| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Research |
1 Department of Radiology, Seoul National University Bundang Hospital, 300
Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi-do, 463-707, Korea.
2 Department of Radiology, Asan Medical Center, University of Ulsan College of
Medicine, Research Institute of Radiology, Seoul, Korea.
3 Health Promotion Center, Seoul National University Bundang Hospital,
Seongnam-si, Gyeonggi-do, Korea.
4 Department of Radiology and Institute of Radiation Medicine, Seoul National
University Hospital, Seoul National University College of Medicine, Seoul,
Korea.
5 Department of Internal Medicine, Seoul National University Bundang Hospital,
Seongnam-si, Gyeonggi-do, Korea.
Received October 30, 2007;
accepted after revision April 13, 2008.
Address correspondence to K. H. Lee
(kholee{at}snubhrad.snu.ac.kr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. In 2005 and 2006, 1,015 Korean adults (609
men and 406 women; mean age, 51 years) underwent screening CTC using a 16-MDCT
scanner and an automated CO2 delivery system. During the study
period, the protocols were changed to use less vigorous purgation and lower
radiation doses; fecal tagging (n = 890) and primary 3D
interpretation (n = 966) were generally used. CTC results were
categorized as C0, inadequate; C1, no significant polyp; C2, one or two 6- to
9-mm polyps; C3, polyps 
10 mm or three 6- to 9-mm polyps; and C4,
mass. Patients with positive CTC results were referred to gastroenterologists
for follow-up or management planning.
RESULTS. Categories C0–C4 were assigned to 21 (2.1%), 916 (90.2%), 54 (5.3%), 23 (2.3%), and one (0.1%) patients, respectively. Fifty-four patients with C4 (n = 1), C3 (n = 20), or C2 (n = 33) underwent subsequent optical colonoscopy: complete (n = 53) and incomplete (n = 1). Per-patient positive predictive values (PPVs) for categories C3–C4 and C2–C4 were 90% (18/20) and 74% (39/53), respectively. Per-polyp PPVs at 10- and 6-mm thresholds were 92% (22/24) and 69% (45/65), respectively. The diagnostic yield for advanced neoplasm was 1.5% (15/1,015).
CONCLUSION. Our results seem comparable to Western experiences, showing that a successful screening CTC program can be reproduced in an Asian population.
Keywords: Asian • CT colonography • screening
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Western results of screening CT colonography (CTC) [7, 8] are encouraging, especially for Korea [9] and Japan, where the number of CT scanners in use per population is the greatest in the world [10]. However, it still remains unclear whether the success of screening CTC in the "centers of excellence" [11, 12] can be generalized to other institutions. This issue is important [13] if CTC is to become a truly successful screening tool. Large studies in screening subjects have been limited to Western populations [7, 8, 11, 14–16]. Despite many reports about the technical aspects of CTC, no results have been published of screening CTC in an Asian population.
The purpose of this study was to report the results of screening CTC in an asymptomatic average-risk Asian population in 2005 and 2006.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bowel Preparation
We occasionally changed the protocols for bowel preparation, scanning, and
interpretation (Table 1) as we
introduced new techniques into our practice. Two radiologists who interpreted
CTC studies changed these protocols on the basis of published study results,
our own experience, and advice from radiologists at other institutions. They
took into consideration technical feasibility, resource availability,
estimated effects on patient comfort and compliance, and estimated effects on
polyp detection accuracy. The protocol changes and their rationales were
announced to our staff.
|
At the beginning of the study period, the bowel preparation entailed a low-residue diet the day before examination and two doses of 45 mL of oral sodium phosphate (Colclean, Taejoon Pharmaceutical). At the end of the study period, we used a low-residue diet the day before examination, 8.1 g of oral magnesium citrate (25 g of Magcorol powder, Taejoon Pharmaceutical), 10 mg of bisacodyl suppository (Dulcolax, Boehringer Ingelheim), and three doses of 200 mL of 4.6% weight/volume (w/v) barium for stool tagging (EZ-CT, Taejoon Pharmaceutical). Over the study period, we gradually changed to use less vigorous purgation, taking special caution to increase patients' compliance. Accordingly, we regularly revised a patient brochure illustrating the importance of bowel preparation and giving instructions about taking medications.
