Abstract
Aim: To prospectively estimate the safety, feasibility and accuracy of 1.0T open-magnetic resonance (MR)-guided percutaneous biopsy in free-hand of focal lesions located in the hepatic dome. Materials and Methods: All percutaneous MR-guided biopsies of the lesions were performed at the Shandong Medical Imaging Research Institute. Thirty-six patients with focal liver lesions located in the hepatic dome measuring 6-36 mm in the maximum diameter were included in this study. Lesions were divided into two groups on the basis of maximum nodule diameters: ≤1.5 cm (n=14) and >1.5 cm (n=22). Surgical pathology of nodules or clinical and imaging follow-up for at least 12 months were used to establish final diagnosis. Diagnostic accuracy, specificity, sensitivity, average procedure time and complications were recorded and analyzed. Results: All obtained samples were sufficient for diagnosis. Accuracy, sensitivity and specificity of MRI-guided percutaneous liver lesions biopsy in diagnosing malignant tumors were 97.2%, 96.7% and 100%, respectively. Accuracy, sensitivity and specificity were 92.9%, 90.9% and 100% for the lesions that were 1.5 cm or smaller in diameter and 100%, 100% and 100% for lesions larger than 1.5 cm in diameter, respectively. There was no significant difference between the two groups (p>0.05). The average procedure time for each pass of the needle from skin entry to the target lesion was 1.1 min and the total procedural time was 28.5 min. Biopsy-induced complications included peri-hepatic hemorrhage in 8.3% (3/36) of cases. No serious complications occurred. Conclusion: 1.0T open-MR-guided free-hand percutaneous biopsy is a safe, feasible and highly accurate diagnostic method for pathological diagnosis of focal liver lesions located in the hepatic dome.
An accurate and timely diagnosis of liver lesions is fundamental to providing patients with malignancy a potential for optimized intervention. Although, in some cases, the combination of imaging, clinical and biochemical evidence can be obtained to provide a definite diagnosis (1), histopathology of some lesions is essential for clinical intervention, especially for suspected malignancy. Image-guided percutaneous liver biopsy is known as a safe and accurate diagnostic procedure for the evaluation of focal or diffuse hepatic disease. Ultrasound (US) and computed tomography (CT) are the most popular modalities for puncture guidance (1-4), although it is a challenge for them when the lesions are located in the hepatic dome. In this situation, it is usually insufficiently visualized by sonography, which has a limited sonic window due to the overlapped lung and/or ribs, and, thus, CT-guided procedure needs to achieve percutaneous transpulmonary biopsy or go through a long procedure that includes the risk of pneumothorax and/or increased difficulty in achieving a successful puncture (5). There are several techniques, including the use of an artificial fluid (artificial pleural effusion or artificial ascites) or air, to overcome these limitations (6). However, all these methods ask for the operators having a certain degree of surgical skill and need a second surgical trauma, not to mention adding the procedure time (7). In recent years, magnetic resonance imaging (MRI)-guided percutaneous biopsy has been successfully used in different anatomical organs (8), such as the lungs, brain and prostate. Magnetic resonance (MR) offers superior soft-tissue and true multi-planar imaging capabilities, which promote the display and location of hepatic dome lesions. In this study, we apply the technique of free-hand combined with MR fluoroscopy to prospectively evaluate the feasibility, safety and accuracy of MR-guided percutaneous biopsy of focal lesions located in the hepatic dome using a 1.0T high-field scanner.
Materials and Methods
The Institutional Review Board of our Hospital reviewed and approved this study plan. Both research informed consents and procedural informed consents were obtained from all participating patients prior to study participation. The risks, benefits and alternatives were discussed in detail.
