Endoscopic Ultrasonography Assessment for Ampullary and Bile Duct Malignancy

From 1989 through 1992, endoscopic ultrasonography (EUS) was undertaken preoperatively to evaluate the extent of primary tumor, involvement of regional lymph nodes, and distant metastases in 22 patients with ampullary carcinoma and 18 patients with bile duct carcinoma. The results were compared with histopathological findings according to the TNM staging system. The accurate rate in assessing the extent of cancer invasion was 82% for ampullary carcinoma, 66% for common hepatic duct carcinoma, and 78% for common bile duct carcinoma. The accuracy of EUS in predicting regional lymph node metastasis was 59% for ampullary carcinoma, 56% for common hepatic duct carcinoma, and 67% for common bile duct carcinoma. Invasion of the portal vein was correctly predicted by EUS in 2 of 3 patients. None of the 3 patients with liver metastasis was detected by EUS. Therefore, endoscopic ultrasonography is an effective method in the evaluation of the extent of cancer invasion of ampullary and bile duct carcinoma as well as the involvement of regional lymph nodes preoperatively. However, due to its limited penetration depth, EUS is inadequate in the assessment of liver metastasis.


sent time.Alt
ough the 5-year survival rate after a Whipple operation for ampullary and distal end bile duct carcinoma is about 50%, there is also a 5 to 10% mortality after this operation.A correct preoperative evaluation of the extent of the primary tumor, involvement of regional lymph nodes, and distant metastases will contribute to choos- ing an appropriate therapeutic method and determin- ing the prognosis.

Since 1989, we have used endoscopic ultrasonography (EUS) assessment for ampullary and bile duct ca - cinoma preoperatively, and compared the results with the surgical explorations and pathological findings for resected specimens for evaluating the accuracy of EUS according to the TNM staging system 1 ].


PATIENTS AND METHODS


Subjects

From 1989 through the end of 1992, 22 patients with ampullary ca

en aged 1
to 72 years, mean 50.4 years) and 18 patients with bile duct carcinoma (10 men and 8 women aged 36 to 72 years, mean 61.2 years) were examined by EUS pre- operatively.All 40 patients underwent surgical exploration, and all the lesions were resected.The results of EUS assessment were compared with the histological findings of the resected specimens.In addition to the 40 patients with resectable tumor, there were other 6 patients with nonresectable tumor.Among these patients, invasion of portal vein was found in 3 patients, with 2 of 3 patients correctly predicted by EUS.Liver metastasis was found in 3 patients, and none of these were detected by EUS.


Instrument

All studies were performed with an echoendoscope (Olympus GF-UM2 and GF-UM3).A water-fil

d balloon o
a combination with the water-filling method was used for getting a better acoustic coupling between the transducer and the mucosa.


FIGURE

EUS image shows a hypoechoic tumor (T) destroying the whole layer structures of the papilla of Vater and invading the pancreatic head (PH) with dilatation of the common bile duct (CBD).

Bile duct carcinoma was imaged as a hypoechoic or h perechoic inhomogeneous lesion with or without invading the surrounding structures (Fig. 2).

Lymph nodes with a hypoechoic pattern and clearly delineated margin were generally considered as malig-

nt.Lymph nodes with
hyperechoic pattern and an uncleared margin were considered as benign.


Manipulative Method

The echoe doscope was inserted into the second por- tion of the duodenum closing to the ampulla of Vater.

After identification of the common bile duct and the pancreatic head using the portal vein as a landmark, the echoendoscope was gradually withdrawn.Careful scan- ning and ass

sment were done for the extent of
tumor, enlarged regional lymph nodes, and distant metastasis.


Interpretation of Endosonography

Ampullary cancer was diagnosed when a hypoechoic mass was imaged in the region of ampulla of Vater with destruction of the normal structure of

pulla, w
th or without infiltration into the pancreas or other surrounding structures (Fig. 1).


RESULTS

Among the 40 patients in whom malignant lesions were resected completely and the results of EUS assessment were compared with the histological find- ings for resected specimens, the accuracy rate for assessing the extent of cancer invasion was 82% for ampullary carcinoma (Table I), 66 for common hepatic duct carcinoma (Table II), and 78% for com- mon bile duct carcinoma (Table III).

There were other 6 patients with nonresectable tumor due to invasion of portal vein or liver metasta- sis.Invasion of the portal vein was correctly redicted by EUS in 2 of 3 patients.None of the 3 patients with liver metastasis was detected by EUS.

The accuracy of EUS in predicting the regional lymph node metastasis was 59% for ampullary carci- FIGURE 2 EUS i

ge shows a
ypoechoic tumor (lower T) at the end of dilated common bile duct with strongly suspected infiltration of portal vein (PV) as well as the middle portion of common bile duct (upper T).   noma (Table IV), 56% for common hepatic duct car- cinoma (Table V), and 67% for common bile duct carcinoma (Table VI).


DISCUSSION

The Whipple operation is the standard method for curing ampullary and distal end bile duct carcinoma.

The 5-year survival rate after this operation for ampullary carcinoma is about 50% [2], and it also carries a 5% mortality even in expert hands [3,4].So it is very important to evaluate the extent of primary tumor as well as to rule out major vascular involve- ment and distant metastases before surgery, that is, to select a patient with a resectable tumor for this operation.For elderly patients, if the extent of ampullary carcinoma is still within the duodenal wall, some- times a local resection is much better and safer.If invasion of major vascular structures or distant metastases are detected before surgery, which means the carcinoma is unresectable, then an alternative nonsurgical palliative procedure, such as endoscopic biliary stenting, will be available for them, because it carries a very low morbidity and mortality and can drain the bile juice well.EUS has been under development for more than a decade.It now plays a well-established role in the staging for gastrointestinal tract malignancies preoperatively.It has an accuracy of 83 to 88% and 82 to 86% in determining the extent of ampullary carcinoma and bile duct carcinoma, respectively [5][6][7][8].The accu- racy rate in our group was 82% in assessing the extent of ampullary carcinoma, and 66% in common hepatic duc carcinoma, and 78% in common bile duct carci- noma.Understaging of ampullary and bile duct carci- noma might be due to microscopic infiltration or incorrect assessment of the infiltration.Overstaging of bile duct carcinoma by EUS might be due to compres- sion of the blood vessels simulating infiltrati n by tumor mass.

The overall accuracy of EUS in predicting lymph node metastases was 54 to 92% and 53 to 65% in ampullarff carcinoma and bile duct carcinoma, respec- tively [5][6][7][8].The accuracy in our group was 59% in ampullary carcinoma, 56% in common hepatic duct carcinoma, and 67% in common bile duct carcinoma.

The accuracy of EUS in assessing lymph node metas-   tases for both ampullary carcinoma and bile duct car- cinoma is not very high.In general, lymph nodes with a hypoechoic pattern and clearly delineated margin were considered as malignant; lymph nodes with a hyperechoic pattern and indistinct margin were con- sidered as benign.However, sometimes it is difficult to distinguish a malignant lymph node metastasis from a nonmetastatic lymph node abnormality.Criteria for defining metastatic as well as nonmetastatic lymph nodes should be improved for increasing the accuracy of EUS.Invasion of the portal vein in ampullary and bile duct carcinoma was correctly detected by EUS in 2 of 3 patients.But none of the 3 patients with liver metas- tases was detected by EUS due to its limited penetra- tion depth.


CONCLUSION

EUS is accurate for assessing the extent of primary tumors of ampullary and bile duct carcinoma.It is sat- isfactory in p