Limited Diagnostic Yield of Routine Gastroscopy in FIT-Positive Patients

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for allowing me to review this paper entitled "Limited Diagnostic Yield of Routine Gastroscopy in FIT-Positive Patients" by Khader et al. The authors investigate the prevalence of upper gastrointestinal lesions at gastroscopy in patients with positive FIT that also underwent colonoscopy and compare findings between those with normal and abnormal colonoscopy. The authors found that clinically significant upper GI lesions in patients with positive FIT were rare regardless of colonoscopy findings, with no cases of upper GI malignancy detected. The study is of interest and adds to the body of evidence in the literature confirming the low clinical utility of gastroscopy in patients with positive FIT. However, I believe there are some concerns that should be addressed. Specific comments:

 

As only a small subset of the entire cohort of patients also underwent gastroscopy in addition to colonoscopy, do the authors think these patients different from those who did not undergo gastroscopy? Did these patients also have upper GI symptoms, e.g. dyspepsia, or were they asymptomatic and only had positive FIT?

 

Regarding description of bowel preparation quality in the methods and results sections I suggest specifying what BBPS scores corresponded to the descriptions of “intermediate,” “good,” and “very good” as it is currently unclear.

 

Section 3.1.6 reports on the prevalence of “colitis” – but it is not clear what kind of colitis this is. I suggest being more specific in terms of what this refers to (e.g. ulcerative colitis? microscopic colitis? non-specific colitis?).

 

There are two sections numbered 3.1.7, so I believe the one entitled “Gastroscopy As a Key Diagnostic Tool in FIT-Positive Patients with Normal sColonoscopy Findings” should be section 3.1.8. There is also a typo in the section title, “sColonoscopy” instead of “Colonoscopy”.

 

In the discussion it is stated that “On the supportive side, Ishtiaq, R. et al. reported that up to 60% of FIT-positive patients with normal colonoscopy results had detectable upper gastrointestinal lesions, suggesting a potentially missed diagnostic oppor tunity when gastroscopy is omitted [22].” However, reference 22 is a paper entitled “Adenoma detection rate vs. adenoma per colonoscopy as quality indicators for colon cancer screening,” which as far as I can tell does not discuss upper gastrointestinal lesions. I think there is an error here, can the authors double-check this?

 

Table 5 is very difficult to interpret and unclear. What is being compared? The prevalence of upper GI findings in patients with normal vs abnormal colonoscopy? Please clarify.

Author Response

Reviewer 1 answers

1. As only a small subset of the entire cohort of patients also underwent gastroscopy in addition to colonoscopy, do the authors think these patients are different from those who did not undergo gastroscopy? Did these patients also have upper GI symptoms, e.g. dyspepsia, or were they asymptomatic and only had positive FIT?

Thank you very much for this insightful comment. Indeed, a portion of the patients underwent gastroscopy despite being asymptomatic, mainly as part of the evaluation following a positive FIT result. In other cases, gastroscopy was performed due to non-specific upper gastrointestinal complaints or upon patient or physicians’ requests.

 

2. Regarding description of bowel preparation quality in the methods and results sections I suggest specifying what BBPS scores corresponded to the descriptions of “intermediate,” “good,” and “very good” as it is currently unclear.

We thank the reviewer for this helpful comment. In the revised manuscript, we have clarified the Boston Bowel Preparation Scale (BBPS) scoring system corresponding to the descriptive categories used. Specifically, the definitions now read as follows: “Bowel preparation quality was assessed according to the Boston Bowel Preparation Scale (BBPS). A total BBPS score of 4-5 was categorized as intermediate, 6-7 as good, and 8-9 as very good. Patients with scores ≤3 were considered to have poor bowel preparation and were excluded from the analysis.”

 

3. Section 3.1.6 reports on the prevalence of “colitis’’, but it is not clear what kind of colitis this is. I suggest being more specific in terms of what this refers to (e.g. ulcerative colitis? microscopic colitis? non-specific colitis?).

