Radiographic diagnosis of early colorectal cancer, with special reference to the superficial type of invasive carcinoma

Incidental 18F-FDG uptake in the colon: value of contrast-enhanced CT correlation with colonoscopic findings

OBJECTIVES

To evaluate the impact of morphological information derived from contrast-enhanced CT in the characterization of incidental focal colonic uptake in ¹⁸F-FDG PET/CT examinations.

METHODS

A total of 125 patients (female: n = 53, male: n = 72) that underwent colonoscopy secondary to contrast-enhanced, full-dose PET/CT without special bowel preparation were included in this retrospective study. PET/CT examinations were assessed for focal colonic tracer uptake in comparison with the background. Focal tracer uptake was correlated with morphological changes of the colonic wall in the contrast-enhanced CT images. Colonoscopy reports were evaluated for benign, inflammatory, polypoid, precancerous, and cancerous lesions verified by histopathology, serving as a reference standard. Sensitivity, specificity, PPV, NPV, and accuracy for detection of therapeutic relevant findings were calculated for (a) sole focal tracer uptake and (b) focal tracer uptake with correlating CT findings in contrast-enhanced CT.

RESULTS

In 38.4% (48/125) of the patients, a focal ¹⁸F-FDG uptake was observed within 67 lesions. Malignant lesions were endoscopically and histopathologically diagnosed in eleven patients, and nine of these were detected by focal ¹⁸F-FDG uptake. A total of 34 lesions with impact on short- or long-term patient management (either being pre- or malignant) were detected. Sensitivity, Specificity, PPV, NPV, and accuracy for sole ¹⁸F-FDG uptake for this combined group were 54%, 69%, 29%, 85%, and 65%. Corresponding results for focal ¹⁸F-FDG uptake with correlating CT findings were 38%, 90%, 50%, 86%, and 80%. This resulted in a statistically significant difference for diagnostic accuracy (p = 0.0001) CONCLUSION: By analyzing additional morphological changes in contrast-enhanced CT imaging, the specificity of focal colonic ¹⁸F-FDG uptake for precancerous and cancerous lesions can be increased but leads to a considerate loss of sensitivity. Therefore, every focal colonic uptake should be followed up by colonoscopy.

Endoscopic management of large colorectal polyps

OBJECTIVES

The optimal treatment for large colorectal polyps (LCPs) is still a controversial issue. The aim of this study was to evaluate the safety and effectiveness of endoscopic polypectomy (EP) of colorectal polyps >or=2 cm in size.

PATIENTS AND METHODS

One hundred fifty-one EP LCPs were performed over a period of 7 years. Diathermal snare was used for pedunculated and pseudopedunculated polyps and endoscopic mucosal resection (EMR) or biopsy forceps polypectomy for sessile and flat polyps. The resected polyps were recovered and collected for histology. At scheduled follow-up visits 1, 3, 6, and 12 months after polypectomy, complications and recurrences were recorded in all patients.

RESULTS

Fifteen polyps were located in the rectum, 84 in the sigmoid colon, 11 in the descending colon, four in the splenic flexure, 11 in the transverse colon, 11 in the hepatic flexure, seven in the ascending colon and eight in the cecum. Fifty-six polyps were sessile, 54 pedunculated, 25 pseudopedunculated, and 16 flat. At histology, most of polyps (131) were adenomas (nine with adenocarcinoma in situ). Five were invasive polypoid carcinomas and required colonic resection. Immediate bleeding occurred in ten patients (7.6%) and it was stopped by endoscopic hemoclips (7), epinephrine injection (1), or surgery (2). There were three perforations (2.3%; all polypoid carcinomas), managed endoscopically (1) or surgically (2). Delayed bleeding occurred in two patients (1.5%) and was treated by endoscopic diathermy and hemoclips (1) or surgery (1). During follow-up, six (4.6%) incompletely excised polyps and three (2.3%) relapses in the site of previous EP were detected and endoscopically removed.

CONCLUSION

EP is relatively safe and effective for benign-appearing LCPs.