Scanning
An automated CO2 delivery system (PROTO-CO2L, E-Z-EM)
was used. No spasmolytics were used. In all patients, unenhanced supine and
prone acquisitions were obtained using a 16-MDCT scanner (IDT16, Philips
Healthcare) with 16 x 1.5-mm collimation, a rotation speed of 0.5
seconds, a pitch of 1.17–1.25, reconstruction thickness of 2 mm, and re
con struction interval of 1 mm. Tube current was automatically modulated. At
the initial stage, the effective tube current ranged from 25 to 45 mAs; we
later lowered it to 13–20 mAs. We empirically lowered our standard tube
potential from 120 to 90 kVp during the second year because we believed this
would not signi ficantly hinder polyp detection in our patients, who usually
have smaller abdominal circumferences than Westerners. At the end of the study
period, the patient dose was estimated to be 0.8–1.0 mSv for combined
supine and prone scans. This effec tive dose was estimated by multiplying the
dose–length product recorded at the time of scanning by a
region-specific normalized con version factor (0.017 mSv ·
mGy–1 · cm–1)
[17].
Interpretation
CTC scans were prospectively interpreted by one of the two radiologists
with prior experience in reading at least 70 cases verified by optical colono
scopy. The workstations for interpretation varied and included the Aquarius
Workstation (Terarecon, AquariusNET), Extended Brilliance Workspace (Philips
Healthcare), and Rapidia (Infinitt). The workstation used depended largely on
availability at the time of interpretation and the reader's preference. Our
standard interpretation method was primary 3D fly-through with 2D problem
solving. We did not use electronic subtraction of tagged feces because the
remaining artifacts distracted the readers. All detected polyps 6 mm were
recorded by storing images displaying lesion appearance, size, and location.
Lesions were measured with an electronic caliper after appropriate
magnification on 3D images. This reviewing method yielded typical reading
times of 
15 minutes for colonic interpretation. If numerous tagged fecal
residues were scattered in a colon segment, the reader also reviewed this
segment in 2D mode. In 49 patients (4.8%), too much fecal residue was
scattered throughout the entire colon, and the readers had to abandon primary
3D interpretation and rely on 2D viewing.
Descriptive reports were made during the early period (n = 448);
however, we later used structured report formats based on the CTC Reporting
and Data System [18]
(n = 567) as this was published (C0, inadequate study that would
potentially miss a polyp 10 mm; C1, no significant polyp; C2, one or two
6- to 9-mm polyps; C3, polyps 10 mm or 3 polyps 6- to 9-mm; and C4,
malignant-appearing mass). Before this guideline was used, the radiologists
generally recommended follow-up optical colonoscopy for patients with polyps
6 mm, with various intervals according to the polyp size and number and
their diagnostic confidence level. Because the structured report was used,
recommendations for follow-up or management planning were provided on the
basis of the aforementioned guideline (C0, repeat examination; C1, follow-up
examination in 5–10 years; C2, follow-up examination in 3 years; C3 and
C4, immediate optical colonoscopy). The radiologists also recorded significant
technical limitations potentially hindering polyp detection, such as
incomplete colon distention or too much fecal residue impeding primary 3D
interpretation.
Management After Positive Results at CTC
The Health Promotion Center physicians explained the CTC results to each
patient. Patients with positive (C2–C4) or inadequate (C0) results were
referred to one of four gastroenterologists for follow-up or management
planning. Referring to the radiologist's recommendation in the CTC report, the
gastroenterologist suggested an appropriate plan to each patient. Patients
with the CTC results of category C0 were generally offered follow-up optical
colonoscopy or CTC. Patients in category C2 or C3 were offered either
immediate or follow-up optical colonoscopy for polypectomy or CTC
surveillance. For patients in category C4, immediate optical colonoscopy for
biopsy was recommended. The timing and method of the follow-up examination
were individualized on the basis of lesion size and number, patient age,
comorbidities, the technical limitations described in the CTC report, the
patient's preference, and the gastroenterologist's own pattern of practice.
Whether a patient should undergo optical colonoscopy was finally deter mined
at the discretion of the gastroenterologist and by the patient's informed
decision.