Patient population. Between October 2014 to November 2015, there were 36 consecutive patients (27 males, 9 females, with ages ranging from 34 to 76 years; mean age=55.6) with 36 liver lesions located in the hepatic dome who underwent percutaneous biopsy by MR guidance using a 1.0-T open MR scanner with free-hand technique. The mean lesions' diameter was 1.8±0.5 cm (maximum diameter range=0.6-3.6). Thirty-one lesions were in the right lobe and 5 in the left lobe. The 36 lesions were divided into two groups on the basis of their size: lesions with a maximum diameter ≤1.5 cm (n=14) and lesions >1.5 cm (n=22). All patients had a clinical suspicion of hepatic malignancy that needed to be confirmed prior to decisions on further clinical intervention, such as surgery, local ablative therapy or systemic chemotherapy. All lesions were located in the hepatic dome with poor visibility at US or non-enhanced CT or with negative results of previous biopsy procedures by US or CT guidance. Patients were excluded due to various factors, such as severe bleeding diathesis and contraindications to MRI.
Equipment and biopsy technique. A 1.0-T open-MRI imager (Panorama HFO; Philips Healthcare, Best, The Netherlands) was used to guide and monitor the percutaneous trans-hepatic needle biopsy, which has a maximum gradient strength of 26 mT/m and a slew rate of 80 T/m/s. A transmit-receive-type flexible surface loop (30 cm in diameter) was placed in the liver region over the area of interest for intraoperative imaging. The coil allows an open approach to the liver. An in-room radiofrequency-shielded liquid crystal monitor (Philips Healthcare) is placed at the side of the magnet, being used for image viewing. We used a coaxial technique in all biopsies. A coaxial system with a 15 or 10 cm, 16-gauge MR-compatible coaxial needle (Wanlin Medical, Qingdao, Shandong, China) and an 18-gauge semi-automated biopsy needle with a length of 22 or 17 cm (TSK TM; TSK Laboratory, Tochigi-shi, Japan) were used to obtain core specimens. All punctures were performed by the same two experienced interventional radiologists, each of whom had at least 6 years of experience in performing interventional MRI in a 0.23T low-field scanner.
Preoperative preparation. In order to confirm and evaluate the lesion size, location and vital structures located in the designed puncturing path or adjacent to the target lesion, such as portal vein, bile duct and diaphragm, contrast-enhanced CT or MRI was obtained before the biopsy procedure (Figure 1A). The biopsy approach was planned to avoid the lung and the above-mentioned important structures. All patients were put on the table of the MR system in a supine decubitus position. All of the entrance points were designed to go through the 1-2 intercostal spaces below the costo-diaphragmatic recess in the coronal image in order to avoid lung injury. Vital signs were continuously monitored by a MR-compatible patient monitoring system (MRGLIFE C Plus™; Schiller Medical, Affoltern, Switzerland) during the biopsy procedure.
Biopsy procedure. In order to confirm the target lesion, a T1-weighted (T1-TFE, TR 8.6 ms, TE 6.9 ms, flip angle 70°, slice thickness/separation 5 mm/1 mm, field of view 375×303, matrix 208×151, breath hold, acquisition time 15.6 s) or T2-weighted (T2-TSE, TR 2,286 ms, TE 90 ms, flip angle 90°, slice thickness/separation 5 mm/1 mm, field of view 400×303, matrix 248×143, respiratory compensation trigger, acquisition time 396 s) three-dimensional gradient field echo sequence was used to locate the initial axial images of the liver lesions and confirm the needle direction and position in the procedure (Figure 1B). After that, the fingertip in conjunction with the target lesions centre were used to define a arbitrarily angulated plane during the MR scan, which showed the entry point and the target lesion in the transverse view image (Figure 1C). Then we adjusted another plane being perpendicular to the above plane, following the pathway of the needle in the coronal axis, also showing the fingertip point and the center of the nodules (Figure 1D). The intersection of the two planes is considered to be the actual entry point. The approach distance (a horizontal line going through the entry point and from the skin entry point to the lesion) in the two planes were measured and taken as the reference index for inserting the biopsy needle. The patient was then moved out of the MR magnet after confirmation of the entry point. After the procedure area was disinfected, all patients were given local anesthesia with 1% lidocaine subcutaneously. A MR fluoroscopy sequence (T1-FFE, TR 10 ms, TE 6.0 ms, flip angle 35°, slice thickness/separation 8 mm/-1 mm, field of view 350×350, matrix 176×146, no breath hold or respiratory compensation trigger, acquisition time 1.6 s) was used to guide the puncturing of the liver lesions in the procedure. The needle was placed with continuous guidance imaging both in the axial and the perpendicular coronal plane after moving patients back into the magnet, so