331jum_online_Layout 1 12/19/13 1:55 PM Page 47
ORIGINAL RESEARCH
Routine Intraoperative Hepatic Sonography Does Not Affect Staging or Postsurgical Hepatic Recurrence in Pancreatic Adenocarcinoma Leonora W. Mui, MD, Lisa J. Pursell, BA, Isadora C. Botwinick, MD, John D. Allendorf, MD, John A. Chabot, MD, Jeffrey H. Newhouse, MD Objectives—The purpose of this study was to evaluate the utility of intraoperative sonography of the liver in the staging of pancreatic adenocarcinoma and its impact on the rate of postoperative tumor recurrence in the liver. Methods—We performed a retrospective analysis of the rate in which intraoperative sonography of the liver changed surgical management in 470 surgical candidates with pancreatic adenocarcinoma. In postsurgical patients, we performed a χ2 analysis to examine whether the patients who underwent hepatic intraoperative sonography had a lower rate of recurrent disease in the liver within the first 6 months of surgery compared to patients who did not undergo the procedure. Received March 10, 2013, from the Department of Radiology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore University Medical Center, Manhasset, New York USA (L.W.M.); Departments of Hepatobiliary Surgery (L.J.P., I.C.B., J.A.C.) and Radiology (J.H.N.), Columbia University Medical Center, New York, New York USA; and Department of Surgery, Winthrop University Hospital, Mineola, New York USA (J.D.A.). Revision requested April 24, 2013. Revised manuscript accepted for publication May 20, 2013. We thank members of the Pancreas Center at Columbia University Medical Center for their datamining efforts. This material was presented at the 96th Scientific Assembly and Annual Meeting of the Radiological Society of North America; November 28–December 3, 2010; Chicago, Illinois. Address correspondence to Leonora W. Mui, MD, Department of Radiology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore University Medical Center, 300 Community Dr, Manhasset, NY 11030 USA. E-mail: [email protected]
Abbreviations
CT, computed tomography; MRI, magnetic resonance imaging doi:10.7863/ultra.33.1.47
Results—Hepatic intraoperative sonography affected management in less than 1% of cases, detecting 1 unsuspected liver metastasis in 470 surgical patients with pancreatic adenocarcinoma. Of 3 patients with equivocal liver lesions identified on preoperative computed tomography or magnetic resonance imaging, hepatic intraoperative sonography excluded metastasis and cleared all the patients for surgical resection. There was no significant difference in postoperative liver recurrence between the group of patients who received intraoperative sonography before resection and patients who did not have the procedure done (P > .99). Conclusions—Routine intraoperative sonography of the liver does not affect staging of pancreatic adenocarcinoma. It may be useful for evaluating equivocal lesions. Key Words—gastrointestinal ultrasound; hepatic metastases; intraoperative sonography; liver; pancreatic adenocarcinoma
P
ancreatic cancer is the 10th most common malignancy and 4th highest cancer killer in adults, with a 5-year overall cumulative survival rate of less than 5%. Surgical resection provides the best curative chance for patients with this malignancy. However, at the time of presentation, about 55% of patients have distant metastases and would not benefit from resection. Preoperative imaging with computed tomography (CT), magnetic resonance imaging (MRI), and endoscopic sonography identifies most surgical candidates. Occasionally, surgical candidates are found during exploratory laparotomy or laparoscopy to have distant metastases, resulting in termination of surgery.
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:47–51 | 0278-4297 | www.aium.org
331jum_online_Layout 1 12/19/13 1:55 PM Page 48
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
Intraoperative sonography of the pancreas and liver is routinely performed during exploratory laparotomy to stage pancreatic cancer (Figure 1). This technique offers excellent spatial and contrast resolution and real-time imaging by applying the ultrasound probe to the organ of interest. Surgical sonography helps delineate the relationship of lesions with vascular and biliary structures and identify small hepatic neoplasms and tumor thrombi.3 Based on studies showing the utility of intraoperative sonography in evaluating liver metastasis in colorectal and hepatocellular carcinoma, intraoperative sonography of the liver is commonly performed in patients with pancreatic adenocarcinoma. The purpose of this study was to evaluate the impact of intraoperative sonography of the liver in the staging and postoperative rate of recurrent liver disease in pancreatic adenocarcinoma.