When indicated, optical colonoscopy was performed by the gastroenterologist (range of experience, 7–20 years), who reviewed the captured images showing the CTC findings before optical colonoscopy. Same-day optical colonoscopy could not be offered because of the limited capacity of the endoscopy unit. Endoscopic techniques included cecal intubation and sequential withdrawal of the scope (CF-Q260AL, Olympus Optical) for polyp detection. Polyp size was measured by comparison with open-biopsy forceps. Polyp locations were visually estimated. Whenever possible, all detected polyps deemed to be of clinical importance were retrieved or biopsy was performed for histologic evaluation.
Polyp Matching
A study coordinator retrospectively translated the descriptive reports into
structured reports according to the aforementioned guideline. In June 2007, a
radiologist with 3 years of experience in CTC interpretation and a
gastroenterologist together reviewed the CTC and optical colonoscopy reports
and captured images and determined per-polyp concordance between the results
of optical colonoscopy and those of CTC using the criteria of Pickhardt et al.
[7]. To be considered a
positive match, a polyp had to be located in either the same segment or an
adjacent segment and the polyp size measured on optical colonoscopy had to be
within a 50% margin of error compared with the size measured on CTC. Polyp
morphology was classi fied as sessile, pedunculated, flat, or mass.
Statistical Analysis
CTC results were tabulated according to the category. Optical colonoscopy
referral rate was calculated. Polyps were categorized according to the size
measured at CTC. In per-polyp analysis, polyps ( 6 mm) for which CTC and
optical colonoscopy findings were concordant were counted as true-positives.
Polyps ( 6 mm) detected at CTC without a matched lesion at optical
colonoscopy were counted as false-positives. Polyps ( 6 mm) detected at
optical colonoscopy but not at CTC were counted as false-negatives. In
perpatient analysis, true-positive cases were defined as patients with at
least one true-positive lesion of a given size category. False-positive cases
were defined as patients with at least one false-positive polyp and without
any true-positive polyps. The positive predictive value (PPV) of CTC findings
was calculated. Neoplastic lesions were defined as adenoma or adeno carcinoma.
Advanced neoplastic lesions were defined as adenomas 10 mm or as any
lesion that contained villous features ( 25%), high-grade dysplasia, or
invasive carci noma, regardless of size
[19–21].
Complications after bowel prep aration or CTC were counted. Descriptive
statistics were reported.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The monthly number of patients who underwent CTC decreased in the second year (Table 1), mainly because the capacity of the optical colonoscopy unit in the Heath Promotion Center increased by recruiting gastroenterologists.
Categories of CTC Results
In 21 patients (2.1%) with otherwise negative findings at CTC, at least one
colonic segment was considered nondiagnostic (C0) because of luminal collapse
or abundant untagged feces. None of these patients chose to undergo optical
colonoscopy or repeat CTC during the study period. Category C4 was reported in
one (0.1%) patient, who underwent subsequent optical colonoscopy. Category C3
was reported in 23 (2.3%) patients, 20 of whom underwent subsequent optical
colono scopy during the study period. Category C2 was reported in 54 (5.3%)
patients, 33 of whom underwent subsequent optical colono scopy. The overall
CTC test-positive rate (C2–C4) was 7.7% (95% CI, 6.1–9.5%)
(Fig. 1).
|
If all patients with either positive (C2–C4) or inadequate (C0) CTC results had undergone subsequent optical colonoscopy, the maximum referral rate would have been 9.8% (99/1,015). If all patients in categories C3 and C4 had undergone subsequent optical colonoscopy, the referral rate would have been 2.4% (24/1,015).
Per-Patient PPV
Concordant findings were identified in 39 of the 53 patients with complete
optical colonoscopy, yielding an overall per-patient PPV of 74% (95% CI,
60–85%). In these patients, at least one matching lesion was neoplastic
in 24 patients, yielding a PPV of 45% (32–60%; 24/53) for neoplastic
polyps. For patients in categories C3 and C4, the PPV for polyps of any
pathology was 90% (68–99%; 18/20), and the PPV for neoplastic polyps was
70% (46–88%; 14/20) (Table
2). The diagnostic yield for advanced neoplasms was 1.5%
(15/1,015).
|
Per-Polyp PPV
The PPV for individual polyps was 69% (45/65) and 92% (22/24) at 6- and
10-mm thresholds. At the 6-mm threshold, the PPVs for sessile, pedunculated,
and flat polyps and mass were 68% (34/50), 75% (9/12), 50% (1/2), and 100%
(1/1), respectively. The per-polyp PPVs for neoplasms were 43% (28/65) and 58%
(14/24) at 6- and 10-mm thresholds (Table
3 and Fig. 2A,
2B,
2C). Histologic diagnoses for
matched neoplasms included tubular ade noma (n = 18), tubulovillous
adenoma (n = 6), and adenocarcinoma (n = 4). Seventeen le
sions were advanced neoplasm, including 14 lesions 10 mm, two smaller
adenomas with villous features, and a 6-mm intra mucosal adenocarcinoma.
