Introduction
Giant tumors located in the rectal and perirectal regions cause compression and distortion in normal anatomic structures, so it may be difficult to determine the origin of the tumor and decide the correct diagnosis and treatment (1-6). Masses such as rectal adenocarcinoma, sarcoma, neuroendocrine tumor, leiomyoma, ovarian mass, rectal gastrointestinal stromal tumor (GIST), prostate adenocarcinoma, lymphoma, neurogenic tumor and congenital cyst should be considered in the differential diagnosis of giant rectal and perirectal tumors (6). Although rectoscopy-guided biopsy is the most important method for diagnosis, it is insufficient to evaluate intramural and extramural extension of the mass (1,5). Pre-operative imaging is necessary to determine the location of the mass (submucosal/intramural/extramural) and its extension, to evaluate the need for surgical approach to the patient and to decide the surgical technique to be applied (1-7). Computed tomography (CT) and magnetic resonance imaging (MRI) are used in the diagnosis and evaluation of rectal and perirectal tumors (1-7). MRI is more appropriate because it does not contain radiation and has high soft tissue resolution (1-5). Diffusion-weighted imaging (DWI) may also provide important findings in use with MRI (8-10).
To the best of our knowledge, there is no detailed study in the literature regarding MRI and DWI characteristics that may help in the differential diagnosis of giant rectal and perirectal tumors. Therefore, we aimed to identify some specific qualitative and quantitative MRI and DWI findings that may be important in the differential diagnosis of giant rectal and perirectal masses.
Methods
Written consent was obtained from the patients and approval was obtained from the ethics committee in accordance with the Helsinki Declaration Criteria before the initiation of study (University of Health Sciences, Ümraniye Training and Research Hospital Ethics Committee of Clinical Studies, decision no: 235, date: 23/01/2019). Lower abdominal MRI reports performed in a single center between January 2014 and January 2019 were reviewed for rectal and perirectal masses using hospital information system. MRI examinations of patients with ≥5 cm rectal and perirectal masses were detected retrospectively. Patients without histopathological diagnosis (n=15) and patients with incomplete or inadequate MRI examination for evaluation (n=7) were excluded. Eighty-one patients (36 female, 44% and 45 male, 56%) with ≥5 cm mass in the rectal and perirectal region and pre-operative MRI and DWI were included in our study. The mean age was 54±15 years.
Contrast-enhanced lower abdominal MRI examinations were performed using a 1.5 Tesla MRI (Magnetom Avanto®, Siemens Healthineers, Erlangen, Germany). The contrast-enhanced lower abdominal MRI parameters were: sagittal T2-weighted turbo spin-echo (TSE) BLADE (TR/TE: 5000/112 ms, FOV: 350, matrix: 256×256, section thickness, 7 mm), coronal fat-suppressed T2-weighted TSE BLADE (TR/TE: 4000/84 ms, FOV: 320, matrix, 256×320, section thickness, 5 mm), axial T2- weighted fast spin-echo sequence (TR/TE: 3700/88 ms, FOV: 320, matrix: 182×320, section thickness, 5 mm), pre-contrast and post-contrast axial T1-weighted 3D VIBE ( TR/TE: 4.5/2.1 ms, FOV: 380, matrix: 195×320, deflection angle, 15 °C section thickness, 5 mm), diffusion weighted single shot echo planar imaging with b values 0 and 1000 s/mm2 (FOV: 430, TR/TE: 5200/58, matrix: 115×192, section thickness: 5 mm, NEX: 3).
Pre-operative MRI and DWI examinations were re-evaluated by a radiologist (F.K.) experienced in abdominal imaging who was blinded to the histopathological information of the patients. Longest diameter (mm) of the masses, contour characteristics (regular/irregular), signal intensity characteristics (hypo-, iso-, hyper-intense) in T1- and T2-weighted sequences, contrast patterns (mild/intense, homogeneous/heterogeneous), tumor content (hemorrhagic signal, cystic component), invasion, lymphadenopathy and metastasis were noted. In addition, the apparent diffusion coefficient (ADC) values and ratios of the masses were calculated. For ADC calculation, circular region of interest (ROI) measurements ranging from approximately 100 to 200 mm2 were used. It was ensured that no lumens and extra-mass structures entered into the ROI. ROI measurements were performed in three different locations in the mass and gluteus maximus muscle in the same section. The ratio of the ADC value of the mass to the ADC value of the muscle (ADC ratio) was calculated for the standardization with the obtained mean ROI measurements.
Statistical Analysis
All patients had histopathological results. The relationship between qualitative (signal intensities, contrast enhancement pattern, presence of lymphadenopathy and metastasis) and quantitative imaging findings and histopathological results were statistically analyzed. The suitability of the parameters to normal distribution was evaluated by Shapiro-Wilk test. The Mann-Whitney U test was used for the comparison of the quantitative data and the chi-square test was used for the comparison of the qualitative data. Significance was evaluated at p<0.05.