it can display deviations of the needle both in the axial and coronal orientations. After adjustment of the biopsy needle orientation while puncturing the lesions, according to the continuous imaging in the two orthogonal image planes, images were acquired with a frame rate of about 1 image per 1.6 second in this procedure (so called MR fluoroscopy) (Figure 1E and F). When the needle tip was located within the target lesion edge correctly, moving out the operation table from the center of the MRI scanner and the needle biopsy system was fired. Two to four specimens were taken from each lesion until sharply cut tissue cores at least 1.5 cm in total specimen length were considered as adequate sampling of the lesion specimen (3, 9). Additional biopsies were performed depending on whether the sample quantity was considered sufficient for diagnosis by the radiologists. No cytopathologist was on-site during the procedure. The needle was removed in the situation that the specimen quantity was considered to be sufficient for diagnosis and then the procedure was finished. The average time of puncturing the target lesion (from skin entry to the lesion) and the total procedural time (from getting the first image to the withdrawal of the biopsy needle) were recorded. After the biopsy, fat suppressed T2W-turbo spin echo (TSE) sequences (TR/TE of 1,600 and 110 ms, respectively) in axial and coronal orientation with a slice thickness of 5 mm were performed to confirm whether there is post-interventional hematoma or biloma or not. The core specimens were fixed in 10% formalin for pathological diagnostic examination. Immunohistochemical examinations were applied in some cases when needed based on the pathologist's opinion.
Case of a 62-year-old man who underwent magnetic resonance imaging (MRI)-guided percutaneous liver biopsy. A: Contrast-enhanced transverse computed tomography (CT) scan obtained before biopsy showed a 2.5 cm-diameter liver lesion (arrow) in the hepatic dome of liver segment 8. B: A T2-weighted magnetic resonance (MR) image was obtained to locate the initial axial images of the liver lesion (arrow). C, D: The fingertip shows the entrance point and the lesion (arrow) in the transverse view image and coronal view image. E, F: During the MR fluoroscopy procedure, puncturing the lesion in the transverse view image and the coronal view image. G, H: Transverse and coronal view images of a fast diagnostic sequence, including the whole needle trajectory, show the needle tip inserted into the lesion, thus permitting the measurement of the distance from the needle tip to the diaphragm. The biopsy specimen revealed that colon cancer metastasized to the liver.
Post-procedural evaluation. All patients were hospitalized. After the biopsy procedure, the patients were monitored for 48-72 h for potential post-biopsy complications. All patients were treated conservatively with monitoring vital signs and given hemostatics for 1-2 days. We routinely give patients MR imaging 2 days after liver biopsy and obtain hematocrit levels one day after biopsy to check whether there is subclinical bleeding or not. If a large or rapidly clinically important hemorrhage was found, we gave more hemostatics (when hemoglobin content decreased more than 10 g/l but less than 30 g/l) or hepatic arteriography if necessary (when hemoglobin decreased more than 30 g/l). All signs of procedure-related complications were classified as minor or major according to the Society of Interventional Radiology Clinical Practice Guidelines' criteria (10).
Statistical analysis. If the acquisition of a tissue sample was sufficient for pathological analysis, it was defined as technical success. The specific histological types of percutaneous biopsy were recorded. The final diagnosis was established by surgical nodules' pathology or clinical and imaging follow-up for at least 12 months. The correct malignant results from liver biopsy were confirmed if evidence of malignancy was found at surgical pathology or the histological findings in the samples were compatible with the known primary malignant tumor and if nodule(s) increased in size during the follow-up protocol. If no malignant tumor was identified at examination both in the specimen and surgical histology or there was no lesion growth found on the subsequent follow-up CT or MRI for more than 12 months without any anti-tumor treatment, it was considered to be the correct benign result. Diagnostic accuracy, sensitivity, specificity, positive predictive value and negative predictive value were calculated by comparing percutaneous biopsy histological diagnoses with final diagnoses obtained in surgery or clinical follow-up course. We used Fisher's exact test to compare accuracy, sensitivity and specificity in the two groups (nodules with diameter ≤1.5 cm and nodules with diameter >1.5 cm). Statistical analysis was performed by using SPSS 17.0 statistical software (SPSS, Chicago, IL, USA). The statistical test was two-sided and p<0.05 was considered to indicate significant differences.
