| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Department of Radiology, University of Michigan, Ann Arbor, MI 48109.
2 Present address: William Beaumont Hospital, 3601 West Thirteen Mile Rd., Royal
Oak, MI 48073.
3 Present address: Department of Radiology, Providence Hospital, Southfield, MI
48075.
Received March 2, 2005;
accepted after revision May 3, 2005.
Address correspondence to H. V. Nghiem
(hnghiem{at}beaumont.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Imaging-guided core biopsy is sensitive, safe, and accurate in the diagnosis of malignant lesions of the pancreas. Benign biopsy findings cannot be used to exclude the presence of a neoplasm, and repetition of a biopsy should be considered if there is high clinical suspicion of malignancy.
Keywords: biopsy • core biopsy • CT • fine-needle aspiration • pancreas • sonography
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although needle biopsy during laparotomy has been performed with the fine-needle aspiration (FNA) technique since the 1960s [1] and with the core technique since the 1970s [4], advances in endoscopy and cross-sectional imaging allow less invasive methods of diagnosis and biopsy that avoid the morbidity and mortality of a major surgical procedure. Endoscopic techniques such as endoscopic sonography and ERCP provide additional diagnostic information about tumor resectability, aid in evaluation for biliary and pancreatic ductal obstruction, allow brush and FNA biopsies, and allow concurrent placement of an endobiliary stent. CT and sonography are commonly used to evaluate and characterize pancreatic masses and to guide percutaneous pancreatic biopsy.
Percutaneous imaging-guided FNA has been in use since the 1970s [5] but requires experienced radiologists and cytopathologists for greatest effectiveness [6]. Even experienced personnel find it challenging to diagnose pancreatic adenocarcinoma from cytologic samples [7]. As an alternative, imaging-guided percutaneous biopsy of the pancreas with an automated-fire core biopsy needle avoids the need for experienced cytopathologists. Although percutaneous core biopsy of pancreatic masses has been performed since the 1980s [8], the performance of core biopsy has been reviewed in few studies. To evaluate the sensitivity and accuracy of percutaneous pancreatic core biopsy, we reviewed consecutive imaging-guided biopsies performed at our institution during a 6-year period.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Biopsies were performed on an outpatient basis by members of our cross-sectional interventional team. The results of a platelet count and coagulation screen were reviewed before biopsy. Any abnormalities in these laboratory values were corrected before biopsy. All patients provided informed consent for the procedure. The radiologists performing the biopsies chose CT or sonographic guidance on the basis of their preference. CT-guided biopsy was performed with a helical single-detector scanner (Ct/i, GE Healthcare). CT-guided biopsy was routinely performed without IV contrast medium; a contrast-enhanced abdominal CT scan obtained before biopsy was used as a reference. Sonographically guided biopsies were performed with electronically focused sector transducers ranging in frequency from 2.5 to 5.0 MHz (Logiq 700, GE Healthcare; ATL 5000, ATL; and Philips 5000, Philips Medical Systems). Before sonographically guided biopsy, color-flow Doppler sonography was routinely used to identify major blood vessels. Standard practice for biopsy of pancreatic masses at our institution is a coaxial technique with a 17-gauge introducer needle and an 18-gauge fully automated biopsy gun for acquisition of two or three cores of tissue. All biopsies were performed through an anterior abdominal approach, avoiding traversing the colon, small bowel, and liver. (On occasion, when a transgastric approach does not provide adequate access to the mass, we use a transhepatic or transenteric approach.) Local anesthesia with lidocaine and IV conscious sedation with fentanyl citrate (Sublimaze, Akron) and midazolam hydrochloride (Versed, Ben Venue) was routinely used. Patients receiving IV conscious sedation underwent hemodynamic and pulse oximetric monitoring. After biopsy, the core samples were sent to the pathology department for later review. Patients were monitored in the recovery area after the biopsy procedure, typically for 4 hours, before discharge.
We identified a total of 119 biopsies of native pancreatic glands performed on 114 patients during the study period. Four biopsies in four patients were performed with the FNA technique only (all performed very early in the study period) and thus were excluded from the review. Our final study group comprised 115 core biopsies performed on 110 masses in 110 patients (57 men and 53 women) ranging in age from 36 to 85 years (mean age, 62.5 years). The mean mass size was 3.8 cm (range, 1.2-8.1 cm). Location of the mass within the pancreas included 79 in the head, 18 in the body, and 13 in the tail. Biopsy of 18 masses was initially performed with CT guidance. The mean mass size was 3.1 cm (range, 2-5.6 cm). Twelve of these masses were in the head of the pancreas, one in the body, and five in the tail. Biopsy of 92 masses was initially performed with sonographic guidance. The mean size of these masses was 3.9 cm (range, 1.2-8.1 cm). Sixty-seven of these masses were in the head of the pancreas, 17 in the body, and eight in the tail. The five repeated core biopsies were performed with CT (n = 3) or sonographic (n = 2) guidance. The mean size of these masses was 3.2 cm (range, 2.0-4.3 cm). Three of these masses were in the head of the pancreas, one was in the body, and one was in the tail.
