Article Text

PDF

Novel prep-less X-ray imaging capsule for colon cancer screening: a feasibility study
  1. Nathan Gluck1,
  2. Elizabeth E Half2,
  3. Vered Bieber3,
  4. Doron Schwartz4,
  5. Yulia Ron1,
  6. Ian Gralnek3,
  7. Amir Klein2,
  8. Jesse Lachter2,
  9. Moshe Shoni Levy5,
  10. Menachem Moshkowitz1,5,
  11. Nadir Arber1,5
  1. 1Department of Gastroenterology, Tel Aviv Sourasky Medical Centre, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
  2. 2Department of Gastroenterology, Rambam Health Care Campus, Haifa, Israel
  3. 3Department of Gastroenterology, Ha’emek Medical Centre, Technion Institute of Technology, Haifa, Israel
  4. 4Department of Gastroenterology, Soroka Medical Center, Ben-Gurion University, Beer Sheva, Israel
  5. 5Integrated Cancer Prevention Center, Tel Aviv Sourasky Medical Centre, Sackler Faculty of Medicine, Tel Aviv University, Tel-Aviv, Israel
  1. Correspondence to Professor Nadir Arber, The Integrated Cancer Prevention Center, Tel-Aviv Sourasky Medical Center, Tel-Aviv 6423906, Israel; nadira{at}tlvmc.gov.il

Statistics from Altmetric.com

Message

The X-ray imaging capsule overcomes the requirement for bowel preparation, a major deterrent for colorectal cancer (CRC) screening. This prospective multicentre study examined capsule safety and efficacy (compared with fecal immunochemical test (FIT)) for polyp detection confirmed by ensuing colonoscopy. In 45 analysed patients, capsule and FIT sensitivity for polyps was 44% and 37%, respectively. Capsule sensitivity increased to 78% when >50% of the colon surface area was imaged, with a linear correlation between imaged area and sensitivity. An updated scanning algorithm, retrospectively implemented on these study data, dramatically increased numbers of subjects with imaged area >50%, from 21 to 41/45 (46% to 91%). Capsule specificity for polyps approached 90%. Transit time was 52±32 hours. No device-related serious adverse events occurred. Average X-ray dose was 0.05 mSv.

In more detail

Many patients forego screening by colonoscopy due to the need for cathartic preparation,1 2 posing a major obstacle to implementation of current screening programmes.3 4 A test not requiring laxative preparation could increase adherence with screening.1 Moreover, inadequate preparation for colonoscopy impairs adenoma and cancer detection rates and often necessitates repeating of the procedure.5 A prep-less modality would not suffer from these limitations. In a previous proof-of-concept study, a prep-less screening capsule was shown to be safe, capable of 3D reconstruction of the colonic wall and lumen and identifying adenomatous polyps.6 In a prospective multicentre comparative study, we studied the safety and efficacy of a first version of the capsule for identifying colonic polypoid lesions and masses compared with FIT and colonoscopy.

The system has previously been described and optimised for correctly imaging polyps greater than 6 mm6 ,7 (online supplementary video). A scanning control algorithm (SCA), optimised based on clinical data and capsule motility statistics, has since been embedded in the system, allowing procedures to be performed autonomously in the patients' familiar environment. Novel pressure sensors accurately time capsule entrance into the caecum and also suggest contact with protruding lesions, that is, polyps (figure 1). Patients underwent the capsule procedure and concomitant FIT followed by colonoscopy as the gold reference, with removal of all visualised polyps. Analysis of capsule data was blinded to FIT and colonoscopy results. Sensitivity and specificity of the capsule findings were calculated with reference to FIT and colonoscopy, as a function of the percentage of the total colon surface area effectively imaged by the scanning capsule.

Supplementary file 2

Figure 1

Pressure sensing by the capsule. A scanned segment of the colon (with margins before and after) is depicted as a function of time. Scanning events by the capsule are represented by vertical grey lines. The capsule senses pressure and temperature using embedded sensors. The pressure plot is shown in black and temperature plot in red. Pressure peaks (seen here between the first and second cluster of scans) are marked by circles.

