Radiologic evaluation of gastrointestinal leiomyosarcomatosis

Peritoneal leiomyosarcomatosis originating from gastrointestinal leiomyosarcomas: CT features

PURPOSE

To evaluate the computed tomographic (CT) features of peritoneal leiomyosarcomatosis (PL) originating from gastrointestinal leiomyosarcoma and compare the CT features of primary gastrointestinal leiomyosarcomas between patients with and those without PL.

MATERIALS AND METHODS

The authors reviewed the medical and surgical-pathology records and the CT scans of 89 patients with gastrointestinal leiomyosarcoma to determine the prevalence of PL. Also, the CT scans of the patients with PL were evaluated for the morphologic appearance of peritoneal lesions. The CT features of the primary tumors were compared between the patients with and those without PL. Student t and Fisher exact tests were performed.

RESULTS

Metastases were present in 39 patients. PL was seen in 25 patients. Other metastatic sites included liver in 26 patients, lymph nodes in 11, lungs in three, and bone in one. In the 25 patients with PL, CT findings included multiple discrete peritoneal nodules in 22 patients and massive peritoneal lesions in three. Smudged omental infiltration was combined with other findings in 10 patients. The sites of the peritoneal masses were mesentery in 18 patients, omentum in 11, paracolic gutters in 11, pelvic cavity in eight, perihepatic space in three, and perisplenic space in one. Four patients had ascites. There were statistically significant differences in size of primary tumor (mean diameter, 12.5 cm vs 9.2 cm) (P =.01) and prevalence of exophytic growth of primary tumor (P =.02) between the patients with and those without PL.

CONCLUSION

CT findings can indicate PL in the differential diagnosis. The prevalence of PL appears to be higher when it originates from large primary tumors.

CT-guided aspiration core needle biopsy of gastrointestinal wall lesions

PURPOSE

The purpose of this work was to examine the accuracy and safety of CT-guided core biopsy of gastrointestinal wall lesions.

METHOD

We reviewed over 1,200 CT-guided abdominal core biopsy results from 1989 through 1998. Forty-five were from gastrointestinal wall lesions (16 gastric, 3 duodenal, 7 small intestine, and 19 colon).

RESULTS

A definitive histologic diagnosis was obtained from the core biopsy in 41 patients (91%) without complication. The gastric lesion diagnoses were mesenchymal tumor of smooth muscle origin (eight), lymphoma (one), adenocarcinoma (three), and normal (one). Duodenal core biopsy diagnoses were inflammation (one) and normal (one). Small intestinal core biopsy diagnoses were non-Hodgkin lymphoma (four), metastatic leiomyosarcoma (one), carcinoma (one), and tuberculosis (one). Colon core biopsy diagnoses were carcinoma (11), lymphoma (2), actinomycosis (2), granulomatous inflammation (1), metastatic squamous cell carcinoma (1), chronic abscess (1), and mesenchymal tumor of smooth muscle origin (1).

CONCLUSION

Percutaneous CT-guided core needle biopsy from gastrointestinal wall lesions is safe and accurate, especially in submucosal lesions, and should be considered in selected cases with negative endoscopic biopsy or in which endoscopic biopsy is not possible.