Results
In our study, 81 giant rectal and perirectal masses, 60 of which were malignant and 21 were benign, were detected. The mean longest diameter was 86±32 mm (range: 50-208 mm) in malignant masses and 66±15 mm (range: 50-110 mm) in benign masses. Malignant masses (n=60) included 48 rectal adenocarcinomas, two prostate adenocarcinomas, three leiomyosarcomas, three ovarian cancers, and four rectal GISTs. Benign mass group (n=21) consisted of two ovarian fibromas, five mature teratomas, five leiomyomas, three perirectal abscesses, two ovarian cystadenomas and four endometriomas. Most of the malignant masses (47/60, 78%) had irregular contours. All benign masses (20/21, 95%) except one perirectal abscess had regular contours. All rectal GISTs (n=4) had regular peripheral contours and mild heterogeneous contrast enhancement beginning in the arterial phase (Figure 1). Unlike other malignant tumors (n=56), heterogeneous, iso-hyperintensity was observed in central sections in T2-weighted sequences of all rectal GISTs (n=4), whereas hypointense signal characteristics were detected in peripheral sections (Figure 2 a,b). Peripheral marked diffusion restriction was present in rectal GISTs (Figure 2 c,d). Most rectal adenocarcinomas (41/48, 85%) had intense enhancement. Lymphadenopathy (43/60, 71%), adjacent organ invasion (28/60, 46%) and distant metastasis (16/60, 26%) were observed only in malignant masses. Significant statistical difference was found between the benign and malignant groups in terms of signal intensities in T1- and T2-weighted sequences (p<0.05). Table 1 shows the distribution of benign and malignant masses according to qualitative MRI characteristics.
The mean ADC values and ratios were significantly lower in malignant masses than benign masses (p<0.05). Table 2 summarizes the quantitative DWI findings of benign and malignant masses. In giant rectal adenocarcinomas (n=48), the mean ADC value was 0.87±0.16×10-3 mm2/s and the ADC ratio was 0.57±0.10 (Figure 3 a,b,c). The mean ADC value of mucinous rectal adenocarcinomas (4/48) was 0.98±0.25×10-3 mm2/s. In non-mucinous rectal adenocarcinomas (44/48), the mean ADC value was found to be low with 0.86±0.15×10-3 mm2/s and was close to the ADC value of mucinous rectal adenocarcinomas. Rectal GISTs, leiomyosarcomas, prostate cancers, and ovarian cancers yielded low ADC values (0.83±0.07×10-3 mm2/s, 0.60±0.08×10-3 mm2/s, 0, 52±0.16×10-3 mm2/s, and 0.77±0.15×10-3 mm2/s, respectively) and ADC ratios were obtained in (0.53±0.06, 0.45±0.05, 0.41±0.09, and 0.54±0.05, respectively) (Figure 4 a,b,c).
Discussion
Differential diagnosis of giant rectal and perirectal tumors can be difficult in pre-operative imaging and sometimes even in histopathological evaluation. Therefore, taking into account some specific findings, evaluating pre-operative MRI and DWI findings in detail may help in making the correct diagnosis and deciding the appropriate treatment method (2,4,11-13). DWI is a an imaging modality showing the movement of water in the tissue (8-10). The ADC map is generated from DWI and allows us to detect and measure diffusion restriction. Increased cellularity leads to diffusion restriction (8-10). Diffusion restriction is hyperintense on DWI and hypointense on ADC map (8-10). In addition, ADC value provides us the quantification of diffusion and can be helpful in the differential diagnosis of benign and malignant tumors (8-10).