Thank you for this valuable comment. We agree that clarification was necessary. The cases labeled as “colitis” in our dataset refer specifically to ulcerative colitis as documented in the endoscopic and diagnostic records of the participating centers. We have updated the text in Section 3.1.6 (Prevalence of Gastrointestinal Disorders) to specify this, as follows: “Ulcerative colitis was the most commonly diagnosed inflammatory condition, affecting 356 patients (0.93%).”

 

4. There are two sections numbered 3.1.7, so I believe the one entitled “Gastroscopy As a Key Diagnostic Tool in FIT-Positive Patients with Normal sColonoscopy Findings” should be section 3.1.8. There is also a typo in the section title, “sColonoscopy” instead of “Colonoscopy”.

We thank the reviewer for noticing these important details. We have carefully reviewed the manuscript and corrected both issues. The section “Gastroscopy as a Key Diagnostic Tool in FIT-Positive Patients with Normal Colonoscopy Findings” has now been renumbered as Section 3.1.8, and the typographical error (“sColonoscopy”) has been corrected to “Colonoscopy.”

 

5. In the discussion it is stated that “On the supportive side, Ishtiaq, R. et al. reported that up to 60% of FIT-positive patients with normal colonoscopy results had detectable upper gastrointestinal lesions, suggesting a potentially missed diagnostic oppor tunity when gastroscopy is omitted [22].” However, reference 22 is a paper entitled “Adenoma detection rate vs. adenoma per colonoscopy as quality indicators for colon cancer screening,” which as far as I can tell does not discuss upper gastrointestinal lesions. I think there is an error here, can the authors double-check this?

Thank you for catching this, you're absolutely right. The citation to Ishtiaq et al. was an inadvertent miscitation introduced during reference updating. That paper evaluates colonoscopy quality indicators (ADR vs APC) and does not report on upper-GI findings in FIT-positive cohorts. We have corrected the discussion text and replaced the miscitation with appropriate sources that specifically address the diagnostic yield of concomitant gastroscopy in FIT-positive patients, namely:

• Choe et al., 2023 (systematic review of gastroscopy after positive FIT and colonoscopy).
• Shah et al., 2023 (systematic review and meta-analysis on upper GI endoscopy in FIT/FOBT-positive subjects).
• Planade et al., 2021 (impact of systematic upper endoscopy performed with screening colonoscopy).

The revised sentence now reads: “Several studies have shown that adding upper endoscopy can identify upper-GI lesions in a meaningful proportion of FIT-positive individuals, supporting a selective role for gastroscopy in this setting (Choe et al., 2023; Shah et al., 2023; Planade et al., 2021).”

6. Table 5 is very difficult to interpret and unclear. What is being compared? The prevalence of upper GI findings in patients with normal vs abnormal colonoscopy? Please clarify.

Thank you for this important comment. Indeed, Table 5 represents the p-values corresponding to Figure 9, which compares the prevalence of upper GI findings detected by gastroscopy between FIT-positive patients with normal colonoscopy results and those with abnormal colonoscopy results.

To make this clearer, we have revised the table title, legend, and figure accordingly. The table is now presented as a supplementary file for easier interpretation, and the corresponding p-values have been added directly onto Figure 9 to visually represent the statistical comparisons.

Table Appendix 1: Statistical comparison (p-values) of upper gastrointestinal findings in FIT-positive patients with normal versus abnormal colonoscopy results (corresponding to Figure 9). This table summarizes the p-values derived from Chi-square or Fisher’s exact tests comparing the frequency of each gastroscopic finding (e.g., gastritis, hernia, esophagitis, duodenitis, ulcers, reflux esophagitis, and severe gastritis) between the two groups.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This is a large retrospective trial examining the role of upper endoscopy in the evaluation of patients with post FIT test positive colonoscopy. They examined all patients that underwent colonoscopy for this indication and are able, based on their population of data to determine the utility of examining the upper GI tract when there are or are not colonoscopic findings or abnormalities. I believe this is of great importance as there is a great strain on resources and need to streamline after FIT + testing who will need or benefit from further evaluation. The data here supports the notion that if the colonoscopy is normal there is little need to pursue upper GI tract evaluation in the absence of clinical concern. 