Materials and Methods Our Institutional Review Board approved this retrospective study with waiver of informed consent. We searched the electronic clinical data warehouse of the Pancreas Center of a major medical center to identify 470 surgical candidates with pancreatic adenocarcinoma between 1995 and February 2011 who underwent intraoperative sonography during exploratory laparotomy or laparoscopy. These patients were deemed resectable on the basis of preoperative CT, MRI, or endoscopic sonography that showed no convincing evidence of metastases. Before surgery, the noninvasive imaging was reviewed at an interdisciplinary conference consisting of surgeons, radiologists, gastroenterologists, and interventionalists to discuss management. During exploratory laparotomy, the surgeons performed visual and manual inspections of the peritoneal cavity, liver, stomach, small bowel, omentum, celiac axis, and head of the pancreas. Figure 1. Pancreatic cancer pathway. EUS indicates endoscopic sonography; and IOUS, intraoperative sonography.
48
Most of the patients (97% [456 of 470]), had open intraoperative sonography, whereas 2% (11 of 270) underwent laparoscopic intraoperative sonography, and 1% (3 of 470) were converted from laparoscopic to open intraoperative sonography. Intraoperative sonography was performed by surgeons who were credentialed by the American College of Surgeons based on both basic and advanced abdominal ultrasound courses involving open and laparoscopic sonography on live pigs. By the study end date, the surgeons had 15, 10, 9, and 6 years of experience with intraoperative sonography. The surgeons used highresolution real-time B-mode intraoperative sonography of the liver with a T-shaped curvilinear array transducer with frequency of 5 MHz using sterile technique. Each of the 8 segments of the liver was evaluated systematically. The right hepatic lobe was interrogated in the transverse direction by identifying the right hepatic vein and scanning posteroanteriorly from the diaphragm to the inferior edge to cover segments 7, 6, 8, and 5. Then the probe was turned longitudinally to the inferior vena cava, and images were scanned right to left to scan segment 4 between the middle hepatic vein and falciform ligament. This scan was repeated at 5-cm intervals from the diaphragm down through the inferior edge of segment 4. The lateral aspect of the left lobe was scanned by getting the heart in view and sweeping through the liver to the left of the falciform ligament. Finally, the probe was placed under the left lateral segment of the venous venosum to evaluate the caudate lobe. Indeterminate liver lesions were biopsied and sent for intraoperative frozen-section analysis. For patients found unresectable at surgery, the medical center’s electronic records were reviewed for radiologic reports, surgical notes, and pathologic reports. We examined the rate at which intraoperative sonography of the liver changed surgical management by detecting liver metastases missed by preoperative imaging and surgical exploration. We reviewed the electronic records of 118 patients who underwent clinical evaluation between 2009 and 2011 to identify 3 cases with equivocal liver lesions on preoperative CT or MRI. In these patients, we specifically investigated the impact of hepatic intraoperative sonography in their surgical management. To examine the impact of hepatic intraoperative sonography on postoperative liver recurrence, we compared the rate of recurrent disease in the liver only between patients who underwent intraoperative sonography at surgery and those who did not undergo the procedure. We investigated the database for patients who received followup care at our institution for at least 6 months after resection between 1995 and 2010 and identified 232 patients
J Ultrasound Med 2014; 33:47–51
331jum_online_Layout 1 12/19/13 1:55 PM Page 49
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
who underwent intraoperative sonography at resection and 58 patients who did not undergo the procedure at surgery. We assumed that any patient with negative preoperative imaging findings, in whom surgical and pathologic findings indicated complete resection of the primary tumor and who developed hepatic metastases after surgery with no evidence of tumor in any other site at the time the hepatic metastases appeared, probably had hepatic metastases too small to find at the time of surgery. Presumably, these liver lesions were present at surgery but too small to detect by inspection or intraoperative sonography. Followup imaging included CT, MRI, or positron emission tomography/CT. A χ2 analysis was performed to test statistical significance.