Non-neoplastic matched lesions included nonspecific in flammation (n
= 6), hyperplastic polyps (n = 5), granulomatous inflammation
(n = 3), fibro epithelial polyp (n = 1), and inadequate
specimen (n = 2).
|
|
|
|
Cancer Detection
In each of four patients (C2, C3, C3, and C4, respectively), one lesion
detected at CTC (measuring 6, 24, 30, and 40 mm) was confirmed as being or
harboring adenocarcinoma. Two of these patients under went complete
polypectomy at the time of optical colonoscopy, and two underwent surgery for
definitive resection.
Missed Polyps
Three individual polyps measuring 6–9 mm, including two adenomas and
one nonspecific inflammation, were identified at opti cal colonoscopy but not
at prospective CTC in three (6%) of the 53 patients who had complete optical
colonoscopy. One of these polyps had a sessile shape, one was pedunculated,
and one was flat. In each of these patients, additional polyps of similar or
larger size were found at prospective CTC, showing concordant optical
colonoscopy findings; and categories for CTC results would have been unchanged
even if the missed lesions had been detected at CTC. Otherwise, no additional
polyps 6 mm were detected at optical colonoscopy.
Complications
In one patient (0.1%; 95% CI, 0.003–0.5%), a symptomless extraluminal
air collection was observed along the ascending colon at the retroperitoneum
on CTC images that spontaneously disappeared at follow-up CT without
complication. It is unclear whether this air collection was an iatrogenic
perforation or preexisting pneumatosis coli. Other wise, no clinically notable
complications occurred in any patient.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
10 mm was 92%. The diagnostic yield for advanced neoplasm was 1.5%
(15/1,015). Our CTC positive rate (2.4% for C3 and C4, 7.7% for C2–C4) was lower than in recent large studies of screening CTC (4.7% for C3 and C4, 10.8–12% for C2–C4) [8, 14]. Of the many factors contributing to this difference, one plausible explanation is that polyp pre valence was presumably lower in our cohort. Our population included asymptomatic average-risk Asians with a mean age of 51 years, compared with the population in the study by Pickhardt et al. [8], which consisted of U. S. residents, of whom 3.5% were symptomatic patients and 6.8% were high-risk patients, with a mean age of 58 years. Although this is debatable [6], polyp prevalence is generally believed to be lower in Asians than in Westerners. The reported prevalence of advanced neoplasms in Asia is 4–4.5% [6, 22, 23], compared with 5.4–11.7% in the West [19–21]. The difference in age is also significant. More than a twofold increase of advanced neoplasm has been observed from the fifth to seventh decades [20].
Our per-patient PPV (90% for C3 and C4, 74% for C2–C4) was lower than
that in the recent results of Pickhardt et al.
[8] (95% and 92% at 10- and
6-mm thresholds); how ever, our per-patient PPV was higher than theirs in
their initial screening trial (67% and 59%)
[7,
24] and higher than in the
results of Copel et al. [25]
in patients with incomplete optical colonoscopy (78% and 58%). This
variability can be attributable to differences in polyp prevalence, patient
selection for sub sequent optical colonoscopy, reader experi ence, and
technical details. These PPVs tend to be higher than those of double-contrast
barium enema (67% and 62%)
[26], especially for polyps
10 mm.
One might argue that the low CTC positive rate in this study is attributed
to low sensitivity in polyp detection. We cannot directly assess the
sensitivity of our screening CTC because not all patients were evaluated with
optical colonoscopy. Moreover, PPV was not calculated for all patients whose
CTC results were positive because some of them, especially in the C2 group,
did not undergo optical colonoscopy. This may be explained by the tendency of
gastroenterologists and patients to defer optical colonoscopy when the
neoplastic risk is small [13,
27,
28]. In our patients who
underwent complete optical colonoscopy, the false-negative rate was 6%, close
to the 7% in the study by Pickhardt et al.