Author Response

Thank you very much for your response; we appreciate it.

Reviewer 3 Report

Comments and Suggestions for Authors

This is a well written paper investigating the diagnostic value of endoscopy in FIT-positive patients. FIT-positivity is a common reason why patients see their gastroenterologists. There are a few concerns that need to be addressed prior to publication. 

1. The discussion did a nice summary about the controversial findings in the similar studies. Suggesting more about the screening guidelines modifications could provide novel insight to the topic.
2. clarification of definition: “Severe gastritis” should be defined histologically or endoscopically—criteria are not clear. "Normal" and "abnormal" colonoscopy should be more detailed and classified as non-invasive and invasive if data available. 
3. Multivairiate analysis should be used to rule out confounders. 

Author Response

Reviewer 3 answers

1. The discussion did a nice summary about the controversial findings in the similar studies. Suggesting more about the screening guidelines modifications could provide novel insight to the topic.

We thank the reviewer for this thoughtful and constructive suggestion. We agree that discussing potential implications for future screening guideline modifications would add valuable perspective to the manuscript. Accordingly, we have expanded the Discussion section to include the following paragraph: “Our findings support a more selective approach to upper gastrointestinal evaluation following a positive FIT. Given the very low rate of clinically significant upper GI pathology observed, routine gastroscopy in all FIT-positive individuals with normal colonoscopy findings does not appear justified. Instead, screening guidelines could consider recommending gastroscopy only for subgroups with specific risk factors, such as age above 60 years, chronic PPI use, anemia, or persistent upper GI symptoms. Incorporating these stratified criteria into national CRC screening protocols could improve cost-effectiveness while minimizing unnecessary procedures.”

 

2. Clarification of definition: “Severe gastritis” should be defined histologically or endoscopically, criteria are not clear. "Normal" and "abnormal" colonoscopy should be more detailed and classified as non-invasive and invasive if data available.

We thank the reviewer for this helpful comment. We have clarified both definitions in the revised manuscript for greater transparency and methodological accuracy.

Regarding “severe gastritis”, this term in our dataset was based on endoscopic appearance, as histologic data were not consistently available across centers. We have now specified that: “Severe gastritis was defined endoscopically as marked erythema, mucosal friability, erosions, or ulcerations extending over a large gastric area.”

We fully agree that subclassifying abnormal colonoscopic findings into non-invasive and invasive categories would provide valuable additional insight. However, this level of pathological detail was not consistently available in our retrospective dataset, as endoscopy reports did not uniformly include histopathologic confirmation or lesion depth characterization. Therefore, classification into non-invasive and invasive categories could not be reliably performed. We have clarified this limitation in the Methods section.

 

3. Multivairiate analysis should be used to rule out confounders

Thank you for your comment. We agree that multivariate regression would provide additional adjustment for potential confounding variables. However, the retrospective nature of our dataset did not include several key clinical covariates such as comorbidities, medication use, or detailed laboratory parameters. Therefore, a multivariate model could not be reliably constructed without introducing bias from incomplete data. We have acknowledged this as a limitation in the Discussion section. ‘’Another limitation of this study is that multivariate analysis could not be performed to adjust for potential confounders such as comorbidities, medication use, or other clinical variables, as these data were not uniformly available in the retrospective records.’’

 

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for allowing me to review the revised version of this paper entitled "Limited Diagnostic Yield of Routine Gastroscopy in FIT-Positive Patients" by Khader et al. In this revised version the authors have addressed the concerns raised regarding the original manuscript and I have no further comments.