There are many studies in the literature regarding the efficacy of MRI and DWI in the diagnosis and post-treatment evaluation of rectal cancer. In a study by Li et al. (11), when compared with conventional MRI (71.42%), combined use of MRI (92.85%) and DWI (b value: 1000 s/mm2) in patients with rectal adenocarcinoma (n= 84) was observed to significantly increase diagnostic accuracy. In another study, Gu et al. (14) evaluated DWI (b value: 1000 s/mm2) and positron emission tomography (PET) CT findings in 33 patients with rectal adenocarcinoma. A significant negative correlation was found between ADC that shows increased tumor cell number and standard uptake value that shows metabolic activity, and a positive correlation was detected between total diffusion index and total lesion glycolysis (14). Çolakoğlu and Erden (15) evaluated the findings of DWI (b value: 1000 s/mm2) in mucinous (n=18) and non-mucinous (n=44) rectal adenocarcinomas. Diffusion properties and ADC values of mucinous and non-mucinous adenocarcinomas were found to be different (15). In mucinous adenocarcinomas (1.631±0.375×10-3 mm2/s), higher ADC values were obtained compared to non-mucinous adenocarcinomas (0.921±0.157×10-3 mm2/s), and a threshold value of 1.27×10-3 mm2/s was calculated with high sensitivity (94.4%) and specificity (94.4%) (15). In a similar study, Nasu et al. (16) investigated the DWI (b value: 1500 s/mm2) and ADC characteristics of mucinous (n=15) and tubular (n=66) rectal adenocarcinomas. A higher ADC value was detected in mucinous adenocarcinomas (1.49±0.34×10-3 mm2/s) than tubular adenocarcinomas (0.80±0.15×10-3 mm2/s) (16). In our study, the number of mucinous rectal adenocarcinomas (n=4) was low and the mean ADC value was close to non-mucinous rectal adenocarcinomas (n=44). In another study, Bassaneze et al. (17) calculated the mean ADC value before chemotherapy as 1.01±0.05×10−3 mm2/s in locally advanced rectal adenocarcinomas (n=33). In our study, low ADC values (0.87±0.16×10-3 mm2/s) were obtained in giant rectal adenocarcinomas (n=48). In addition, the ADC ratio was calculated in our study and not only rectal adenocarcinomas, but also other rectal and perirectal masses ≥5 cm in size were evaluated.
The degree of DWI is associated with the b value (9,10). At high b, a strong diffusion effect and diffusion restriction are achieved, and the noise signal ratio is reduced (9,10). Therefore, using the appropriate b value is important for image quality and accurate evaluation (9,10). Chen et al. (18) used 10 different b values ranging from 0 to 2000 s/mm2 in the evaluation of rectal cancers by DWI. The most suitable b value combination was found to be 0 and 1000 s/mm2 (18). In our study, b values of 0 and 1000 s/mm2 were used.
In another study, the role of MRI in the differential diagnosis of postoperative rectal cancer recurrence (n=17) and benign lesion (n=8) was investigated (19). It was reported that malignant and benign lesions could be distinguished with high sensitivity (100%), specificity (91.7%) and accuracy (96.6%) rates with a threshold ADC value of 1.21×10-3 mm2/s (19). In the dynamic contrast test, the time-intensity curve has also been reported to have a high accuracy rate (92%) in distinguishing between malignant and benign masses (19). No significant difference was found between the malignant (2.84±1.52) and benign (2.58±0.80) groups in terms of signal intensity changes in T2-weighted sequences (p>0.05) (19). In contrast, in our study, a significant difference was found between the benign and malignant groups in terms of T1- and T2- weighted signal intensities.
In 14 patients with <5 cm four rectal GISTs and >5 cm 10 rectal GISTs, Jiang et al. (20) observed that six of the GISTs were irregularly contoured and 11 showed heterogeneous enhancement on MRI and CT (20). Kim et al. (4) defined the MRI characteristics of rectal GISTs as eccentric mural masses with regular contour, calcification, necrosis or ulceration, that has hypointensity in T1-weighted sequences, iso-hyperintense signaling in T2-weighted sequences and intensely heterogeneous contrast enhancement. In our study, which included fewer rectal GISTs, all rectal GISTs had contours and signal characteristics were similar. However, the enhancement characteristics were heterogeneous but milder. It was thought that GISTs ≥5 cm in our study may be the cause of peripheral heterogeneous mild enhancement. In addition, the hypointense appearance of peripheral sections of rectal GISTs and heterogeneous hyperintense appearance of central sections of T2-weighted sequences was one of the striking findings of our study. However, in our study, low ADC values and ratios were also obtained in rectal GISTs.
There are some limitations of our study. One of them is the retrospective nature of the study with a small number of benign masses. The other is that qualitative and quantitative MRI findings were evaluated by a single radiologist.
Conclusion
The evaluation of specific MRI characteristics and ADC ratio measurements have an important role in the differential diagnosis of giant rectal and perirectal masses, and guide the planning of patient treatment.
Ethics Committee Approval: University of Health Sciences, Ümraniye Training and Research Hospital Ethics Committee of Clinical Studies, (decision no: 235, date: 23/01/2019).
Informed Consent: Written consent was obtained from the patients.
Peer-review: Externally peer-reviewed.
Author Contributions: Surgical and Medical Practices - F.K., Ö.D.; Concept - F.K., Ö.D.; Design - F.K., Ö.D.; Data Collection and/ or Processing - F.K., Ö.D.; Analysis and/or Interpretation - F.K., Ö.D.; Literature Search - F.K., Ö.D.; Writing Manuscript - F.K., Ö.D.
Conflict of Interest: No conflict of interest was declared by the authors.
Financial Disclosure: The authors declared that this study received no financial support.