点击显示 收起
1 From the Section of Vascular and Interventional Radiology, Department of Diagnostic Radiology, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 325, Houston, TX 77030. Received February 9, 2001; revision requested March 20; revision received June 6; accepted July 5.
| ABSTRACT |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
MATERIALS AND METHODS: The authors retrospectively reviewed the records of 29 patients with lesions located within (n = 24) or around (n = 5) the pancreatic head who underwent CT-guided FNAB with the posterior transcaval route due to obstruction of the anterior approach by bowel, liver, and/or other structures. A coaxial needle technique was used, with an outer 18-gauge needle positioned posterior to the IVC and an inner 22-gauge needle traversing the IVC to obtain a biopsy of each lesion. All biopsy specimens were subjected to cytologic evaluation. Medical records of all patients were evaluated for complications.
RESULTS: All lesions were safely accessed with the posterior transcaval approach without major complications. The biopsies revealed a malignant process in 21 patients, benign pancreatic cysts in two patients, and pancreatitis in one patient. There were five false-negative biopsy results. Minor complications occurred in four patients (small retroperitoneal hematomas occurred in three and abdominal pain occurred in one).
CONCLUSION: CT-guided coaxial FNAB by means of a posterior transcaval approach is a safe method for obtaining samples from lesions in or around the pancreatic head.
Index terms: Biopsies, 771.1262 • Computed tomography (CT), guidance, 771.1211 • Pancreas, biopsy, 771.1262 • Pancreas, cysts, 771.312 • Pancreas, neoplasms, 771.312, 771.321, 771.33 • Pancreatitis, 771.291
| INTRODUCTION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
At our institution, we have routinely performed percutaneous CT-guided FNAB of masses in or around the head of the pancreas with a posterior approach and a coaxial needle technique, with the fine needle traversing the inferior vena cava (IVC) before entering the target lesion. The purpose of our study was to evaluate the safety and feasibility of this technique.
| MATERIALS AND METHODS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Diagnostic abdominal CT scans of all patients were available prior to the biopsy procedure. The masses were located within the head of the pancreas, within the uncinate process of the pancreas, or within both in 24 patients. In the other five patients, CT revealed enlarged peripancreatic lymph nodes (n = 2), infiltrative soft-tissue lesions around the pancreatic head (n = 2), or a cystic lesion posterior to the pancreatic head (n = 1). The largest dimensions of the masses ranged from 1.0 cm to 8.0 cm (mean, 2.7 cm). Sixteen masses (55%) were 1–2 cm, 12 (41%) were 2–4 cm, and only one was more than 4 cm. For all patients, review of the diagnostic CT scan suggested that an anterior approach would involve transgression of the gastrointestinal tract (stomach, colon, or small bowel), the liver, or both. Two patients, in addition, had large venous collateral vessels located anterior to the pancreatic lesion. No substantial IVC abnormality (eg, intraluminal clot or filter) was detected on the CT images in any of the patients.
Twenty-one patients had no personal history of cancer; eight had a known extrapancreatic malignancy at the time of FNAB. Eleven patients had undergone previous nondiagnostic biopsy procedures (endoscopic US–guided FNABs in eight patients and CT-guided biopsies at outside hospitals in three patients).
Prothrombin time, partial thromboplastin time, and platelet counts were routinely obtained in all 29 patients prior to the biopsy to rule out any hemorrhagic diathesis. All biopsies were performed after the patients received drugs for conscious sedation and local anesthesia.
All procedures were performed with a commercially available CT scanner (CTi Smart View; GE Medical Systems, Milwaukee, Wis). The patients underwent unenhanced CT examinations in the prone position, and a needle course was outlined by means of the CT scans. The images were carefully scrutinized to help avoid the right renal artery, which courses posterior to the IVC, and to ensure that the needle entered below the diaphragm. Biopsies were performed by seven attending radiologists or by a resident or fellow under direct supervision by an attending radiologist.
