucosa and in the carcinoma (8.9% + 3.3% and 8.8% 13%) were comparable. Mean S-phase fraction decreased to 7.8% + 1.1% in the nondysplastic esophageal mucosa (P NS) and to 2.2% + 2.0% in the gastric mucosa (P < 0.0001). Mean G2/M-phase fractions in the nondysplastic esophageal mucosa (4.0% _ 1.2%), dysplastic esophageal mucosa (3.6% + 0.9%), and the carcinoma (3.0% + 0.8%) did not differ significantly. Correlation Between Histology and Flow Cytometry The distribution of regions with an abnormal DNA content in comparison with the histological results z J 4 5c~7 A 1 i 0 z 32"~ B 4 rd Z d t DNA CONTENT ANALYS I S P " ' " ~ :v"s" 61:;,... '8" i, 'g,... { 6 F, ~ F L U O R E S C E N C E CHANNEL NO. DNA CONTENT ANALYSIS are shown in Figure 4. An area of 3 samples with identical DNA indices was found near the carcinoma; it contained 1 sample with moderate and 1 sample with mild dysplasia. However, DNA aneuploidy did not always correlate with dysplasia on a sample-by-sample basis. Discussion Although it is generally accepted that achalasia is a predisposing factor for the development of cancer, the true incidence still remains a matter of debate. Recent endoscopic series suggest that the risk of developing squamous cell carcinoma is increased 8-33-fold in comparison with the normal population. %6 As in our patient, carcinoma of the esophagus often arises at an earlier age in persons with achalasia than in normal patients. Symptoms of achalasia generally precede the development of cancer by a mean of 17-20 years. 4 6 Because of the increased cancer risk in patients with achalasia, some investigators advise endoscopic surveillance) '6 To improve the efficacy of such endoscopic screening programs, it would be helpful to dispose of a biomarker to detect patients with an increased cancer risk at the precancerous stage. In ulcerative colitis and in Barrett's esophagus, dysplasia is used as a histological marker for an increased cancer risk. 9'1 Flow cytometric abnormalities such as DNA aneuploidy and an increased G2/M fraction have also been assessed as additional markers in these diseases. 9x4 16 These studies have shown that esophagus 2 9 9 mild dysplasia moderate dysplasia carcinoma 0 Z stomach F L U O R E S C E N C E C H A N N E L N O. Figure 3, (A) DNA histogram of a biopsy specimen taken from the tumor showing a near-diploid aneuploid peak with a DNA index of 1.1. The corresponding aneuploid G2/M peak can be seen on the right. The first peak on the left represents chicken erythrocytes. (B) DNA abnormalities in the form of a shoulder without a clearly defined second G1 peak are present in another tumor biopsy specimen. Figure 4. Schematic diagram of the esophagectomy specimen showing the distribution of flow cytometric and histological abnormalities. Circles denote sites where samples were taken for flow cytometry and histological analysis. Closed circles denote sites where minor DNA abnormalities in the form of a shoulder were measured by flow cytometry. Numbers indicate DNA indices of aneuploid samples. The heavy line encloses an area of samples with the same DNA index.

548 PORSCHEN ET AL. GASTROENTEROLOGY Vol. 108, No. 2 genomic instability and clonal evolution are associated with the progression to dysplasia and carcinoma. Unfortunately, in achalasia, details regarding the histological findings in the esophageal mucosa adjacent to the tumor are seldom found in the literature. In one study, Iv leukoplakia was described; in another study, 18 the squamous cell carcinoma originated in an area of severe dysplasia. This shortcoming prompted us to perform a systematic search for histological and flow cytometric alterations in an esophagectomy specimen with an achalasia-associated carcinoma. Areas with mild to moderate dysplasia were found in the esophageal mucosa adjacent to the tumor. Results of histology and flow cytometry were not always concordant. However, the correlation between flow cytometry and histology was similar to that found in specimens of Barrett's esophagus and ulcerative colitis resected for dysplasia or cancer. 14'19 Because tissue samples were divided for flow cytometry and histology, it cannot be excluded that the correlation between these two parameters might be slightly influenced by this kind of sample preparation. In the carcinoma, 4 of 10 samples contained a neardiploid aneuploidy and 4 additional samples showed a shoulder. Because only minor DNA abnormalities were observed in the tumor (DNA index, 1.06-1.10), it cannot be excluded that in the samples with a shoulder subtle DNA, aberrations existed that could not be resolved by flow cytometry despite the low coefficient of variation. Two tumor samples were diploid. Sasaki et al. 2 have also shown intratumoral heterogeneity in DNA ploidy of esophageal squamous cell carcinomas. In the esophageal mucosa, additional aneuploid sampies were detected. It could be objected that these peaks with small DNA abnormalities (DNA index, 1.18-1.22) containing a low fraction of the total number of ceils in the histogram might represent false aneuploid peaks produced by autolysis of the samples. 21 However, several observations argue against this view. Near-diploid peaks were not observed in all samples of the esophageal mucosa but only in 4 of 16. The amount of cell debris was low in all samples and not increased in the samples with a near-diploid aneuploidy. Near-diploid peaks were not found in the gastric samples, although in the stomach, autolysis might proceed faster due to the presence of hydrochloric acid. Near-diploid peaks also did not exist in samples taken in patients with reflux esophagitis (personal observation). In Barrett's esophagus 9'19 and in ulcerative colitis, 14 small near-diploid peaks were also shown by flow cytometry. In some of these patients, cytogenetic evaluation confirmed a corresponding karyotypic abnormality. 