Histological features responsible for brownish ... - Wiley Online Library

bs_bs_banner

doi:10.1111/jgh.12059

GASTROENTEROLOGY

Histological features responsible for brownish epithelium in squamous neoplasia of the esophagus by narrow band imaging Hiromitsu Kanzaki,* Ryu Ishihara,* Shingo Ishiguro,‡ Kengo Nagai,* Fumi Matsui,* Takeshi Yamashina,* Takashi Ohta,* Sachiko Yamamoto,* Noboru Hanaoka,* Masao Hanafusa,* Yoji Takeuchi,* Koji Higashino,* Noriya Uedo,* Hiroyasu Iishi* and Yasuhiko Tomita† Departments of *Gastrointestinal Oncology and †Pathology, Osaka Medical Center for Cancer and Cardiovascular Diseases, and ‡PCL Osaka Inc., Osaka, Japan

Key words brownish area, brownish epithelium, esophageal cancer, esophageal squamous neoplasia, narrow band imaging. Accepted for publication 16 October 2012. Correspondence Dr Ryu Ishihara, Department of Gastrointestinal Oncology, Osaka Medical Center for Cancer and Cardiovascular Diseases, 3-3, Nakamichi 1-chome, Higashinari-ku, Osaka 537-8511, Japan. Email: [email protected] Competing interests: None.

Abstract Background and Aim: Esophageal squamous neoplasias usually appear brown under narrow band imaging as a result of microvascular proliferation, and brownish color changes in the areas between vessels, referred to as brownish epithelium. However, the reasons for the development of this brownish epithelium and its clinical implications have not been fully investigated. Methods: Patients with superficial esophageal neoplasias treated by endoscopic resection were included in the study. Areas of mucosa with brownish and non-brownish epithelia were evaluated histologically. Results: A total of 68 superficial esophageal neoplasias in 58 patients were included in the analysis. Of the 68 lesions, 32 were classified in the brownish epithelium group, and 36 in the non-brownish epithelium group. Brownish epithelium was significantly associated with a diagnosis of high-grade intraepithelial neoplasia or invasive cancer (P < 0.0001). Thinning of the keratinous layer, thinning of the epithelium, and cellular atypia were significantly associated with brownish epithelium by univariate analysis, and thinning of the keratinous layer and thinning of the epithelium were confirmed to be independent factors by multivariate analysis. The odds ratios were 9.6 (95% confidence interval: 2.0–46.3) for thinning of the keratinous layer, and 4.6 (95% confidence interval: 1.1–19.4) for thinning of the epithelium. Conclusions: Brownish epithelium is an important finding in the diagnosis of esophageal squamous neoplasia, and may be related to thinning of the keratinous layer, caused by neoplastic cell proliferation, and thinning of the epithelium.

Introduction Esophageal cancer is the sixth most common cause of cancerrelated mortality worldwide.1 The overall survival of patients with esophageal cancer, regardless of its histological type, remains poor. However, a favorable prognosis can be expected if this cancer is detected at an early stage.2–5 Endoscopic examination has been reported to be useful for the early detection of esophageal cancer,6–14 but the usefulness of conventional endoscopy is currently limited because of its poor ability to detect early neoplastic changes.6,7 Narrow band imaging (NBI) is a non-invasive optical technique that uses reflected light to visualize the organ surface.8 NBI can enhance the superficial structure and facilitate the detection of neoplastic lesions,9–11 and is regarded as the standard modality for screening of esophageal squamous neoplasias.9–14 274

Lesions that are suspicious for esophageal squamous neoplasias usually appear brown when using NBI. Brownish areas are regarded as a red-flag sign for esophageal neoplasia, and their importance in the screening of esophageal neoplasias has been confirmed in previous studies.9,10,12,13,15,16 Brownish areas comprise areas of microvascular proliferation, with brownish color changes in the areas between vessels, referred to as brownish epithelium.17 The vascular architecture changes during cancer development, and previous studies have performed detailed analyses of the vascular changes seen in esophageal cancer.9,17,18 However, the cause of the brownish epithelium and its clinical implications have not been fully investigated. A better knowledge of the reasons for the development of brownish epithelium may improve our understanding of the lesion’s character, and increase the chances of its accurate diagnosis. This study, therefore, aimed to clarify the histological

Journal of Gastroenterology and Hepatology 28 (2013) 274–278 © 2012 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd

H Kanzaki et al.

Histological change of brownish epithelium

(a)

(b)

Figure 1 (a) Lesion with brownish color change in the area between vessels. (b) Lesion without brownish color change in the area between vessels.

changes responsible for the development of the brownish epithelium in esophageal squamous neoplasia.