Sixty-five (59%) of the 110 patients referred for imaging-guided biopsy had undergone at least one previous biopsy attempt, 10 patients having undergone two and one patient having undergone three previous biopsies. Techniques used included endoscopic sonography (n = 42), esophagogastroduodenoscopy-ERCP (n = 24), CT-guided percutaneous biopsy (n = 4), and surgical biopsy (n = 5). All four of the CT-guided biopsies had been performed at outside institutions.
In the initial 110 biopsies, four of the core samples did not contain pancreatic tissue, only blood clot (n = 1); gastric mucosa (n = 1); extrapancreatic fibroadipose tissue (n = 1); or skeletal muscle, intestinal mucosa, and extrapancreatic fibroadipose tissue (n = 1). Five patients underwent repeated percutaneous core biopsy, including three of the four patients with insufficient samples for diagnosis. In the fourth patient, core biopsy was repeated to confirm a result suspicious for malignancy. In the fifth patient, the initial biopsy revealed only fibrosis without evidence of malignancy; the repeated biopsy showed pancreatic adenocarcinoma.
Three of the 110 patients were excluded from statistical analysis. Two of these three patients, who did not have a history of pancreatitis and who had solitary masses of the pancreatic head, had core biopsy results consistent with chronic pancreatitis. Because neither patient had follow-up clinical or imaging data available, the final diagnosis was uncertain. The third patient excluded from analysis had a biopsy result that did not provide enough information for a diagnosis without a repeated biopsy or further clinical or imaging follow-up at our institution. The other 107 patients had surgical, clinical, or imaging follow-up findings available with a mean follow-up period of 7.5 months (range, 1-24 months).
A biopsy result was considered true-positive if the result of histopathologic analysis was positive or strongly suggestive of malignancy. A biopsy result was considered true-negative if the result of histopathologic analysis was negative for malignancy without clinical evidence or subsequent imaging follow-up findings suggestive of malignancy. Conversely, a biopsy result was considered false-negative if the result of histopathologic analysis was negative for malignancy but clinical or imaging follow-up findings suggested malignancy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We identified nine true-negative biopsy findings, representing chronic pancreatitis (n = 7), scar tissue (n = 1), and normal pancreatic tissue (n = 1). Each biopsy had been performed with sonographic guidance. These results were confirmed with surgical biopsy in five cases: three cases of chronic pancreatitis and the biopsy samples containing scar and normal pancreatic tissue. The other four patients had clinical follow-up findings (mean follow-up period, 11 months; range, 5-18 months) consistent with the histologic diagnosis of chronic pancreatitis.
We identified six false-negative biopsy results, each biopsy having been performed with sonographic guidance. The mean mass size in these cases was 3.0 cm (range, 1.7-4 cm). The final clinical diagnosis was pancreatic adenocarcinoma established by surgical biopsy (n = 1), evidence of metastatic disease at initial (n = 1) or follow-up (n = 1) imaging, death after discharge to hospice (n = 1), or treatment based on the clinical, imaging, and laboratory presentation (n = 2). There were no false-positive biopsy results.
For CT and sonography combined, the sensitivity and accuracy of imaging-guided core biopsy were 93.9% and 94.4%, respectively. For CT alone, the sensitivity and accuracy were both 100%. For sonography alone, the sensitivity and accuracy were 92.5% and 93.3%, respectively. The negative predictive value for imaging-guided biopsy was 60.0%.
Complications occurred during three biopsies, all performed with sonographic guidance; two of these biopsies were of pancreatic head masses and one was of a pancreatic tail mass. Acute hypotension developed in one patient immediately after biopsy. Echogenic free abdominal fluid was found at repeated sonography and necessitated fluid resuscitation and blood transfusions before emergency surgical repair of a right gastric artery laceration. The second patient had abdominal pain, abdominal distention, and a decrease in hematocrit level from 34.1% to 25.6% on the morning after biopsy. These complications were caused by retroperitoneal hemorrhage and necessitated transfusion of two units of packed RBCs with subsequent stabilization of the hematocrit level and resolution of symptoms. The third patient, who had a history of chronic pancreatitis, needed a 2-day hospital admission for pain management beginning the day after biopsy. Follow-up imaging showed no biopsy complications, and there was no biochemical evidence of acute pancreatitis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Zech et al. [9] reviewed 63 core biopsies performed with CT rather than sonographic guidance and found a sensitivity of 78% and an accuracy of 81%. They postulated that the relatively lower sensitivity of CT-guided biopsy in their study compared with sonographically guided biopsies in other studies may have occurred because more difficult cases were triaged away from sonography to CT or because of technical factors such as needle displacement in the time between needle positioning and deployment. In an analysis of the sensitivity of CT-guided biopsy, Brandt et al. [13] pooled data from 211 CT- and 58 sonographically guided percutaneous core and FNA biopsies performed between 1985 and 1989 and found a lower sensitivity for CT. They suggested that several factors may have contributed to the relatively higher sensitivity of sonography, including the possibility that more easily accessible masses may have been triaged to sonography, that sonography may be more accurate because it allows continuous real-time sonographic observation of the needle tip in relation to the mass, and that because fewer radiologists were involved with the sonographically guided biopsies in their study, experience may also be an important factor.