9 These near-diploid peaks cannot be produced by an aggregation between a chicken erythrocyte nucleus and a human G1 nucleus because this event would result in a DNA index of 1.57. At the present the biological significance of an increased cell proliferation in the nondysplastic esophageal mucosa is unknown. It cannot be explained by the presence of erosions and ulcerations in the esophageal mucosa because the mean S-phase fraction (4.9% + 1.9%) as well as the mean G2/M-phase fraction (2.0% + 1.0%) in patients with reflux esophagitis is significantly lower (personal observations). In patients with Barrett's esophagus, an increased G2/tetraploid fraction represents a risk factor for the progression to high-grade dysplasia or carcinoma during follow-up. 9 In conclusion, a systematic study in an esophagectomy specimen of a patient with achalasia-associated squamous cell carcinoma showed histological and flow cytometric abnormalities. Although this first report shows the potential applicability of histological and flow cytometric analyses, prospective endoscopic studies with long follow-up periods are required before these parameters can be used as biomarkers of an increased cancer risk in achalasia. References 1. Chuong JJH, DuBovik S, McCallum RW. Achalasia as a risk factor for esophageal carcinoma. A reappraisal. Dig Dis Sci 1984;29: 1105-1108. 2. Goodman P, Scott LD, Verani RR, Berggreen CC. Esophageal adenocarcinoma in a patient with surgically treated achalasia. Dig Dis Sci 1990;35:1549-1552. 3. Hankins JR, McLaughiin JS. The association of carcinoma of the esophagus with achalasia. J Thorac Cardiovasc Surg 1975;69: 355-360. 4. Just-Viera JO, Haight C. Achalasia and carcinoma of the esophagus. Surg Gynecol Obstet 1969;128:1081-1095. 5. Brossard E, Ollyo JB, Fontotliet C, Savary M, Monnier P. Achalasia and squamous cell carcinoma of the esophagus: is an endoscopic surveillance justified (abstr)? Gastroenterology 1992; 102:A4. 6. Meijssen MAC, Tilanus HW, van Biankenstein M, Hop WCJ, Ong GL. Achalasia complicated by oesophageal squamous cell carcinoma: a prospective study in 195 patients. Gut 1992;33:155-158. 7. Borchard F. Pr~kanzerosen der Speiser6hre: Krebsrisikoerkrankungen und Dysplasien. Verdauungskrankheiten 1990;8:69-84. 8. Correa P. Precursors of gastric and esophageal cancer. Cancer 1982; 50:2554-2565. 9. Reid B J, Blount PL, Rubin CE, Levine DS, Haggitt RC, Rabinovitch PS. Flow-cytometric and histological progression to malignancy in Barrett's esophagus: prospective endoscopic surveillance of a cohort. Gastroenterology 1992;102:1212-1219. 10. Riddell RH, Goldman H, Ransohoff DF, Appelman HD, Fenoglio CM, Haggitt RC, Aren C, Correa P, Hamilton SR, Morson BC, Sommers SC, Yardley JH. Dysplasia in inflammatory bowel disease: standardized classification with provisional clinical applications. Hum Pathol 1983;14:931-968. 11. Watanabe H, Jass JR, Sobin LH. Histological typing of oesophageal and gastric tumours. 2nd ed. Berlin: Springer Verlag, 1990. 12. Vindelov LL, Christensen I J, Nissen NI. A detergent-trypsin

February 1995 DNA PLOIDY AND DYSPLASIA IN ACHALASIA 549 method for the preparation of nuclei for flow cytometric DNA analysis. Cytometry 1983;3:323-327. 13. Hiddemann W, Schumann J, Andreeff M, Barlogie B, Herman C J, Leif RC, Mayall BH, Murphy RF, Sandberg AA. Convention on nomenclature for DNA cytometry. Cytometry 1984; 5:445-446. 14. Levine DS, Rabinovitch PS, Haggitt RC, Blount PL, Dean P J, Rubin CE, Reid BJ. Distribution of aneuploid cell populations in ulcerative colitis with dysplasia or cancer. Gastroenterology 1991; 101:1198-1210. 15. LSfberg R, BrostrSm O, Karl6n P, Ost A, Tribukait B. DNA aneuploidy in ulcerative colitis: reproducibility, topographic distribution, and relation to dysplasia. Gastroenterolog~j 1992;102: 1149-1154. 16. Porschen R, Robin U, Schauseil S, Borchard F, hengels K J, Strohmeyer G. DNA aneuploidy in Crohn's disease and ulcerative colitis: results of a comparative flow cytometric study. Gut 1992; 33:663-667. 17. Camara-Lopes LH. Carcinoma of the esophagus as a complication of megaesophagus: an analysis of seven cases. Am J Dig Dis 1961;6:742-756. 18. Eckardt VF, Junginger T, Gabbert HE, Bettendorf U. Superficial esophageal carcinoma in achalasia detected by endoscopic surveillance. Z Gastroenterol 1992;30:411-414. 19. Reid B J, Haggitt RC, Rubin CE, Levine DS, Rabinovitch PS. Barrett's esophagus. Correlation between flow cytometry and histology in detection of patients at risk for adenocarcinoma. Gastroenterology 1987;93:1-11. 20. Sasaki K, Murakami T, Murakami T, Nakamura M. Intratumoral heterogeneity in DNA ploidy of esophageal squamous cell carcinomas. Cancer 1991;68:2403-2406. 21. Alanen KA, Joensuu H, Klemi PJ. Autolysis is a potential source of false aneuploid peaks in flow cytometric DNA histograms. Cytometry 1989; 10:417-425. Received February 7, 1994. Accepted October 18, 1994. Address requests for reprints to: Rainer Porschen, M.D., First Department of Internal Medicine, Eberhard-Karls-University, Otfried- MiJller-Str. 10, D-72076 TiJbingen, Germany. Fax: (49) 70-7129- 2095. Supported by a grant of the Deutsche Krebshilfe (M 24/90/Po2) and the Ernst and Berta Grimmke-Stiftung.