This study was approved by the National Hospital Organization Shimoshizu Hospital Ethics Committee. This was not considered to be a clinical trial, as the procedures were performed as a part of routine clinical practice. Written informed consent was obtained from the patients to perform colonoscpy. Informed consent was obtained to perform abdominal US, but written forms were waived. Written informed consent for inclusion in the study was waived, as patient records were anonymized and retrospectively analyzed.

The medical records of patients who were treated at the National Hospital Organization Shimoshizu Hospital from March, 2010 to January, 2015 were retrospectively analyzed. Enrolled patients were required to meet the following inclusion criteria: Subjected to abdominal US prior to colonoscopy, computed tomography (CT), or magnetic resonance imaging; underwent surgery at the National Hospital Organization Shimoshisu Hospital; and diagnosis pathologically confirmed. The patients underwent abdominal US for anemia, abdominal pain and bowel obstruction. Certain patients were subjected to abdominal US for screening. Following diagnosis of CRC with abdominal US, colonoscopy was performed in all the patients. The exclusion criteria were as follows: Subjected to abdominal US after the diagnosis of CRC with colonoscopy; subjected to abdominal US with the suspicion of CRC with CT or magnetic resonance imaging; and not subjected to surgery. The enrolled patients were restricted to those whose surgical specimens were available to investigate the depth of invasion. The enrolled patients included 5 men (aged 74.0±0.8 years) and 10 women (aged 73.0±12.0 years).

Abdominal US was performed by Senior Fellows of the Japan Society of Ultrasonics in Medicine (M.T. and F.S) using the SSA-700A diagnostic US system (Toshiba Medical Systems Corporation, Ohtawara, Japan) with a 3.75-MHz curved-array probe (PVT-375BT; Toshiba Medical Systems) or an 8.0-MHz linear-array probe (PLT-805AT; Toshiba Medical Systems) in the US unit. The small and large intestines were scanned following routine abdominal US when intestinal diseases, such as ileus, were suspected, or when the patients had anemia.

The diagnostic criterion for CRC was localized irregular wall thickening (Fig. 1A) or a hypoechoic mass with a hyperechoic mass (pseudokidney sign; Fig. 1B) (10). The former is a common finding in patients with CRC (12), while the latter represents tumor tissue and air in the residual lumen (13).

Wall thickness was measured with abdominal US between the mucosa and serosa borders. Wall thickness was analyzed to differentiate between CRC and the surrounding normal colonic wall. The US findings were evaluated in terms of shape, stratification and contour. Shape was divided into wall thickening (W; Fig. 1A) and mass (M; Fig. 1B). Stratification was observed due to the different layers of the colonic wall (12) and patients were divided into two groups, namely preserved (Fig. 1C) or lost stratification (Fig. 1D). Irregular contour is considered to be an US characteristic of CRC (10). A proportion of the patients had a smooth contour (Fig. 1E), while the majority exhibited irregular contour (Fig. 1F).

The depth of invasion was determined by two pathologists (K.F. and T.K). The analyzed specimens were obtained via surgical resection. Patients referred to other hospitals for pathological analysis and those treated conservatively were excluded from the analysis.

The blood test variables analyzed were white blood cell count, hemoglobin (Hb), C-reactive protein (CRP), carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA 19-9).

The mean wall thickness was compared between CRC and the surrounding normal colonic wall with one-way analysis of variance. The Chi-square test was applied to analyze the correlation between the shape of CRC (W or M) and stratification or contour. The Chi-squared test was also applied to analyze the correlation between depth of invasion and the shape of CRC (W or M), stratification, or contour. The threshold value of wall thickness to diagnose CRC was investigated with receiver operating characteristic (ROC) curve analysis. A P-value of <0.05 indicated statistically significant differences. JMP 10.0.2 software (SAS Institute, Cary, NC) was used for all statistical analyses.

The wall thickness in CRC and in the surrounding normal colonic wall was measured and plotted in Fig. 2A. The average wall thickness was 2.8±0.4 mm in the surrounding normal tissue and 12.7±5.2 mm in CRC. The wall was significantly thicker in CRC compared with the normal colonic wall (P<0.0001). The thickness of normal colonic wall was <3.0 mm, while it was >4.3 mm in CRC. As shown in Fig. 2A, there may be a threshold value for the diagnosis of CRC using wall thickness. ROC curve analysis was performed to investigate the threshold value for the diagnosis of CRC using abdominal US. The calculated threshold value was 4.3 mm. The sensitivity and specificity at this value were both 100%.

To determine whether there is an association between the shape of CRC and stratification or contour, Chi-square test was performed (Table I). Stratification was preserved in W, while it was lost in M (P=0.0196). The correlation between shape and contour was not significant (P=0.4356).

To analyze the association between the depth of invasion and the shape, stratification, or contour, Chi-square test was performed (Table II). No significant correlation was observed between any of the variables.

To assess the background of patients diagnosed with CRC using abdominal US, blood test variables were analyzed (Table III). The Hb level was below the normal range, whereas the CRP, CEA and CA 19-9 levels were above the normal range.