Although we use CT guidance for biopsy of pancreatic masses in a limited subset of patients, we favor the use of sonographic guidance when technically feasible because of the portability of the equipment, lack of ionizing radiation, ability to view the biopsy needle tip in real time, and efficiency [14]. Our reported sensitivity and accuracy of 93.9% and 94.4% are concordant with results of studies of percutaneous core biopsy of pancreatic masses. In our study, we found similar overall rates of accuracy and sensitivity for CT and sonographic guidance, although we performed relatively few CT-guided biopsies in the time frame of the study.
That we did not identify any false-positive biopsy results for pancreatic malignancy is consistent with the body of data on pancreatic core biopsy [9, 10, 12, 13]. Evaluation of specificity is limited because relatively few cases proceed to surgical resection; thus, confirmation of the histopathologic diagnosis typically relies on concordance between the biopsy results and the final clinical diagnosis or outcome. However, if the false-positive rate for pancreatic core biopsy is confirmed to be at or near zero, core biopsy may have an advantage over FNA, in which false-positive results have been reported, yielding specificity ranging from 82% to 100% [15-19].
False-negative results of core biopsy do occur [9-13]. They have been postulated to relate to sampling error and interpretation difficulties from the strong desmoplastic reaction elicited by pancreatic adenocarcinoma [9] and to technical errors related to needle misplacement [12], a problem more commonly seen with small masses [13]. However, most studies of FNA biopsy of pancreatic masses by the intraoperative, percutaneous, and endoscopic routes commonly show false-negative results and biopsies in which there is not enough information for a diagnosis [15, 20-27]. The reported sensitivity of percutaneous imaging-guided FNA biopsy varies between 85% and 100% [18, 28-31].
The clinical impact of false-negative biopsy results is best expressed by the negative predictive value, which in our study was 60%. We believe that the negative predictive value of percutaneous core biopsy is unacceptably low for reliable exclusion of the presence of pancreatic malignancy, and we concur with other investigators that negative results of percutaneous biopsy, whether core or FNA, should be viewed with caution [12, 17]. In the appropriate clinical situation, however, a nonmalignant biopsy result may be considered adequate evidence to support cautious clinical and imaging follow-up of a patient with chronic pancreatitis.
Although some proponents of FNA have suggested a higher complication rate for core biopsy relative to FNA [32], our complication rate of 2.6% using the core biopsy technique compares favorably with reported rates of 0-3.3% [9-13] for core and 0-4.9% for FNA [13, 17-19, 31] biopsy of pancreatic masses performed with imaging guidance.
Patients are commonly referred to our percutaneous interventional service having undergone a biopsy, usually by the endoscopic route, that did not yield enough information for a diagnosis. We have found that the sensitivity and accuracy of pancreatic biopsy remain high despite the failure of previous attempts to obtain a confident histologic diagnosis.
Limitations of this study included the nonstandardized short-term follow-up stemming from its retrospective nature and the gathering of the patient population from a tertiary care, academic medical center with a large oncology referral base.
In conclusion, imaging-guided percutaneous core biopsy is sensitive, safe, and accurate in the diagnosis of malignant tumors of the pancreas, even in cases in which other initial biopsy methods do not give enough information for a diagnosis. Most of our biopsies were performed with sonographic guidance. We continue to favor the use of sonography as the primary imaging technique for percutaneous biopsy of pancreatic masses. Benign biopsy findings cannot be used to exclude the presence of pancreatic malignancy, and repetition of the biopsy should be considered if there is high clinical suspicion of the presence of malignancy. However, careful clinical and imaging follow-up may be appropriate in the care of patients with a history of chronic pancreatitis and benign biopsy results.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Kothary, S. Dieterich, J. D. Louie, D. T. Chang, L. V. Hofmann, and D. Y. Sze Percutaneous Implantation of Fiducial Markers for Imaging-Guided Radiation Therapy Am. J. Roentgenol., April 1, 2009; 192(4): 1090 - 1096. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