After the patient was properly prepared and draped, local anesthetic was instilled. The tip of an 18-gauge (10- or 15-cm-long) guiding needle (Chiba biopsy needle; Cook, Bloomington, Ind) was advanced percutaneously to a position immediately posterior to the IVC. The needle-tip position and the trajectory of the needle were adjusted based on CT scans obtained intermittently after partial needle advancement; we did not use CT fluoroscopy in any of our patients. In two patients, interposed pleura and lung precluded a direct vertical posterior approach; in these patients, a skin entry site below the transverse plane of the tumor was selected and the needle was angled upward into the pancreatic lesion, thus avoiding the posterior costophrenic pleural recess. When the needle angle and position were appropriate, the distance from the needle tip to the target lesion was measured on the CT image. A 22-gauge Chiba biopsy needle (Cook) of appropriate length was then advanced coaxially through the guiding needle, through the IVC, and into the target lesion (Figs 1, 2). If necessary, a curved 22-gauge biopsy needle was used to sample a particular portion of the lesion (eg, a nonnecrotic, solid part of the tumor; Fig 3). The tip of the needle was grasped with a hemostat and bent to impart a gentle curve to it. After the CT scan had documented that the position of the needle tip was satisfactory, an aspirate was obtained. The biopsy needle was then withdrawn; the guiding needle was left in place for further passes if required.
|
|
|
|
|
|
All biopsies were performed on an outpatient basis unless the patient was already hospitalized for another reason. Inpatients were returned to their hospital room for observation. Outpatients were observed in the radiology nursing unit, and their vital signs were frequently monitored. The patients were discharged after 1 hour if no complications were evident and only if each patient was accompanied by a responsible companion. Patients were given a discharge sheet with instructions to call the emergency center of the hospital if symptoms suggestive of complications appeared. Patients with clinical and/or radiologic evidence of complications were observed for a longer period of time, which, although not recorded, ranged from 4 to 6 hours. These patients underwent a repeat follow-up CT examination if the attending interventional radiologist who performed the biopsy believed it necessary. The medical records of each patient were retrospectively reviewed for evidence of delayed complications. Clinical follow-up ranged from 4 days to 20 months (mean, 5 months) and was based on subsequent hospital visits (27 patients) or telephone inquiry (two patients). If medical records indicated that a patient had developed one or more hematomas during the biopsy procedure, the size of the hematomas was measured (by S.G.) on images retrieved from optical disk archives (at our institution, all images from CT-guided interventional procedures are archived on optical disks).
When cytologic analysis provided an unequivocal result, it was considered to be the definitive diagnosis. However, if the cytologic specimen showed no evidence of malignancy, available clinical, radiologic, and histologic follow-up information was reviewed to determine if the specimen diagnosis at FNAB was accurate.
| RESULTS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
In eight of the 29 patients, there was no cytologic evidence of malignancy. Two patients had cystic lesions at CT examination, and the fine-needle aspiration yielded fluid with no malignant cells. The lesions remained stable on follow-up CT scans obtained 8 months after FNAB in one of the two patients and 1 year after FNAB in the other patient; the lesions were diagnosed as chronic benign pancreatic cysts. Another patient, who had a history of established chronic pancreatitis, presented with worsening abdominal pain and the suggestion of a mass in the region of the head of the pancreas. Because there was a strong clinical suspicion of malignancy, this patient underwent one endoscopic US–guided biopsy followed by two CT-guided FNABs; all three biopsies revealed features of both acute and chronic pancreatitis with no evidence of malignancy. At the time of this writing, the patient is being monitored in a clinical follow-up regimen; the presumptive diagnosis is acute pancreatitis superimposed on chronic pancreatitis. The FNAB results from all three of these patients were considered true-negative and prevented unnecessary operations. Thus, FNAB enabled a diagnosis in 24 of 29 patients (83%).
Subsequent biopsy results from the other five patients (17%) with negative CT-guided FNAB results later established the presence of malignancies. One patient with two negative biopsy results underwent an endoscopic US–guided biopsy that revealed a neuroendocrine tumor. An endoscopic US–guided biopsy in another patient yielded ductal adenocarcinoma. Laparoscopic biopsy of a liver lesion in one patient and a mesenteric nodule in another revealed metastatic adenocarcinoma, presumably from pancreatic cancer. The last patient underwent pancreatoduodenectomy, and histopathologic analysis of the gross pathology specimen revealed an islet cell tumor. These false-negative FNAB results were attributed to sampling error.
