|Year : 2018 | Volume
| Issue : 3 | Page : 574-577
The diagnostic accuracy of colonoscopic brush cytology in diagnosis of colorectal malignancies: A study of 49 patients
Simorjot Kaur1, Reetika Sharma1, Vijay Kaushal1, Anchana Gulati1, Brij Sharma2
1 Department of Pathology, Indira Gandhi Medical College, Shimla, Himachal Pradesh, India
2 Department of Gastroenterology, Indira Gandhi Medical College, Shimla, Himachal Pradesh, India
|Date of Web Publication||12-Jun-2018|
Postgraduate Student, Department of Pathology, Indira Gandhi Medical College, Shimla - 171 001, Himachal Pradesh
Source of Support: None, Conflict of Interest: None
Aims: To find the spectrum and frequency of lower gastrointestinal malignancies and diagnostic accuracy of colonoscopic brush cytology in their diagnosis.
Materials and Methods: This was a prospective study carried out on 49 patients in the Department of Pathology over 1 year. Brushing material was smeared directly onto at least two clean glass slides. The air dried smears were stained with May Grunwald Giemsa stain. The endoscopic biopsies were examined grossly and were fixed in 10% formalin, processed and stained with Hematoxylin and Eosin stain, respectively. Special stains were used wherever required.
Observation and Results: The study was done on 49 patients presented with colorectal and anal lesions. Age of the patients ranged from 17 to 72 years with male to female ratio being 1.57:1. On statistical analysis, the sensitivity of colonoscopic brush cytology was found to be 85.71% and specificity 61.53%. The accuracy came out to be 79.16%.
Conclusion: Brush cytology is a reliable, safe, inexpensive, and rapid method of diagnosing gastrointestinal lesions. Since brushing is a relatively noninvasive procedure, routine use of brushings of colonoscopically visible lesions should be done, in addition, to biopsy to increase the diagnostic yield.
Keywords: Brush cytology, diagnostic accuracy, lower gastrointestinal tract
|How to cite this article:|
Kaur S, Sharma R, Kaushal V, Gulati A, Sharma B. The diagnostic accuracy of colonoscopic brush cytology in diagnosis of colorectal malignancies: A study of 49 patients. J Can Res Ther 2018;14:574-7
|How to cite this URL:|
Kaur S, Sharma R, Kaushal V, Gulati A, Sharma B. The diagnostic accuracy of colonoscopic brush cytology in diagnosis of colorectal malignancies: A study of 49 patients. J Can Res Ther [serial online] 2018 [cited 2020 Jul 9];14:574-7. Available from: http://www.cancerjournal.net/text.asp?2018/14/3/574/176416
| > Introduction|| |
Malignancies of gastrointestinal tract (GIT) are one of the leading causes of death worldwide. Brush cytology has proven to be a valuable method in the diagnosis of upper GIT. When combined with biopsy, its accuracy can be as high as 95–100%. Colorectal cancer (CRC) is the fourth most commonly diagnosed cancer and the second leading cause of cancer related death in the United States. Colonoscopy has revolutionized the approach to the diagnosis and management of patients with colorectal neoplasia. A variety of diagnostic techniques is currently available for the assessment of the nature of the lesion, e.g., biopsy, brush cytology, and lavage cytology. It is less invasive and a risk-free technique compared to biopsy. This study was carried out to find out the diagnostic accuracy of colonoscopic brush cytology.
| > Materials and Methods|| |
This study was a prospective study carried out on 49 patients in the Department of Pathology over 1 year. The samples were collected from outdoor/indoor patients attended the Department of Gastroenterology and Medicine in whom colonoscopic findings were suggestive of malignancy. Brushing material was smeared directly onto at least two clean glass slides. The air dried smears were stained with May Grunwald Giemsa stain. The endoscopic biopsies were examined grossly and were fixed in 10% formalin, processed and stained with Hematoxylin and Eosin stain, respectively. Special stains were used wherever required. Validation of cytological diagnosis was done on the basis of histological examination.