Methods Endoscopic examination and resection. The current clinical investigation was carried out during routine endoscopic resections for esophageal squamous lesions, and patients with superficial esophageal neoplasias treated with endoscopic resection were included in the study. Prior to endoscopic resection, the lesions were evaluated by endoscopic examination with conventional imaging and NBI, followed by chromoendoscopy with iodine solution in all patients. Biopsy specimens were taken from iodine-unstained lesions. Lesions that are histologically diagnosed as mucosal high-grade neoplasias were treated with endoscopic resection. Patients were excluded if they had previously undergone surgery, chemotherapy, or radiotherapy for esophageal cancer. The endoscopy procedures were carried out using a high-resolution magnifying upper-gastrointestinal endoscope (GIF-Q240Z or GIF-H260Z, Olympus, Tokyo, Japan). The structure enhancement function of the video processor was set at a level of B8 (strongest enhancement level for microstructure) for NBI observation. A black soft hood (MB-162 for GIF-Q240Z or MB-46 for GIFH260Z, Olympus) was mounted on the tip of the endoscope to maintain an adequate distance between the tip of the endoscope zoom lens and the mucosal surface during observations. The endoscopic procedure was carried out with the patients under intravenous sedation with midazolam (Dormicam, Yamanouchi Pharmaceutical, Tokyo, Japan) and pentazocine (Pentazin, Sankyo Pharmaceuticals, Tokyo, Japan). Using NBI, lesions were divided into lesions with brownish epithelium (Fig. 1a) and lesions with non-brownish epithelium (Fig. 1b). A representative area (2–5 mm) of the lesions specified as the region of interest was marked with marker dots. The lesions were then resected by endoscopic submucosal dissection or endoscopic mucosal resection. Further analyses were conducted in the region of interest. Written informed consent was obtained from all patients before endoscopic examination and resection. Institutional review board approval was granted for a retrospective chart review and analysis of procedural data. Histological evaluation and further treatment. All specimens were cut into 2-mm slices and embedded in paraffin.

Figure 2 Thinning of the epithelium. The epithelium in the diseased area was ⱕ 75% of the surrounding normal epithelium. Black single parenthesis shows thinning of the epithelium.

Sections were cut from the paraffin blocks, and stained with HE and Prussian blue. Histological assessments of brownish epithelium and non-brownish epithelium were conducted in the areas marked with marker dots. The depth of cancer involvement was classified according to the Japanese Classification of Esophageal Carcinoma,19 and intraepithelial neoplasias were classified into low-grade intraepithelial neoplasia (LGIN) or high-grade intraepithelial neoplasia (HGIN), according to the World Health Organization classification.20 Intraepithelial cancer was included in HGIN. The following histological features of the mucosa possibly associated with brownish epithelium were evaluated. Thinning of the epithelium in the area was assessed, and classified as thinning (ⱕ75% of the surrounding normal epithelium) (Fig. 2) or nonthinning. Thinning of the keratinous layer in the area was assessed and classified as thinning (ⱕ3 layers) or non-thinning (Fig. 3). Cellular atypia was assessed and classified as high grade, moderate grade, or low grade, based on nuclear irregularity, mitotic figures, loss of polarity, chromatin pattern, and nuclear/cytoplasmic ratio. Destruction of basal cell layer, presence of surface differentiation in squamous neoplastic cells, and presence of extravascular red blood cells (RBCs; characterized by RBCs > 1% of the mucosal area) were examined by HE staining. The presence of hemosiderin


275


H Kanzaki et al.

Table 1

Figure 3 Thinning of the keratinous layers (ⱕ 3 layers). Black single parenthesis shows thinning of the keratinous layers.

pigment (characterized by hemosiderin pigment > 1% of the mucosal area) was examined by Prussian blue staining. Adequate cut-off values were determined by two investigators (HI and RI) based on the endoscopic and histological findings in some representative cases. The thicknesses of the epithelium and the keratinous layer may differ depending on how the specimen was fixed or sliced. The influence of such factors was minimized by assessing thinning of the epithelium as the ratio of the diseased area to the surrounding normal epithelium, and thinning of the keratinous layer as the number of layers, because the keratinous layer was too thin to be compared with the normal epithelium. Histological assessments were conducted by a pathologist (SI), who was blinded to the NBI findings and to the clinical characteristics of the patients. Statistical analysis. The associations between the presence of brownish epithelium and each histological finding were assessed. We aimed to analyze at least 60 cases (30 cases in each group) to ensure adequate power in the statistical analyses. If a patient had more than one lesion at the time of resection, each lesion was considered as a separate entry for the purposes of statistical analysis. Univariate analysis of the association between brownish epithelium and each histological finding was performed using Yates’ c2 test for comparisons of variables. In multivariate analysis, independent factors were determined by Cox regression hazards models. Histological findings with significant or marginal associations with brownish epithelium were included in the multivariate analysis. The model fit was assessed using the Hosmer– Lemeshow test. A two-sided P-value of < 0.05 was considered statistically significant. A Bonferroni-adjusted P-value was used for multiple comparisons to control for experimental errors due to multiple testing. All analyses were carried out using computer software (SPSS version 11.0, SPSS, Chicago, IL, USA).