All patients tolerated the procedure well; none had major complications (ie, complications requiring a blood transfusion, angiographic or surgical evaluation, or hospital admission) either during or after the procedure. Minor complications occurred in four patients. Three patients developed small (2.6 x 2.4-cm, 1.5 x 1.9-cm, and 2.2 x 2.7-cm) retroperitoneal hematomas (Fig 4) that were detected on CT images obtained to verify the position of the 22-gauge biopsy needle during the second (two patients) or third (one patient) needle pass. Two of these patients underwent a follow-up CT examination (conducted 3 hours later in one patient and 5 hours later in the other); neither examination showed an increase in the size of the hematomas. All three patients were observed for several hours, remained completely asymptomatic, did not require blood transfusion, and were discharged in stable condition. One patient had mild abdominal pain after the procedure; a limited abdominal CT scan did not reveal any marked abnormality, the patient’s pain was relieved with analgesic medications, and the patient was discharged in stable condition.
|
|
| DISCUSSION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
It is generally acceptable to traverse the gastrointestinal tract or liver in an immunocompetent patient when performing FNAB of abdominal masses (1–6,10,15). In the study by Brandt et al (1), none of the patients had complications related to the biopsy route. Many other studies (2–6) have also shown that transgression of the bowel during percutaneous pancreatic biopsy is unlikely to result in substantial complications. Although some authors (1) state that traversing the colon is acceptable during pancreatic biopsies, others (3,6) avoid the colon, especially when using large-bore needles.
Acute pancreatitis is an uncommon but potentially fatal complication of needle biopsy of the pancreas (16,18,19). Mueller et al (19), in their series of 184 pancreatic biopsies in 178 patients, observed severe postprocedure pancreatitis in five patients (3%). Although in three of these five patients the needle was known to have transgressed the bowel during biopsy, the authors believed that a definite conclusion as to whether transintestinal biopsy increases the risk of pancreatitis could not be made because bowel transgression probably occurs in many pancreatic biopsies. Ferrucci et al (11) reported a case of gram-negative sepsis after FNAB of the pancreas. They attributed the sepsis to bowel transgression and resultant needle implantation of intestinal flora. Martino et al (20) reported a case of secondary bacterial contamination of an endometrioma, presumably resulting from bowel transgression during percutaneous FNAB. We believe that because bowel transgression can potentially contaminate a sterile biopsy procedure, it is best avoided to preclude even a small risk of infection or leakage. None of the patients in our study developed pancreatitis.
At our institution, we routinely use a posterior approach in needle biopsy of pancreatic head lesions if a review of diagnostic CT images fails to reveal a direct anterior approach that avoids other organs. In our experience, a posterior approach is less painful to the patient because it avoids a transperitoneal puncture. In a posterior approach, the needle is inserted through the paraspinal muscles and advanced lateral to the IVC, between the IVC and the vertebra, or, if these two routes are not feasible, through the IVC itself.
Although it seems logical to avoid major vascular structures located adjacent to the target biopsy site while performing image-guided percutaneous needle biopsies, many studies have shown that the transgression of vascular structures, especially the low-pressure veins, does not markedly increase the complication rate (10,11,16,17,21–23). Translumbar aortography, which involves the use of larger needles and puncture of a much higher-pressure system, has been used safely for many years. Gupta et al (21) obtained FNAB samples from superior mediastinal masses with a US-guided transvenous (jugular or brachiocephalic vein) route in five patients without reported complications. In a series of 12 US-guided supraclavicular biopsies reported by Yang et al (22), the needle had to traverse the jugular vein in four patients, and no complications occurred. Yang et al believed that if vessel penetration is unavoidable, the penetration of veins with thin (22-gauge) needles would be acceptable. We also did not encounter any major hemorrhagic complications in our patients who underwent transcaval biopsies; the appearance of small, clinically silent needle-track hematomas during CT-guided biopsy procedures (noted on CT scans obtained during the biopsy procedure in three of our patients) is not uncommon after percutaneous needle biopsies of abdominal masses (24). On the basis of our experience, we believe that transgression of the IVC with fine needles does not cause substantial bleeding. Because routine postbiopsy CT scans were not obtained and hematocrit values were not measured in our study, it is possible that we may have failed to observe a few hemorrhages; however, none of our patients developed clinically important hematomas. Although inadvertent puncture of the aorta and hilar vessels without substantial complications has been reported (23), and some authors do not regard the puncture of arteries as exceedingly hazardous, we believe that transarterial punctures should be avoided when possible; life-threatening hemorrhages can result from arterial puncture during percutaneous biopsy procedures (25). We took care to identify and avoid the right renal artery, which courses posterior to the IVC. Many of the previous studies on pancreatic biopsies included a number of US-guided procedures (1,4,9), and because it is not possible to detect small biopsy-site or needle-track hematomas at US, it is likely that many small hematomas go undetected. Major hemorrhage after pancreatic biopsy, although uncommon, has been reported in the literature (26); most of these cases resulted from biopsies performed with the anterior approach. These hemorrhages probably resulted from the inadvertent puncture of mesenteric or pancreatic vessels. Pancreatic carcinoma can cause portal vein thrombosis with resultant formation of collateral circulation vessels that can also preclude the use of an anterior approach, as was the case in two of our patients (Fig 2a).