| > Observation and Results|| |
The study was done on 49 patients presented with colorectal and anal lesions. Age of the patients ranged from 17 to 72 years with maximum number of patients presented in age group 60–69 years (13 patients). There were 30 males and 19 females with male:female ratio of 1.57:1. On colorectal biopsy, out of 49 cases with clinical suspicion of malignancy, 35 cases had malignancy and 13 were benign, whereas in 1 case biopsy was inadequate for evaluation. Brushings were positive for malignancy in 23 cases (out of 49 cases) and 13 were benign, whereas 13 cases were given as suspicious for malignancy due to scant representative material/air drying artifacts/obscuring inflammation [Table 1] and [Figure 1]. Out of 35 biopsy proven cases of malignancy, 30 (85.75%) cases were of adenocarcinoma, 1 case each was of squamous cell carcinoma (2.85%), gastrointestinal stromal tumor (2.85%), non-Hodgkin lymphoma (2.85%), carcinoid (2.85%), and villous adenoma with lowgrade dysplasia (2.85%) [Table 2] and [Figure 2]. On cytology, 4 cases were given suspicious of malignancy and 1 case was given as malignant (adenocarcinoma) came out as benign on biopsy. The latter case was presented with a history of constipation. On endoscopy, circumferential growth was seen, but on biopsy, there was normal rectal mucosa. The patient lost the follow-up, and the repeat biopsy could not be performed. The other 4 patients whose cytology was given as suspicious of malignancy, three of them were follow-up cases of rectal carcinoma and one was on treatment for ulcerative colitis.
|Table 1: Comparisons of brush cytology with biopsy of colorectal and anal lesions|
Click here to view
|Figure 1: (a) Adenocarcinoma rectum malignant cells forming glands (May Grunwald Giemsa, ×400); (b) non-Hodgkin Lymphoma dispersed population of malignant lymphoid cells (May Grunwald Giemsa, ×100); (c) granulomatous colitis epithelioid cell granuloma (May Grunwald Giemsa, ×400)|
Click here to view
|Figure 2: (a) Granulomatous coloitis epithelioid cell granuloma (H and E, ×400); (b) adenocarcinoma rectum-malignant cells forming glands (H and E, ×100); (c) signet ring carcinoma colon showing signet ring cells (H and E, ×400); (d) clear cell variant of adenocarcinoma colon showing clear cell change in cytoplasm (H and E, ×100)|
Click here to view
The results were interpreted as follows:
- True positive (TP) - Cytology correctly interprets a case as malignant
- True negative (TN) - Cytology correctly interprets a case as benign
- False positive (FP) - Cytology falsely interprets a case as malignant
- False negative (FN) - Cytology falsely interprets a case as not having malignant lesion.
The following were the results:
Sensitivity: Likelihood that patient with disease has positive test results.
Specificity: Likelihood that patient without disease has negative test results.
Thus, the sensitivity of colonoscopic brush cytology came out as 85.71% and specificity 61.53%. The diagnostic accuracy came out to be 79.16%.
| > Discussion|| |
Of the total 49 cases presented with lesions in lower GIT (colorectal and anal canal), the rectum was the predominant site of involvement accounting for 19/49 (43.33%) cases. Sayeed et al. also reported rectum as the most common site of involvement in lower GIT in their study. A biopsy was positive for malignancy in 35 cases, benign in 13 cases, and inadequate in 1 case. Adenocarcinoma was the most common malignancy 30/35 (85.71%). In addition, there was one case each of moderately differentiated squamous cell carcinoma, mesenchymal tumor, carcinoid, polyp, and adenoma. The peak age of malignancy in large intestine was 60–69 years in our study, whereas a study by Celestino et al. reported lower GIT malignancy in patients over the age of 40 years. Among the benign lesions, granulomatous colitis was nonspecific colitis, were the most common lesions detected in 2 cases each. Mortensen et al. compared the accuracy of direct vision brush cytology at colonoscopy with colonoscopic biopsies, colonoscopy alone, and radiology in the diagnosis of colonic strictures. The combined accuracy of cytology and biopsy was 88%. They concluded that direct vision brush cytology has an important place in the accurate tissue diagnosis of colonic strictures, particularly those in the rectosigmoid. Chen  in their study found that cell brushings are often