3D endosonography improves anal cancer staging

Accurate treatment planning for anal carcinomas helps to avoid recurrence. Three-D endosonography, compared with 2D, improves the detection of lymph nodes and of tumor spread, according to Danish researchers.

Dr. Anders F. Christensen and colleagues from Copenhagen examined 30 patients with a 10 MHz rotating endoprobe. Cross-sectional images (2D) of the anal sphincters were stored on a 3D system during retraction of the endoprobe through the anal canal. Afterwards, any projection (e.g., coronal and sagittal) could be reconstructed.

The 2D and 3D images were compared according to five parameters concerning tumor spread:
? invasion into the internal anal sphincter
? invasion into the external anal sphincter
? penetration beyond the external anal sphincter
? location of outer tumor limits
? detection of lymph nodes (the number was noted as well)

"Ideally, these five parameters should be identical in 2D and 3D endosonography," Christensen said.

The 3D method detected a median of five definitive findings, compared with a median of four findings by the 2D method, a significant result, Christensen said.

In eight patients, only 3D endosonography visualized the border of tumor spread. The median number of lymph nodes visualized in 3D was one, in 2D it was zero, another significant finding. In certain patients, 2D endosonography saw no lymph nodes, while 3D reconstruction picked them up.

He concluded that 3D allows accurate measurement of tumor size in any projection, while 2D allows assessment only in the scanning plane.

In another study, Dr. Christiane Kulinna from Vienna, Austria, and colleagues from Munich, Germany, found sagittal and coronal multiplanar reconstructions (MPR) of MDCT superior to standard axial reconstructions in local staging of rectal cancer. The reconstruction takes only an additional five minutes.

Fifty-five patients with histologic proven rectal cancer underwent MDCT of the abdomen. The raw data were acquired with a collimation of 1 mm, a pitch of 6, and an imaging reconstruction with 0.6 mm increment.

MPR in sagittal and coronal planes (5 mm slice thickness, 4 mm increment) were calculated from the raw data and compared to 5 mm axial reconstructions. Two blinded observers evaluated the local T- and N-staging first in axial, then in axial, sagittal, and coronal views within one week. Researchers did not differentiate T1 and T2 tumors.

Interobserver correlation was good, according to Kulinna.

For precise lymph node staging, 12 lymph nodes and five lymph nodes after radiochemotherapy had to be examined histopathologically.

Axial CT slices combined with MPR images showed accuracy rates of 76% for preoperative T-staging and 92% for detection of lymph node metastases. In up to 18% of the cases, MPR, especially sagittal slices, was more informative than axial images about the extent of the tumor invasion, Kulinna said.

In response to an audience member's questions, Kulinna said that based on results from another study, MR is better than CT at looking at the mesorectal fascia. A separate study determined there was no difference in results using 1, 3, and 5 mm slice thickness.