[8]. Although this is debatable
[29], the sensitivity of
technically competent CTC is believed to be consistently high, especially for
polyps 10 mm. A meta-analysis
[29] reported the perpatient
sensitivity at the 10-mm threshold was 95% (95% CI, 92–99%) if MDCT
scanners were used. Sensitivities of 93% (73–98%) and 86% (75–93%)
were reported at 10- and 6-mm thresholds, respectively, in another
meta-analysis [30] using
stricter inclusion criteria for technical details of bowel preparation,
scanning, and inter pre tation, all of which were met by our protocols.
Therefore, it would be reasonable to assume that our sensitivity lies within
the reported ranges.
Our diagnostic yields for neoplasms 10 mm and advanced neoplasms were
1.3% (13/1,015) and 1.5% (15/1,015), respectively, which are seemingly lower
than in Western studies (2.6% for neoplasms 10 mm and 2.4% for advanced
neoplasms) [8,
11]. On the other hand, all of
these diagnostic yields of CTC are numerically lower than the reported yields
of double-contrast barium enema for advanced neoplasms (6.2%)
[26] or than the yields of
optical colonoscopy for neoplasms 10 mm (5.1–9.5%)
[19,
21,
31] and advanced neoplasms
(4–4.5% in Asia, 5.4–11.7% in the United States)
[6,
19–22].
However, these numeric comparisons are not truly meaningful because the
diagnostic yield depends heavily on disease prevalence. It should be noted
again that, compared with others
[6,
8,
11,
19–23,
26,
31], our cohort is likely to
have a lower prevalence of neoplasms considering ethnicity, mean age, and
absence of symptoms or risk factors of colorectal cancer. Our diagnostic
yields represent conservative estimates of actual yields for CTC because
optical colonoscopy was not performed in all patients with positive CTC
results. Further more, our yield for advanced neoplasms would inevitably be
low er than that of optical colonoscopy be cause we reported only lesions
6 mm at CTC.
Over time, we gradually changed our protocols to use less vigorous purgation and lower radiation doses, with the intention of expanding screening CTC. These two factors were important in appealing to pa tients and referring physicians, and the outlook for screening CTC was initially not favorable in our center. These protocol changes were sometimes inconvenient for the reader; however, readers soon became more tolerant. Polyp detection also had to be unhindered, and therefore, gradual changes in the protocols were preferred to sudden reform until the readers were familiar with the new protocols.
Our final radiation dose was 0.8–1.0 mSv, lower than those in other studies [32]. Rigorous dose reduction has been criticized for hampering depiction of extracolonic findings [12, 32]. We did not analyze extracolonic findings, which is a limitation of this study. According to other studies [33, 34], the extracolonic abnormalities that truly require intervention (mainly renal cell carcinoma) were found in less than 1% of patients, and pulmonary lesions and aortic aneurysms that could probably have been detected in noisier lower-dose scans were excluded. Apart from the significance of the extracolonic findings, we believe doses should be lower for Asians, who typically have a smaller body mass index than Westerners [35].
Our study had other limitations because of its retrospective nature. First, our bowel preparation, scanning, and interpretation proto cols were heterogeneous; subgroup analy sis was impractical because of the limited number of true-positive cases and the many confounding variables. Second, the subjects self referred for the screening at their own expense. They were likely to be more health-conscious and of higher socioeconomic status; hence, they were not ideally representative of the general population. Third, the reasons for the false-positive detection at CTC were not evaluated because we did not review the scans in a retrospective manner. We postulate that the main reason for false-positive findings was untagged feces. Fourth, we did not evaluate the patients' acceptance or discomfort regarding the CTC procedure.
In conclusion, despite these limitations, this study represents the first results of screening CTC in an Asian population using current techniques. Our principal interest was to determine whether the success in leading Western centers [8] could be reproduced in an Asian population in which colorectal cancer is rapidly increasing. Considering the lower polyp prevalence presumed in our cohort, our test-positive rate, optical colonoscopy referral rate, PPV, and diagnostic yields were comparable to those in more experienced Western centers, showing that a successful screening CTC program can be reproduced.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