Results Seventeen percent of the surgical candidates (80 of 470) were unresectable, including 4% with distant metastases (19 of 470). Three percent of the operations (14 of 470) were terminated because of liver metastases, which were pathologically proven to be adenocarcinoma. The rate at which intraoperative sonography of the liver detected occult lesions was less than 1%; there was 1 case among the 470 sampled in which intraoperative sonography of the liver detected an unsuspected liver metastasis missed by preoperative CT and surgical exploration (Figure 2). This patient had no evidence of metastatic disease on preoperative imaging and no visible or palpable lesions during surgical exploration. Pathologic examination confirmed adenocarcinoma, and surgery was terminated. In another case, intraoperative sonography revealed additional liver lesions in a patient with diffuse biliary ductal dilatation without discrete lesions on MRI; however, liver metastases visualized and palpated during surgical exploration already disqualified surgery as a treatment option, and additional liver lesions seen on intraoperative sonography did not affect management. Sampling of 118 surgical candidates between 2009 and February 2011 found 3 cases in which preoperative imaging questioned possible liver lesions. At surgical exploration with intraoperative sonography, these lesions were found to be negative for metastasis on the basis of sonographic features. There was another case in which routine hepatic intraoperative sonography revealed an indeterminate lesion, prompting tissue sampling with negative results for metastasis. The 6-month postoperative rate of recurrent disease in the liver was identical in the patients who underwent intraoperative sonography at surgery (7% 17 of 232) and
J Ultrasound Med 2014; 33:47–51
the patients who did not undergo the procedure (4 of 58), with χ2 P > .99 (Figure 3).
Discussion The utility of intraoperative sonography in evaluating patients with hepatobiliary disease has been extensively studied.4–7 Machi et al4–6 found that intraoperative sonography provided beneficial information affecting the staging and management of 89% of hepatic operations, 83% of noncalculous biliary operations, and 73% of pancreatic operations. The overall accuracy of intraoperative sonography in detecting liver metastasis from colorectal carcinoma was 94% in one series, higher than the accuracy of CT (75%) or MRI (80%).4 Zacherl et al8 concluded that intraoperative sonography is crucial in the staging of patients with hepatocellular carcinoma and colorectal metastases since it changed the surgical strategy in about 23% of cases. A prospective study by Guimarães et al9 in 60 patients with a mix of abdominal tumors, including colFigure 2. Impact of hepatic intraoperative sonography (IOUS) on staging patients with pancreatic adenocarcinoma.
49
331jum_online_Layout 1 12/19/13 1:55 PM Page 50
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
orectal cancer, bile duct tumors, hepatic tumors, and 4% of cases with pancreatic tumors, found that intraoperative sonography offered 91% sensitivity, 78% specificity, a 91% positive predictive value, and a 78% negative predictive value for detection of liver metastasis, with results changing the surgical strategy in 41% of cases. A recent interdisciplinary review of intraoperative sonography in pancreatic surgery by Ní Mhuircheartaigh et al10 found the technique highly useful in guiding Puestow and pseudocyst drainage procedures and localizing foci of neuroendocrine tumors and moderately useful for staging resection of nonadenomatous pancreatic masses but of the lowest value for staging adenocarcinoma. Studies of liver intraoperative sonography in the surgical management of pancreatic cancer suggest marginal value for detection of liver lesions in patients with pancreatic adenocarcinoma.11–13 Recent advances in multidimensional thin-section CT and MRI have likely narrowed the margin of benefit of intraoperative sonography for detection of liver lesions in patients with colon cancer metastases. Sahani et al14 found MRI to be as sensitive as intraoperative sonography for liver lesion depiction (87% and 94%, respectively), and that surgical management was altered on the basis of intraoperative sonographic findings in only 4% of cases. Advanced MRI and positron emission tomographic tech-
niques are likely enabling precise detection of liver metastases, obviating the need for routine use of intraoperative hepatic sonography in patients with pancreatic cancer. We found that routine use of intraoperative sonography of the liver affected surgical management in 1 of 470 patients (
ORIGINAL RESEARCH