Results A total of 68 superficial esophageal neoplasias in 58 patients treated by endoscopic resection at the Osaka Medical Center for 276

Characteristics of patients and lesions

Gender (n) Male Female Age (years) Median Range Lesion location Cervical esophagus Upper thoracic esophagus Middle thoracic esophagus Lower thoracic esophagus Lesion size (mm) Median Range Histological diagnosis of the marked area LGIN HGIN LPM

54 4 67 45–83 2 12 35 19 20 2–74 22 43 3

HGIN, high-grade intraepithelial neoplasia; LGIN, low-grade intraepithelial neoplasia; LPM, cancer invading into the lamina propria.

Cancer and Cardiovascular Diseases from May 2011 to March 2012 were included in the analysis. The locations of the lesions were the cervical esophagus in two lesions, upper thoracic esophagus in 12 lesions, middle thoracic esophagus in 35 lesions, and lower thoracic esophagus in 19 lesions. The median size of the lesions was 20 mm (range 2–74 mm) (Table 1). Of the 68 lesions, 32 were classified in the brownish epithelium group, and 36 were classified in the non-brownish epithelium group. Of 33 lesions < 20 mm, 16 showed brownish epithelium and 17 showed nonbrownish epithelium, while of 35 lesions ⱖ 20 mm, 16 showed brownish epithelium and 19 showed non-brownish epithelium (P = 0.82). The histological diagnoses of the areas with brownish epithelium were HGIN or cancer invading into the lamina propria in 31 of 32 lesions, while those of the areas with non-brownish epithelium were HGIN or cancer invading into the lamina propria in 15 of 36 lesions. There was a significant association between brownish epithelium and a diagnosis of HGIN or invasive cancer (P < 0.0001). Fifteen of 46 HGINs or cancers had non-brownish epithelium because some HGIN and cancer cases showed the presence of abnormal cells in the upper half of the epithelium and non-thinning of the keratinous layers (> 3 layers). One of 22 LGINs showed brownish epithelium, probably because of thinning of the epithelium. Among 15 HGIN or invasive cancer cases with non-brownish epithelium, seven showed thinning of the keratinous layer and four showed thinning of the epithelium. Among 31 HGIN or invasive cancer cases with brownish epithelium, 27 showed thinning of the keratinous layer and 21 showed thinning of the epithelium. Among 21 LGIN with non-brownish epithelium, two showed thinning of the keratinous layer and three showed thinning of the epithelium. The only case of LGIN with brownish epithelium showed thinning of the keratinous layer and thinning of the epithelium. Univariate analysis (Table 2) found significant associations between brownish epithelium and thinning of the keratinous layer, thinning of the epithelium, and cellular atypia. There was no significant association between brownish epithelium and presence


H Kanzaki et al.


Table 2 Associations between presence of brownish epithelium and histological findings (univariate analysis)

Thinning of the epithelium Thinning of the keratinous layer Cellular atypia

Presence of surface differentiation Presence of basal membrane Extravascular red blood cells Presence of hemosiderin pigment

ⱕ 75% > 75% ⱕ 3 layers > 3 layers Mild Moderate Severe Yes No Yes No ⱕ 1% > 1% ⱕ 1% > 1%

Brown

Non-brown

22 10 28 4 0 24 8 10 22 16 16 27 5 32 0

7 29 9 27 15 19 2 2 34 28 8 33 3 36 0

P-value 0.0001 < 0.0001 0.0004

0.014 0.032 0.58 1

Table 3 Associations between presence of brownish epithelium and histological findings (multivariate analysis)

Thinning of the epithelium Thinning of the keratinous layer Presence of surface differentiation Cellular atypia

Odds ratio (95% confidence interval)

P-value

> 75% (reference) ⱕ 75% > 3 layers (reference) ⱕ 3 layers

1 4.6 (1.1–19.4) 1 9.6 (2.0–46.3)

