Clinical information

Twenty-eight patients with rectal cancer who received TME surgery at the Hunan Cancer Hospital from October 2016 to April 2017 were selected. Inclusion criteria: (1) patients with MRI-based clinical stage ≥ T3; (2) patients who refuse to receive radiotherapy and chemotherapy before surgery; (3) all patients who underwent endoscopic biopsy and adenocarcinoma was pathologically confirmed before surgery (which was referred to as “rectal cancer” in this paper); (4) patients for whom distant metastasis was not indicated through chest and abdominal CT and bone scintigraphy examination. Those 28 patient pathological specimens were retrospectively measured in 2022. This study was approved by the Hunan Cancer Hospital Ethics Committe and was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki.

Acquisition of preoperative image

All MRI examinations were performed using a 1.5-Tesla MRI scanner (Optima®MR 360; GE Medical Systems, Milwaukee, WI, USA) using a phased-array body coil. The patients were placed in the supine position, and preoperative pelvic enhanced standard MRI sequences were performed on all patients within 1 week of operation. Perpendicular to the rectal axis, the MRI scan spacing ranges from 0.3 to 0.5 mm, and scan thickness is 3 mm. The gross tumor volume (GTV) of the rectal cancer was delineated according to abnormal signals on T2WI. The delineation and thickness measurements were performed by an imaging diagnostic physician with more than 10 years of experience (L.W.).

Processing of pathological specimens

TME surgery was conducted on all 28 patients in this study. Immediately after operation, processing of pathological specimens and measuring tumor retraction factors are illustrated and described in Fig. 1. This process was used to measure and calculate the retraction factor of specimen processing.

Fig. 1

Processing of pathological specimens and measuring tumor retraction factors. A Immediate surgical specimen was cut vertically along the opposite side of the main tumor body, and then the tumor was evenly cut into tissue slices of 5 mm in thickness. B The middle piece of tissue slice was fixed to a foam plate with pins, the gross tumor range was mapped by naked eyes and the tumor depth and height were measured (marked as Dpath and Hpath). C The adjacent 5 mm tissue slices were fixed to the foam plate similarly, then a surgical knot was tied at the midpoint of the mesangial side of the tumor and labeled as the standard point A, and several knot reference points B1, B2, B3, B4 and B5 were set in different directions of the mesentery. D After 24 h formalin dehydration. E The gross tumor was outlined on all layers of sagittal T2WI and was superimposed in the middle image. The tumor height in MRI was the vertical distance between the top and bottom tips of the outlines and marked as HMR.F the corresponding layer of MRI T2WI images at the widest part of the tumor was marked as DMR. BM: basis of measurement

Calculation of tumor retraction factors

Tumor deformation and retraction occurred during the process of TME surgery and postoperative pathological section preparation. The calculation of the retraction factor was divided into two steps: after the in vivo tumor was isolated, the first retraction occurred from the lack of pelvic fascia tractions, which was defined as the “in vivo–in vitro” retraction factor (R1); in the process of pathological specimen preparation, the in vitro tumor shrunk for the second time because of tissue dehydration, which was defined as the specimen processing retraction factor (R2).

Calculation of R 1

The middle tissue slices of immediate surgical rectal cancer specimens (Fig. 1B) were obtained and fixed on a foam plate, as described above. The gross tumor range was mapped by the naked eye. To calculate the retraction factor in the lateral (peripheral) direction, the widest thickness was measured, which was marked as Dpath. The MRI T2WI images of all tumor layers were obtained, and the corresponding layer at the widest part of the tumor was selected (Fig. 1F), which was marked as DMR. The R1 value in the lateral direction was calculated using the following formula: R1-L = Dpath/DMR. To calculate the retraction factor in the superior/inferior direction, the vertical height of the gross tumor in the specimen was measured and marked as Hpath (Fig. 1B). The gross tumor was outlined on all layers of sagittal T2WI and superimposed in the middle image. The tumor height in the MRI was the vertical distance between the top and bottom tips of the outlines and marked as HMR (Fig. 1E). The R1 value in the superior/inferior direction was calculated using the following formula: R1-S/I = Hpath/HMR.

Calculation of R 2

Because the degrees of tumor retraction varied in different directions, R2 was divided into three parts: inferior retraction (R2-I), lateral retraction (R2-L), and superior retraction (R2-S). The retraction factor (R2) of specimen processing was calculated using the changes of relative position between A and B1, B2, B3, B4, and B5 before dehydration (Fig. 1C) and after dehydration (Fig. 1D). AB1-5 represents the distance between A and the different B points. The formulas for calculating the retraction factors are as follows:

$$R_}=\frac1\text}1\text},$$

$$R_}= \left (\frac2\text}2\text}+\frac3\text}3\text}+\frac4\text}4\text} \right)/3$$

$$R_}=\frac5\text}5\text}.$$

Measuring the distance of microcarcinoma

For pathological specimens, the middle 5 mm tissue specimens in 28 cases were divided into three paraffin tissues (separated by BM lines) of the lateral, inferior, and superior sides of the gross tumor (Fig. 2A). All paraffin tissues were sliced continuously at intervals of 2 mm, and each slice was 4 μm thick. The central three pieces were taken, with a total of 252 pieces taken for observation and examination of tumor micro-foci.

Fig. 2

Measurement of micro-foci distance away from rectal gross tumor. A Schematic show of micro-foci distance measurement. B Micro-foci in the superior side. C Micro-foci in the lateral (peripheral) side. BM: basis of measurement; DM-I: distance of micro-foci inferiorly; DM-S: distance of micro-foci superiorly; DM-L: distance of micro-foci laterally

The sections were stained with eosin and methylene blue, and the whole section was first reviewed by a senior pathologist (X.Y.C.) using a low-power microscope (10×) to sketch the peripheral boundary of the primary gross tumor. If the edge of the primary focus showed a spiculated margin, it was sketched close to the root of spiculation. All tumor micro-foci located in the intramural and mesorectum were sketched on the pathological section. The distance from the outer edge of the gross tumor was measured, which was defined as the shortest linear extensive distance from the outboard edge of the cancer nest to the edge of the gross tumor. The tumor micro-foci were classified according to their morphologic type and their positions in the tumor nodule (lymph node metastasis within the mesorectum was not included), tumor budding, endovascular foci, endolymphatic foci, and neural sheath foci [16]. The actual measured distance between the micro-focus below and above the gross tumor was defined as the nearest linear distance between the micro-focus and BM, which was marked as DM-I/DM-S. The actual distance of the lateral micro-focus was defined as the nearest linear distance between the micro-focus and the primary gross tumor, which was recorded as DM-L (Fig. 2). After correction of the respective retraction factors, the in vivo distances of tumor micro-foci were recorded as DM-Iin vivo, DM-Sin vivo, and DM-Lin vivo.

By measuring the DM-I, DM-S, and DM-L values and incorporating the R1 and R2 retraction factors in the respective directions, the distance of micro-foci in vivo was calculated as DM-Lin vivo = DM-L/(R1-L × R2-L), DM-Sin vivo = DM-S/(R1-S/I × R2-S), DM-Iin vivo = DM-I/(R1-S/I × R2-I).

Statistical analysis

SPSS 17.0 software was used for statistical analysis. The distances of lateral, inferior, and superior “in vivo” tumor micro-foci and 95% frequency value for 28 patients were calculated. Independent sample t tests and variance analysis were carried out. Numerical and categorical variables were compared using Chi-square and Student’s t tests. Differences were considered statistically significant for P values ≤ 0.05.