Routine Intraoperative Hepatic Sonography Does Not Affect Staging or Postsurgical Hepatic Recurrence in Pancreatic Adenocarcinoma Leonora W. Mui, MD, Lisa J. Pursell, BA, Isadora C. Botwinick, MD, John D. Allendorf, MD, John A. Chabot, MD, Jeffrey H. Newhouse, MD Objectives—The purpose of this study was to evaluate the utility of intraoperative sonography of the liver in the staging of pancreatic adenocarcinoma and its impact on the rate of postoperative tumor recurrence in the liver. Methods—We performed a retrospective analysis of the rate in which intraoperative sonography of the liver changed surgical management in 470 surgical candidates with pancreatic adenocarcinoma. In postsurgical patients, we performed a χ2 analysis to examine whether the patients who underwent hepatic intraoperative sonography had a lower rate of recurrent disease in the liver within the first 6 months of surgery compared to patients who did not undergo the procedure. Received March 10, 2013, from the Department of Radiology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore University Medical Center, Manhasset, New York USA (L.W.M.); Departments of Hepatobiliary Surgery (L.J.P., I.C.B., J.A.C.) and Radiology (J.H.N.), Columbia University Medical Center, New York, New York USA; and Department of Surgery, Winthrop University Hospital, Mineola, New York USA (J.D.A.). Revision requested April 24, 2013. Revised manuscript accepted for publication May 20, 2013. We thank members of the Pancreas Center at Columbia University Medical Center for their datamining efforts. This material was presented at the 96th Scientific Assembly and Annual Meeting of the Radiological Society of North America; November 28–December 3, 2010; Chicago, Illinois. Address correspondence to Leonora W. Mui, MD, Department of Radiology, Hofstra North Shore–Long Island Jewish School of Medicine, North Shore University Medical Center, 300 Community Dr, Manhasset, NY 11030 USA. E-mail: [email protected]
Abbreviations
CT, computed tomography; MRI, magnetic resonance imaging doi:10.7863/ultra.33.1.47
Results—Hepatic intraoperative sonography affected management in less than 1% of cases, detecting 1 unsuspected liver metastasis in 470 surgical patients with pancreatic adenocarcinoma. Of 3 patients with equivocal liver lesions identified on preoperative computed tomography or magnetic resonance imaging, hepatic intraoperative sonography excluded metastasis and cleared all the patients for surgical resection. There was no significant difference in postoperative liver recurrence between the group of patients who received intraoperative sonography before resection and patients who did not have the procedure done (P > .99). Conclusions—Routine intraoperative sonography of the liver does not affect staging of pancreatic adenocarcinoma. It may be useful for evaluating equivocal lesions. Key Words—gastrointestinal ultrasound; hepatic metastases; intraoperative sonography; liver; pancreatic adenocarcinoma
P
ancreatic cancer is the 10th most common malignancy and 4th highest cancer killer in adults, with a 5-year overall cumulative survival rate of less than 5%. Surgical resection provides the best curative chance for patients with this malignancy. However, at the time of presentation, about 55% of patients have distant metastases and would not benefit from resection. Preoperative imaging with computed tomography (CT), magnetic resonance imaging (MRI), and endoscopic sonography identifies most surgical candidates. Occasionally, surgical candidates are found during exploratory laparotomy or laparoscopy to have distant metastases, resulting in termination of surgery.
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:47–51 | 0278-4297 | www.aium.org
331jum_online_Layout 1 12/19/13 1:55 PM Page 48
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
Intraoperative sonography of the pancreas and liver is routinely performed during exploratory laparotomy to stage pancreatic cancer (Figure 1). This technique offers excellent spatial and contrast resolution and real-time imaging by applying the ultrasound probe to the organ of interest. Surgical sonography helps delineate the relationship of lesions with vascular and biliary structures and identify small hepatic neoplasms and tumor thrombi.3 Based on studies showing the utility of intraoperative sonography in evaluating liver metastasis in colorectal and hepatocellular carcinoma, intraoperative sonography of the liver is commonly performed in patients with pancreatic adenocarcinoma. The purpose of this study was to evaluate the impact of intraoperative sonography of the liver in the staging and postoperative rate of recurrent liver disease in pancreatic adenocarcinoma.