positive when colonic stricture/obstruction prevent the colonoscope from reaching the lesion; biopsy, obviously, cannot be done in such a case. In a study by Bardawil et al. cytology proved to have a sensitivity of 73% and a specificity of 100% whereas tissue biopsy showed a sensitivity of 81% and a specificity of 100%. By combining the two methods, the sensitivity increased to 92%. Similarly, Ehya and O'Hara  also found that combination of the two techniques increased the sensitivity to 90% and improved the overall accuracy of the test. The combined sensitivity for forceps biopsy and cytology was 98 percentage in a study by Farouk et al. However, combining the results of brush cytology and biopsy resulted in a statistically significant increase in sensitivity to 97.4 percentage in the study by Brouwer et al. In our study, the sensitivity of colonoscopic brush cytology was 85.71%, specificity was 61.53%, and diagnostic accuracy of brush cytology was 79.16%. The reason for low specificity in this study was due to the detection of only 8 out of 13 benign lesions on cytology correctly (low true negative cases). Four cases were reported as suspicious and one case as malignant (high false positive) due to reactive changes such as the enlargement of nuclei and prominent nucleoli. False negative cases were might be because of ulcerated growth, the tumor cells were not sampled adequately, or the brushing was done from the site adjacent to the tumor. The false positive cases were might be due to reactive/reparative changes in the lining epithelium, e.g., large hyperchromatic nuclei, prominent nucleoli, inflammatory process, or after erosion. On cytology, it is very difficult to differentiate reparative epithelium from malignant one. Biopsy is the gold standard for this. The limitation of this study was the less number of patients (49) and use of only biopsy for the validation of brush cytology and the absence of other confirmatory tests such as barium roentgenography, surgery, or follow-up. These changes are usually seen in epithelium adjacent to an ulcer or chronic up.
| > Conclusion|| |
Brush cytology is a reliable, safe, inexpensive, and rapid method of diagnosing GI lesions. It is more useful when the lesion is large or multiple or when patients refuse for biopsy. However, brush cytology cannot distinguish between in situ carcinoma and invasive carcinoma. Though the colonoscopic biopsy is used as a routine procedure in the diagnosis of gastrointestinal lesions, brush cytology can be used as an adjunct to tissue biopsy during to distinguish benign from malignant lesions of the GIT. Regardless of the biopsy findings, patients with “suspicious” cytologic reports require careful reevaluation.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Halpern M, Gal R, Rath-Wolfson L, Koren R, Weil R, Avni A. Brush cytology and biopsy in the diagnosis of colorectal cancer. A comparison. Acta Cytol 1997;41:628-32.
Gadoa A, Ebeidb B, Abdelmohsenc A, Axon A. Improving the yield of histological sampling in patients with suspected colorectal cancer during colonoscopy by introducing a colonoscopy quality assurance program. Gastroenterol Res 2011;4:157-61.
Sayeed MA, Islam R, Siraji D, Hoque MG, Mohsen AQ. Colonoscopy: A study of findings in 332 patients. JCMCTA 2007;18:28-31.
Celestino A, Castillo T, Frisancho O, Contardo C, Espejo H, Tomioka C, et al
. Colorectal cancer: Study on 365 cases. Rev Gastroenterol Peru 1996;16:187-96.
Mortensen NJ, Eltringham WK, Mountford RA, Lever JV. Direct vision brush cytology with colonoscopy: An aid to the accurate diagnosis of colonic strictures. Br J Surg 1984;71:930-2.
Chen YL. The diagnosis of colorectal cancer with cytologic brushings under direct vision at fiberoptic colonoscopy. A report of 59 cases. Dis Colon Rectum 1987;30:342-4.
Bardawil RG, D'Ambrosio FG, Hajdu SI. Colonic cytology. A retrospective study with histopathologic correlation. Acta Cytol 1990;34:620-6.
Ehya H, O'Hara BJ. Brush cytology in the diagnosis of colonic neoplasms. Cancer 1990;66:1563-7.
Farouk R, Edwards J, Thorne M, Young W, MacDonald AW, Monson JR, et al.
Brush cytology for the diagnosis of rectal carcinoma. Br J Surg 1996;83:1456-8.
Brouwer R, MacDonald A, Matthews R, Gunn J, Monson JR, Hartley JE. Brush cytology for the diagnosis of colorectal cancer. Dis Colon Rectum 2009;52:598-601.
[Figure 1], [Figure 2]
[Table 1], [Table 2]