Rectal ultrasound with fine needle aspiration: an underutilized modality for delineating and diagnosing perirectal, presacral, and pelvic lesions

Rectal cancer is one of the most commonly diagnosed malignancies with an estimated 40,000 new cases a year in the United States with reported local recurrence rates ranging from 3 % to 9.2 % after treatment 1 2 3 4 5 . These recurrences, as well as other primary/malignant pathological lesions, can manifest as perirectal, presacral, and pelvic lesions. Rectal ultrasound (RUS), computed tomography (CT), and magnetic resonance imaging (MRI) have all been utilized to evaluate and stage these primary and recurrent rectal malignancies 6 7 8 9 10 . However, effective and safe tissue diagnosis in the perirectal, presacral, and pelvic lesions is crucial for effective management as these lesions can encompass a broad differential. The ability to perform RUS fine-needle aspiration (FNA) allows for pathological confirmation that can have considerable clinical impact on managing patients. This modality has been used in diagnosing perirectal, pelvic, and urologic lesions 11 12 13 14 15 . However, there is a paucity of data about the diagnostic yield and inherent risks of RUS-FNA compared to the conventional approaches of CT-guided transgluteal or surgical assessment of perirectal, presacral, and pelvic lesions. The purpose of this study was to evaluate the efficacy and safety of performing RUS-FNA for presacral, perirectal, and pelvic abnormalities.

This study was approved by the Institutional Review Board of Wake Forest Baptist Medical Center. Our retrospective case series used an institutional database to investigate patients who underwent RUS-FNA of perirectal, presacral, or pelvic lesions between August 2005 and September 2016. Twenty-seven patients met criteria during the specified study time period. Data including age, gender, prior imaging modalities utilized, pathology results, and outcomes were collected. All procedures were performed in an outpatient setting using moderate sedation by administering both midazolam and fentanyl or deep sedation with IV propofol by a licensed CRNA. All endoscopic procedures were performed using an Olympus UM130 or UM160 radial and linear echoendoscope with dopplers (Olympus America, Inc, Center Valley, Pennsylvania, United States). Furthermore, RUS-FNA was performed with a 22- or 25-gauge needle by experienced endosonographers at our tertiary referral center. All subjects received prophylactic ciprofloxacin 400 mg prior to FNA and 3 days following the procedure.

Our patient cohort consisted of 12 males (44.4 %) and 15 females (55.5 %) with an average patient age of 51 (range 19 – 80). Information for each case is summarized in Table 1 . Twelve patients (44.4 %) had known prior rectal or colon adenocarcinoma. One patient (3.7 %) had known endometriosis. All but one patient had prior imaging. A perirectal mass was detected at hysterectomy in the patient with no prior imaging. Imaging modalities included CT (22, 81.5 %), MRI (4, 14.8 %), and positron emission tomography (12, 44.4 %). On imaging, a presacral mass was present in 12 patients (44.4 %), a perirectal node was present in two patients (7.4 %), a perirectal abnormality was present in 11 patients (40.7 %), and a pelvic mass was present in one patient (3.7 %). The average size of lesion present on imaging was 3.58 cm (range 0.9 to 16.0 cm). Excluding the 16-cm lesion that was too large to be measured on RUS, the average size of lesion recorded on RUS was 3.12 cm (range 0.9 to 7.6). Eighteen lesions (66.7 %) were hypoechoic and nine lesions (33.3 %) were heterogeneous on RUS. U/S, ultrasound; FNA, fine-needle aspiration; CT, computed tomography; PET, positron emission tomography; I&D, incision and drainage; SUV, standard uptake value; MRI, magnetic resonance imaging; ARF, acute renal failure; GIST, gastrointestinal stromal tumor; RUS, rectal ultrasound FNA pathology distribution was as follows: adenocarcinoma (6, 22.2 %), squamous cell carcinoma (2,7.4 %), benign lymphoid hyperplasia (2,7.4 %), benign epithelial cells (3,11.1 %), benign atypical or nonspecific cells (2,7.4 %), benign reactive or inflammatory cells (2,7.4 %), myelolipoma (1,3.7 %), benign, cystic lesion (dermoid cyst, inclusion cyst, or teratoma) (4,14.8 %), non-diagnostic (2,7.4 %), sarcoma (1,3.7 %), seminal vesicle (1,3.7 %), urothelial bladder cancer (1,3.7 %). All adenocarcinomas were recurrent malignancies. RUS-FNA provided an effective diagnosis in 24 patients, giving a diagnostic yield of 88.8 %, and diagnosed recurrent adenocarcinoma in six patients. RUS-FNA was non-diagnostic in three cases (11 %). In one case, a specimen was initially labeled benign-appearing cells, however, subsequent surgical pathology reported a cystic hamartoma. In a second case, aspecimen was initially labeled benign atypical glandular cells, however, subsequent surgical pathology determined the specimen to be endometriosis. In the third case, a specimen was incorrectly read by pathology as a gastrointestinal stromal tumor (GIST) but subsequent surgical pathology revealed the lesion to be endometriosis. Five individuals (18.5 %) were lost to follow-up. We encountered four complications in two presacral and two perirectal mass FNAs ( Table 2 ). Our complication rate was approximately 25 % with biopsies of presacral masses and 8 % of the perirectal biopsies. No complications were observed with the pelvic mass RUS-FNA. The overall total complication rate was approximately 14.8 % in the form of abscess formation requiring either drainage or surgical intervention. Average needle passes performed in the four cases with complications was 2.5 passes. The location of abscess formation coincided with the original biopsy site. The echo characteristics of the four lesions were as follows: two heterogeneous and two hypoechoic. All four individuals had benign cytopathology on FNA. One out of the four individuals had an extended hospital course requiring two different incision and drainage (I&D) procedures and prolonged course of antibiotics because of an infected sacral teratoma. That individual subsequently improved after intervention but was lost to follow-up. The other individual with a presacral abscess had subsequent abscess excision with improvement of symptoms. The two individuals with perirectal abscesses both presented with fever and rectal pain. Both individuals underwent I&D with no reported complications. The two individuals’ symptoms improved after treatment. RUS, rectal ultrasound; FNA, fine-needle aspiration Does not include a 16-cm lesion that was too large to be measured on RUS.