Materials and Methods Our Institutional Review Board approved this retrospective study with waiver of informed consent. We searched the electronic clinical data warehouse of the Pancreas Center of a major medical center to identify 470 surgical candidates with pancreatic adenocarcinoma between 1995 and February 2011 who underwent intraoperative sonography during exploratory laparotomy or laparoscopy. These patients were deemed resectable on the basis of preoperative CT, MRI, or endoscopic sonography that showed no convincing evidence of metastases. Before surgery, the noninvasive imaging was reviewed at an interdisciplinary conference consisting of surgeons, radiologists, gastroenterologists, and interventionalists to discuss management. During exploratory laparotomy, the surgeons performed visual and manual inspections of the peritoneal cavity, liver, stomach, small bowel, omentum, celiac axis, and head of the pancreas. Figure 1. Pancreatic cancer pathway. EUS indicates endoscopic sonography; and IOUS, intraoperative sonography.
48
Most of the patients (97% [456 of 470]), had open intraoperative sonography, whereas 2% (11 of 270) underwent laparoscopic intraoperative sonography, and 1% (3 of 470) were converted from laparoscopic to open intraoperative sonography. Intraoperative sonography was performed by surgeons who were credentialed by the American College of Surgeons based on both basic and advanced abdominal ultrasound courses involving open and laparoscopic sonography on live pigs. By the study end date, the surgeons had 15, 10, 9, and 6 years of experience with intraoperative sonography. The surgeons used highresolution real-time B-mode intraoperative sonography of the liver with a T-shaped curvilinear array transducer with frequency of 5 MHz using sterile technique. Each of the 8 segments of the liver was evaluated systematically. The right hepatic lobe was interrogated in the transverse direction by identifying the right hepatic vein and scanning posteroanteriorly from the diaphragm to the inferior edge to cover segments 7, 6, 8, and 5. Then the probe was turned longitudinally to the inferior vena cava, and images were scanned right to left to scan segment 4 between the middle hepatic vein and falciform ligament. This scan was repeated at 5-cm intervals from the diaphragm down through the inferior edge of segment 4. The lateral aspect of the left lobe was scanned by getting the heart in view and sweeping through the liver to the left of the falciform ligament. Finally, the probe was placed under the left lateral segment of the venous venosum to evaluate the caudate lobe. Indeterminate liver lesions were biopsied and sent for intraoperative frozen-section analysis. For patients found unresectable at surgery, the medical center’s electronic records were reviewed for radiologic reports, surgical notes, and pathologic reports. We examined the rate at which intraoperative sonography of the liver changed surgical management by detecting liver metastases missed by preoperative imaging and surgical exploration. We reviewed the electronic records of 118 patients who underwent clinical evaluation between 2009 and 2011 to identify 3 cases with equivocal liver lesions on preoperative CT or MRI. In these patients, we specifically investigated the impact of hepatic intraoperative sonography in their surgical management. To examine the impact of hepatic intraoperative sonography on postoperative liver recurrence, we compared the rate of recurrent disease in the liver only between patients who underwent intraoperative sonography at surgery and those who did not undergo the procedure. We investigated the database for patients who received followup care at our institution for at least 6 months after resection between 1995 and 2010 and identified 232 patients
J Ultrasound Med 2014; 33:47–51
331jum_online_Layout 1 12/19/13 1:55 PM Page 49
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
who underwent intraoperative sonography at resection and 58 patients who did not undergo the procedure at surgery. We assumed that any patient with negative preoperative imaging findings, in whom surgical and pathologic findings indicated complete resection of the primary tumor and who developed hepatic metastases after surgery with no evidence of tumor in any other site at the time the hepatic metastases appeared, probably had hepatic metastases too small to find at the time of surgery. Presumably, these liver lesions were present at surgery but too small to detect by inspection or intraoperative sonography. Followup imaging included CT, MRI, or positron emission tomography/CT. A χ2 analysis was performed to test statistical significance.