Sixty-six patients (age 37–79, 40 male) were enrolled, 43 had unresected polyps and 23 were average risk with no prior colonoscopy. Forty-five procedures were analysed. Drop-outs included 13 technical failures and eight physiological/anatomical outliers. All patients continued daily routine on normal diet. Infrequent adverse events, including soft stool, mild abdominal pain, headache and rash under the back sticker, resolved spontaneously within a few hours. One serious adverse event occurred: an abdominal wall haematoma unrelated to the device or capsule procedures. The average total X-ray dose exposure was 0.051 mSv. Average total transit time of ingested capsules was 52±32 hours. Sensitivity of FIT for polyp detection was 37% with a specificity of 100%. A per-patient analysis was performed for all 45 patients. Sensitivity and specificity of capsule findings, as a function of the imaged colonic surface area, are presented in table 1. While overall sensitivity for polyp detection was 44% (p=0.5 relative to FIT), it increased to 78% in the subgroup with colon imaging of >50% (n=19, p=0.006) and to 100% in the subgroup with imaging >70% (n=12, p=0.001). At the same time, specificity remained high at 90% and 86%, respectively.

Table 1

Capsule polyp detection relative to colonoscopy as a function of colon imaging

Overall sensitivity for FIT combined with the capsule was 67%, significantly more than FIT alone (p=0.0046). When FIT was followed by capsule with imaging >50%, an additional increase in sensitivity was observed (89%, p=0.0005).

After trial termination, an improved SCA version was developed and optimised by retrospectively analysing the capsule position data recorded during the study procedures. A substantial improvement in colon scan coverage was achieved: of the 45 procedures, the proportion of individuals obtaining imaging of >50% increased from 47% (n=21) to 91% (n=41) (figure 2).

Figure 2

Colon imaging according to scanning algorithm. Colon scan imaging is expressed as the percentage of colon adequately scanned by the capsule. Distribution of imaging levels in study patients is shown for the original and advanced algorithm.

Find more details on methods and results in the online supplementary file.

Supplementary file 1

Comments

In the first prospective multicentre trial the safety and efficacy of a new imaging modality for CRC screening is shown. The X-ray capsule imaging system can detect polyps without the need for colon cleansing and with minimal if any discomfort to the patient. This advantage may persuade more patients to undergo colon cancer screening.

No device or procedure-related SAE occurred and the procedure was well-tolerated by the participating subjects. Importantly, since the system is operator-independent and runs autonomously, daily routine and familiar environment could be maintained.

A significant drop in transit time (and ensuing study duration) relative to the pilot study, from an average of 76 to 52 hours, will make the procedure more comfortable and easy to perform.

A strong correlation existed between colon imaging and sensitivity for polyp detection: higher surface area imaged correlated strongly with improved sensitivity. Interestingly, >50% overall colon imaging coverage yielded >78% sensitivity for polyp detection. This may be the result of the tendency of the capsule to slow down near polyps, even small polyps, producing higher scan density. For a colon segment to be considered imaged, a certain number of scans is required. This process is not linear. The specificity for polyp detection was consistently high even in subgroups with high sensitivity. The new SCA version retrospectively analysed pushed the majority of patients (91%) above 50% colon imaging, which in this study showed sensitivity of 78%–100% for polyp detection, comparable with performance of colonoscopy. This improvement in colon imaging may yield improvement in the efficacy of the capsule system for colonic polyp detection. A follow-up study with the new SCA algorithm implemented in the system is currently being launched to confirm this notion prospectively.

This study has several limitations. It was conducted with the first version of the X-ray capsule that suffered from some manufacturing defects and a non-optimised SCA algorithm. Additionally, the study was relatively small and enriched with more polyps than expected in the average risk population. This may create statistical bias. Two gastroenterologists alternated as members of the team that analysed the cases, hence interobserver variability during the review process may exist. In addition, the findings from the capsule procedures were not shared with the endoscopist before the follow-up colonoscopy. All limitations will be addressed in the upcoming study. 

References

View Abstract

Footnotes

  • NG and EEH contributed equally.

  • Contributors NG, EEH: acquisition of data, drafting the manuscript. MM: study design, acquisition of data. NA: study design, acquisition of data, critical review of manuscript. JL, DS, AK, VB, YR, IG, MSL: acquisition of data.

  • Funding The study was funded by Check-Cap Ltd, Mount Carmel, Israel.

  • Competing interests NG and NA report receiving consulting fees from Check Cap Ltd.

  • Patient consent Not required.

  • Ethics approval IRB in all participating sites, as well as by the national ethics committee of the Israeli Ministry of Health.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.