RUS-FNA provides unparalleled ability to sample lesions surrounding the perirectal space including presacral and pelvic lesions. Previous studies have highlighted RUS-FNA’s role in diagnosing local pelvic urologic malignancies/masses, in confirming nodal metastases in early rectal cancer, in accurately diagnosing perirectal lesions (CRC and other lesions), and in preventing aggressive surgical interventions for benign conditions 12 13 14 15 16 . Our study is one of the larger descriptive cohort studies that highlights the clinical utility of RUS-FNA for assessing and accessing perirectal, presacral, and pelvic lesions. However, few have reported on diagnostic and safety data on RUS-FNA. Our study shows that RUS-FNA alters management in patients with perirectal, presacral, and pelvic lesions. Notably, our study had a diagnostic accuracy of 88.8 % which coincided with previous reports of FNA procedures in perirectal, intraluminal, and pelvic lesions 15 16 17 . Surgery was avoided in 55.5 % of our subjects and clinically impacted approximately 60 % of subjects, indicating the importance of RUS-FNA’s ability to obtain a tissue diagnosis and inform a decision about institution of medical and/or surgical therapy. These findings suggest that RUS-FNA is an accurate and useful clinical tool in management of patients with presacral, perirectal, and pelvic lesions. Although RUS-FNA is relatively safe, we found a significantly higher complication rate. Approximately 15 % of our patients developed an abscess. This higher complication rate is in stark contrast to the relatively uncommon reported complications with EUS-FNA from the upper gastrointestinal tract. Studies show that upper gastrointestinal FNAs appear to have fewer complication rates compared to lower gastrointestinal FNAs. The reported complication rate performing an endoscopic ultrasound (EUS)-FNA of the pancreas is 1 % to 2.5 % 18 19 . In comparison, a study evaluating adverse events (AEs) in lower gastrointestinal EUS-FNA reported AEs in 20.6 % of cases, mostly in the form of bleeding and pain, with 5.6 % of those events being serious 20 . Interestingly, few infectious complications have been reported in upper and lower gastrointestinal FNAs 19 20 . RUS-FNA has been proven a safe method for tissue sampling, with incidence of bacteremia similar to or less than that seen in diagnostic colonoscopy 21 . A lesion’s characteristics appear to contribute to risk of complications. An increased risk of febrile episodes or sepsis has been observed in upper gastrointestinal FNAs of cystic lesions 20 22 23 . It is unclear why our study showed such a high rate of infectious complications. Two out of the four lesions were heterogeneous and not purely solid, which may have increased the likelihood of infectious complications. Studies suggest that biopsy of presacral lesions does not add to the surgical strategy and that biopsies vary in accuracy 24 . However, the utility of tissue sampling has been increased with improved techniques and preoperative treatments. Presacral lesions such as Ewing sarcomas, osteosarcomas, lymphomas, and fibrous tumors are examples of lesions that could benefit from neoadjuvant therapy 25 . Patients’ medical treatment would be improved by preoperative biopsy of such lesions. Certainly, continued use of prophylactic antibiotics, minimization of needle passages, and use of experienced endosonographers can minimize complications in RUS-FNA. To our knowledge, there is little data in the literature outlining the diagnostic yield or complication rate of CT-guided biopsy via a transgluteal approach for perirectal, presacral, or pelvic lesions. Success rates with CT-guided prostate biopsies have been upward of 95 % to 97 % and a study performed in 2003 showed a 93 % diagnostic yield of pelvic lesions by an extraperitoneal approach 26 27 28 Despite this lack of data, CT-guided percutaneous biopsy has been described as being an appropriate method for biopsy of lesions located in perirectal, presacral, and posterior pelvic regions and superior in distant or metastatic disease 29 30 . However, there are disadvantages to transgluteal CT-guided FNA, including pain, patient discomfort due to lying in prone position for an extended period of time, and risk of gluteal vessel, sciatic nerve, and sacral plexus injury 29 . It can be argued that RUS-FNA may palliate many of the aforementioned disadvantages by minimizing pain, allowing patients to lie in the left lateral decubitus position, and providing the proceduralist with closer anatomic proximity to lesions to accurately obtain a tissue diagnosis and improve staging of primary/recurrent malignancies. Our study is limited by a small sample size of 27 patients and the retrospective design. Admittedly, there are concerns about later complications possibly being missed as patients could have gone to their local community hospital or physician rather than returning to our facility. All biopsies could not be corroborated with surgical specimen pathology because results of FNA biopsies dictated medical decision-making and prevented some patients from having a surgical intervention. Also, our study was not a comparative study to differentiate the diagnostic yield between CT-guided biopsy versus RUS-FNA. Ideally, these two imaging modalities should have much larger studies for comparison in diagnostic yield and complication rates.

In summary, RUS-FNA is an accurate and relatively safe method for obtaining tissue diagnosis of presacral, perirectal, and pelvic lesions when performed by experienced endosonographers. While the diagnostic yield of RUS-FNA is high and the potential to affect clinical decision-making is real, the risk of complication is not negligible. RUS-FNA should only be performed if the result will substantially alter clinical management, and the decision to perform RUS-FNA should be made by a multidisciplinary team.