Results Seventeen percent of the surgical candidates (80 of 470) were unresectable, including 4% with distant metastases (19 of 470). Three percent of the operations (14 of 470) were terminated because of liver metastases, which were pathologically proven to be adenocarcinoma. The rate at which intraoperative sonography of the liver detected occult lesions was less than 1%; there was 1 case among the 470 sampled in which intraoperative sonography of the liver detected an unsuspected liver metastasis missed by preoperative CT and surgical exploration (Figure 2). This patient had no evidence of metastatic disease on preoperative imaging and no visible or palpable lesions during surgical exploration. Pathologic examination confirmed adenocarcinoma, and surgery was terminated. In another case, intraoperative sonography revealed additional liver lesions in a patient with diffuse biliary ductal dilatation without discrete lesions on MRI; however, liver metastases visualized and palpated during surgical exploration already disqualified surgery as a treatment option, and additional liver lesions seen on intraoperative sonography did not affect management. Sampling of 118 surgical candidates between 2009 and February 2011 found 3 cases in which preoperative imaging questioned possible liver lesions. At surgical exploration with intraoperative sonography, these lesions were found to be negative for metastasis on the basis of sonographic features. There was another case in which routine hepatic intraoperative sonography revealed an indeterminate lesion, prompting tissue sampling with negative results for metastasis. The 6-month postoperative rate of recurrent disease in the liver was identical in the patients who underwent intraoperative sonography at surgery (7% 17 of 232) and
J Ultrasound Med 2014; 33:47–51
the patients who did not undergo the procedure (4 of 58), with χ2 P > .99 (Figure 3).
Discussion The utility of intraoperative sonography in evaluating patients with hepatobiliary disease has been extensively studied.4–7 Machi et al4–6 found that intraoperative sonography provided beneficial information affecting the staging and management of 89% of hepatic operations, 83% of noncalculous biliary operations, and 73% of pancreatic operations. The overall accuracy of intraoperative sonography in detecting liver metastasis from colorectal carcinoma was 94% in one series, higher than the accuracy of CT (75%) or MRI (80%).4 Zacherl et al8 concluded that intraoperative sonography is crucial in the staging of patients with hepatocellular carcinoma and colorectal metastases since it changed the surgical strategy in about 23% of cases. A prospective study by Guimarães et al9 in 60 patients with a mix of abdominal tumors, including colFigure 2. Impact of hepatic intraoperative sonography (IOUS) on staging patients with pancreatic adenocarcinoma.
49
331jum_online_Layout 1 12/19/13 1:55 PM Page 50
Mui et al—Intraoperative Hepatic Sonography in Pancreatic Adenocarcinoma
orectal cancer, bile duct tumors, hepatic tumors, and 4% of cases with pancreatic tumors, found that intraoperative sonography offered 91% sensitivity, 78% specificity, a 91% positive predictive value, and a 78% negative predictive value for detection of liver metastasis, with results changing the surgical strategy in 41% of cases. A recent interdisciplinary review of intraoperative sonography in pancreatic surgery by Ní Mhuircheartaigh et al10 found the technique highly useful in guiding Puestow and pseudocyst drainage procedures and localizing foci of neuroendocrine tumors and moderately useful for staging resection of nonadenomatous pancreatic masses but of the lowest value for staging adenocarcinoma. Studies of liver intraoperative sonography in the surgical management of pancreatic cancer suggest marginal value for detection of liver lesions in patients with pancreatic adenocarcinoma.11–13 Recent advances in multidimensional thin-section CT and MRI have likely narrowed the margin of benefit of intraoperative sonography for detection of liver lesions in patients with colon cancer metastases. Sahani et al14 found MRI to be as sensitive as intraoperative sonography for liver lesion depiction (87% and 94%, respectively), and that surgical management was altered on the basis of intraoperative sonographic findings in only 4% of cases. Advanced MRI and positron emission tomographic tech-
niques are likely enabling precise detection of liver metastases, obviating the need for routine use of intraoperative hepatic sonography in patients with pancreatic cancer. We found that routine use of intraoperative sonography of the liver affected surgical management in 1 of 470 patients (