Endoscopic US–guided FNAB has been shown to enable a cytologic diagnosis in 80%–90% of pancreatic neoplasms (12–14). However, this technique has some limitations. As noted by Voss et al (14), pancreatic tumors located in the uncinate process may prove difficult to sample in endoscopic US–guided biopsy procedures because the position of the endoscope in the second part of the duodenum decreases the force of the biopsy needle below the level necessary to penetrate pancreatic tissue. The presence of prominent stromal fibrosis in pancreatic lesions may also interfere with needle penetration. Occasionally, the presence of vascular structures in the bowel wall can preclude endoscopic US–guided needle biopsy. Also, the success of endoscopic US–guided FNAB is very operator dependent. Eight of our patients had previously undergone nondiagnostic endoscopic US–guided biopsies.
Percutaneous needle biopsy of the pancreas has become a well-established technique (1–9,27) performed either to help confirm a diagnosis of ductal adenocarcinoma or to enable the diagnosis of other conditions appearing as pancreatic masses at imaging examinations. FNAB is the most commonly used biopsy technique because it combines a high diagnostic yield and a low complication rate. Although large-bore needle biopsy has been advocated by some authors (4), it was not used in our series because of safety concerns. It is generally accepted that fine-needle passage through vascular structures does not lead to substantial complications; it seems prudent, however, to avoid passing large-bore cutting needles through the IVC. In all our patients, FNAB allowed the cytopathologist to view the aspirated cellular material in an immediate microscopic evaluation. This immediate cytologic evaluation permits the cytopathologist and the radiologist to determine whether subsequent passes are required either to obtain additional specimens or to perform ancillary studies. An immediate cytologic result also allows the radiologist to decide whether to attempt a biopsy of a different area of the lesion if the initial aspirate showed only necrotic material, fibrotic tissue, or nonspecific inflammatory cells. This is especially important in pancreatic biopsies, because pancreatic tumors are known to be associated with extensive fibrosis and inflammation in the surrounding tumor-free pancreatic tissue (7,27). The coaxial needle technique (used in all of our patients), along with the use of curved 22-gauge biopsy needles (used in many of our patients), allows multiple samples from different areas of the tumor to be obtained safely and expeditiously through the same guide needle.
The reported accuracy of CT-guided pancreatic biopsy ranges from 50% to 94% (1–9,27). In our study, FNAB was effective in establishing the diagnosis in 83% of patients. The five false-negative biopsies in our study were probably related to sampling error. Pancreatic malignancies initiate an intense desmoplastic reaction around the tumor (27) or obstruct the major pancreatic duct, causing either acute or chronic pancreatitis. Hence, on CT scans, the pancreatic mass may appear larger than the actual tumor itself is, thereby increasing the chances of sampling error. Tumor necrosis or bleeding can also lead to the insufficient recovery of tumor cells. The sampling error in our series could also have been related to the fact that we did not inject an intravenous contrast agent during the CT-guided biopsy procedures; the use of contrast material may help in identifying small, isoattenuating tumors or in demonstrating necrotic regions within a tumor, thus facilitating a more accurate sampling (1,6).
In conclusion, our results indicate that the technique of CT-guided, percutaneous, transcaval, coaxial FNAB provides easy and safe alternative access to masses in or around the head of the pancreas or the uncinate process of the pancreas. This technique also facilitates the acquisition of multiple biopsy samples from different regions of the tumor.
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|