0.035

Yes (reference) No

1 1.2 (0.18–7.9.1)

0.85

Moderate (reference) Severe

1 1.2 (0.17–7.7)

0.88

0.005

of surface differentiation after Bonferroni’s correction. In multivariate analysis (Table 3), thinning of the keratinous layer and thinning of the epithelium were confirmed to be independent factors. The odds ratios were 9.6 (95% confidence interval [CI]: 2.0–46.3) for thinning of the keratinous layer and 4.6 (95% CI: 1.1–19.4) for thinning of the epithelium. Hosmer–Lemeshow tests indicated that the model achieved a sufficient goodness of fit (P = 0.996).

Discussion In this study, brownish epithelium was significantly associated with a diagnosis of HGIN or invasive cancer. Thinning of the keratinous layer and thinning of the epithelium were independently associated with the presence of brownish epithelium. NBI is a real-time, on-demand endoscopic imaging technique designed to enhance visualization of the vascular network and surface texture of the mucosa.8 NBI images are provided by feeding the 415-nm image in the blue and green channels, and the 540-nm image in the red channel of the monitor. A brownish color change in the object due to decreased reflection of the blue and

green channels is caused by the absorption of 415-nm light. Transmission of 415-nm light, which is strongly absorbed by hemoglobin and only penetrates the surface of tissues, is dependent on the hemoglobin distribution and surface structure of the tissue. When the emitted short-wavelength light is reflected or backwardscattered by the surface, the light does not contact the hemoglobin, while when the emitted short-wavelength light penetrates the surface or is forward-scattered by the tissue components, the light contacts and is absorbed by hemoglobin during its optical path. Surfaces with high reflectance, thus, appear bright in color, while surfaces with lower reflectance appear dark or brownish in color. A dense distribution of hemoglobin in the areas between vessels associated with extravascular RBCs could be responsible for the brownish color of the epithelium because hemoglobin strongly absorbs 415-nm light. However, in the current study, extravascular RBCs occupied less than 1% of the total mucosal area in most cases, and such small amounts of hemoglobin would be unlikely to be responsible for the brownish color change of the area. Moreover, there was no difference in the volume of extravascular RBCs between lesions with brownish and non-brownish epithelia. These results suggest that extravascular RBCs may not be the cause of the brownish epithelium. Low reflectance of the surface may cause a brownish color change in the epithelium. Squamous epithelium has high reflectance,21 and therefore appears brighter in NBI, compared with columnar epithelium. This property may be derived from the keratinous layer or from the whole epithelium. It seems likely that the reflectance would decrease with thinning of the keratinous layer and epithelium, causing the area to appear brown in color. The close relationship between thinning of the keratinous layer and the epithelium suggests that these two findings may be responsible for the development of the brownish color. The vascular architecture changes with the development of squamous neoplasias.9,17,18 However, the degree of neoplastic cell proliferation cannot be diagnosed by the vascular changes. Thinning of the keratinous layer or epithelium is caused by replacement or occupation of the epithelium by neoplastic cells, and most areas with brownish epithelium therefore demonstrated proliferation of neoplastic cells into the upper half of the epithelium. Brownish epithelium was also closely associated with a diagnosis of HGIN or cancer requiring the presence of abnormal cells in the upper half of epithelium. The combination of vascular changes and brownish epithelium might, thus, help predict the development and proliferation of neoplasias, and facilitate the accurate diagnosis of esophageal lesions. Resection is required for HGIN or cancer, and differentiating between these conditions and LGIN is thus an important step in deciding on a suitable treatment strategy.20 Histological diagnosis based on biopsy specimens taken from unsuitable sites may result in misdiagnosis. However, the close association between brownish epithelium and HGIN or cancer suggests that these areas represent good targets for confirming the diagnosis. Moreover, the thin or absent keratinous layer in brownish epithelial areas means that neoplastic cells might be easily obtained from these areas by forceps. In conclusion, brownish epithelium on NBI is an important finding for the diagnosis of esophageal squamous neoplasias. The brownish coloration may be caused by thinning of the keratinous layer and the epithelium.


277


H Kanzaki et al.

Acknowledgment The authors are grateful to Mr Kazuhiro Gono of the Olympus Medical Systems Corporation for his support in the interpretation of the optical mechanism of brownish epithelium.

