Scholarlink

Purpose <p>Colorectal cancer (CRC) is a leading cause of cancer-related death, and timely and accurate recognition of endoluminal pathology is crucial. While CT colonography (CTC) is validated for luminal assessment with a sensitivity concordant to endoscopy, most radiology referrals to the tumor board are based on unprepared CT abdominal studies without bowel preparation or fecal tagging. The diagnostic yield of these routine unprepared CT scans and the influence of radiologist subspecialty, remain uncertain. This study evaluated the positive predictive value of CTC and unprepared CT for suspected endoluminal pathology and examined the impact of gastrointestinal (GI) subspeciality reporting.</p> Methods <p>We reviewed colorectal tumor board outcomes from 2022 at St George’s Hospital, London, United Kingdom. Patients referred to the tumor board by radiology were identified and analyzed through patient records. Radiological and endoscopic concordance was assessed using composite endpoints.</p> Results <p>Of the 106 radiology-initiated referrals to the tumor board in 2022, 61 (58%) were for suspected endoluminal pathology. Overall positive predictive value (PPV) was 79% (42 true positives and 11 false positives). The PPV was 91% for CTC and 70% for unprepared CT. GI subspecialist reporters identified 44% more endoluminal lesions on unprepared CT than non-specialist reports (<i>p</i> &lt; 0.0001), but without a significant difference in PPV (67% vs. 78%, <i>p</i> = 0.543). No significant difference in colorectal cancer detection was observed between CTC and unprepared CT (<i>p</i> = 0.8).</p> Conclusion <p>Unprepared CT demonstrates a good PPV (70%) for detecting endoluminal pathology, with over half of identified lesions being malignant. Although PPV is comparable between GI and non-GI radiologists, GI subspecialists refer significantly more cases for further evaluation, emphasizing the importance of subspeciality expertise. Radiologists should confidently raise suspicion of endoluminal pathology to ensure timely referral for direct visualization and tumor board discussion.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The role of routine imaging in identifying endoluminal colorectal pathology, a United Kingdom clinical experience

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月22日

查看全文
  • Narendranath Govindarajah,
  • Daniel Livingstone,
  • Robert Mitchell,
  • Keith Farrell-Dillon,
  • Edward Antram,
  • Sharmin Malekout,
  • Nigel Beharry,
  • Kunal Patel,
  • Nirav Patel,
  • Kesavan Kandiah,
  • Anita Wale

摘要

Purpose

Colorectal cancer (CRC) is a leading cause of cancer-related death, and timely and accurate recognition of endoluminal pathology is crucial. While CT colonography (CTC) is validated for luminal assessment with a sensitivity concordant to endoscopy, most radiology referrals to the tumor board are based on unprepared CT abdominal studies without bowel preparation or fecal tagging. The diagnostic yield of these routine unprepared CT scans and the influence of radiologist subspecialty, remain uncertain. This study evaluated the positive predictive value of CTC and unprepared CT for suspected endoluminal pathology and examined the impact of gastrointestinal (GI) subspeciality reporting.

Methods

We reviewed colorectal tumor board outcomes from 2022 at St George’s Hospital, London, United Kingdom. Patients referred to the tumor board by radiology were identified and analyzed through patient records. Radiological and endoscopic concordance was assessed using composite endpoints.

Results

Of the 106 radiology-initiated referrals to the tumor board in 2022, 61 (58%) were for suspected endoluminal pathology. Overall positive predictive value (PPV) was 79% (42 true positives and 11 false positives). The PPV was 91% for CTC and 70% for unprepared CT. GI subspecialist reporters identified 44% more endoluminal lesions on unprepared CT than non-specialist reports (p < 0.0001), but without a significant difference in PPV (67% vs. 78%, p = 0.543). No significant difference in colorectal cancer detection was observed between CTC and unprepared CT (p = 0.8).

Conclusion

Unprepared CT demonstrates a good PPV (70%) for detecting endoluminal pathology, with over half of identified lesions being malignant. Although PPV is comparable between GI and non-GI radiologists, GI subspecialists refer significantly more cases for further evaluation, emphasizing the importance of subspeciality expertise. Radiologists should confidently raise suspicion of endoluminal pathology to ensure timely referral for direct visualization and tumor board discussion.