References 1 Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J. Clin. 2005; 55: 74–108. 2 Kodama M, Kakegawa T. Treatment of superficial cancer of the esophagus: a summary of responses to a questionnaire on superficial cancer of the esophagus in Japan. Surgery 1998; 123: 432–9. 3 Ishihara R, Tanaka H, Iishi H et al. Long-term outcome of esophageal mucosal squamous cell carcinoma without lymphovascular involvement after endoscopic resection. Cancer 2008; 112: 2166–72. 4 Fujishiro M, Yahagi N, Kakushima N et al. Endoscopic submucosal dissection of esophageal squamous cell neoplasms. Clin. Gastroenterol. Hepatol. 2006; 4: 688–94. 5 Igaki H, Kato H, Tachimori Y et al. Clinicopathologic characteristics and survival of patients with clinical Stage I squamous cell carcinomas of the thoracic esophagus treated with three-field lymph node dissection. Eur. J. Cardiothorac Surg. 2001; 20: 1089–94. 6 Hashimoto CL, Iriya K, Baba ER et al. Lugol’s dye spray chromoendoscopy establishes early diagnosis of esophageal cancer in patients with primary head and neck cancer. Am. J. Gastroenterol. 2005; 100: 275–82. 7 Dawsey SM, Fleischer DE, Wang GQ et al. Mucosal iodine staining improves endoscopic visualization of squamous dysplasia and squamous cell carcinoma of the esophagus in Linxian, China. Cancer 1998; 83: 220–31. 8 Gono K, Obi T, Yamaguchi M et al. Appearance of enhanced tissue features in narrow-band endoscopic imaging. J. Biomed. Opt. 2004; 9: 568–77. 9 Yoshida T, Inoue H, Usui S, Satodate H, Fukami N, Kudo SE. Narrow-band imaging system with magnifying endoscopy for superficial esophageal lesions. Gastrointest. Endosc. 2004; 59: 288–95. 10 Muto M, Nakane M, Katada C et al. Squamous cell carcinoma in situ at oropharyngeal and hypopharyngeal mucosal sites. Cancer 2004; 101: 1375–81.

278

11 Kumagai Y, Inoue H, Nagai K, Kawano T, Iwai T. Magnifying endoscopy, stereoscopic microscopy, and the microvascular architecture of superficial esophageal carcinoma. Endoscopy 2002; 34: 369–75. 12 Ishihara R, Takeuchi Y, Chatani R et al. Prospective evaluation of narrow-band imaging endoscopy for screening of esophageal squamous mucosal high-grade neoplasia in experienced and less experienced endoscopists. Dis. Esophagus 2010; 23: 480–6. 13 Ezoe Y, Muto M, Uedo N et al. Magnifying narrowband imaging is more accurate than conventional white-light imaging in diagnosis of gastric mucosal cancer. Gastroenterology 2011; 141: 2017–25. 14 Ishihara R, Inoue T, Hanaoka N et al. Autofluorescence imaging endoscopy for screening of esophageal squamous mucosal high-grade neoplasia: a phase II study. J. Gastroenterol. Hepatol. 2012; 27: 86–90. 15 Takenaka R, Kawahara Y, Okada H et al. Narrow-band imaging provides reliable screening for esophageal malignancy in patients with head and neck cancers. Am. J. Gastroenterol. 2009; 104: 2942–8. 16 Lee YC, Wang CP, Chen CC et al. Transnasal endoscopy with narrow-band imaging and Lugol staining to screen patients with head and neck cancer whose condition limits oral intubation with standard endoscope (with video). Gastrointest. Endosc. 2009; 69: 408–17. 17 Ishihara R, Inoue T, Uedo N et al. Significance of each narrow-band imaging finding in diagnosing squamous mucosal high-grade neoplasia of the esophagus. J. Gastroenterol. Hepatol. 2010; 25: 1410–15. 18 Kumagai Y, Toi M, Inoue H. Dynamism of tumour vasculature in the early phase of cancer progression: outcomes from oesophageal cancer research. Lancet Oncol. 2002; 3: 604–10. 19 Japan Esophageal Society. Japanese classification of esophageal cancer, tenth edition: part I. Esophagus 2009; 6: 1–25. 20 Gabbert HE, Shimoda T, Hainaut P, Nakamura Y, Field JK, Inoue H. Squamous cell carcinoma of the esophagus. In: Hamilton SR, Aaltonen LA, eds. Pathology and Genetics of the Digestive System: World Health Organization Classification. Lyon: IARC Press, 2000; 11–19. 21 Wilmink GJ, Ibey BL, Tongue T et al. Development of a compact terahertz time-domain spectrometer for the measurement of the optical properties of biological tissues. J. Biomed. Opt. 2011; 16: 047006-1–10.