Characterization of specific lesion type, as provided by histopathological examination.
Results
In all patients, the percutaneous biopsy procedure was technically successful with only one puncture and tolerated by all patients. All samples obtained were sufficient for histological evaluations for all 36 target lesions, whereas additional immunohistochemical analyses were conducted in 8 specimens. The only adverse effect that occurred in 25 patients was puncture site discomfort or tolerable mild pain, which subsided after a few hours (range=1-4.5 h). Three patients had a negligible self-limited peri-hepatic hemorrhage without need of additional treatment. No tumor seeding or diaphragmatic injury occurred. There were no serious procedure-related complications. The average procedure time for each puncture of the needle from skin entry to the lesion was 1.1 min, while the total procedural time (from the first image acquisition to the removal of the needle) was 28.5 min (range=21-47 min). The results of MRI-guided percutaneous liver biopsy revealed 29 (29/36, 80.6%) malignant and 7 (7/36, 19.4%) benign lesions. The final diagnoses, according to independent surgical histopathological findings (n=26) or clinical follow-up (n=10), were 30 malignant and 6 benign lesions. These results are shown in Table I, whereas results from each group, based on nodule size, are shown in Table II.
In the 35 (29 malignant and 6 benign) cases, surgical pathology (n=26) was consistent with biopsy results. Nine (4 malignant and 5 benign) cases were confirmed by clinical follow-up. One nodule that biopsy classified as a cirrhotic nodule increased in size as evidenced at the-ninth month follow-up, accompanied by alpha-fetoprotein (AFP) expression (827 ng/ml) and was subsequently diagnosed as hepatocellular carcinoma. The diagnostic performance of MR-guided percutaneous biopsy of hepatic dome lesions using a 1.0-T open high-field MRI scanner with free-hand and MR fluoroscopy technique in diagnosing malignant tumours was as follows: sensitivity 96.7% (29/30), specificity 100% (6/6), accuracy 97.2% (35/36), positive predictive value of 100% (29/29) and negative predictive value of 85.7% (6/7), whereas sensitivity, specificity and accuracy were 90.9% (10/11), 100% (3/3) and 92.9% (13/14), respectively, for lesions 1.5 cm or smaller and 100% (19/19), 100% (3/3) and 100% (22/22), respectively, for lesions larger than 1.5 cm. There was no significant difference in accuracy, sensitivity and specificity in the two groups (all p>0.05 by Fisher's exact test) (Table III).
Results of each group based on nodule size.
Accuracy of magnetic resonance imaging (MRI)-guided percutaneous liver biopsy in the two groups.
Discussion
Percutaneous US-guided biopsy is a well-established and widely used method for histopathological diagnosis of focal liver lesions. However, US requires greater operator skills and experience, as well as a lower intrinsic contrast than CT or MR (11). In addition, lesions situated in the hepatic dome are more difficult for US because the lesions are often only partially visible owing to acoustic shadowing from the lung base and ribs even in patients on deep respiration state (6, 12).
CT-guidance has been widely used in liver biopsy. However, CT exposes patients to ionizing radiation and many lesions identified with use of iodinated contrast material are not visible on non-contrast images. As noted in Sainani et al. patient series (13), approximately 20% of focal liver lesions biopsied under CT guidance were isoattenuating on common CT images. Biopsy was not attempted due to poor conspicuity of the isoattenuating lesions on CT for many of these lesions. Besides, when lesions were located in the hepatic dome, percutaneous trans-pulmonary CT-guided liver biopsy could induce pleural puncture and concomitant risk of pneumothorax, which is theoretically higher than for percutaneous lung biopsy because the needle traverses two pleural surfaces to reach the liver (14). To avoid this situation, CT-guided percutaneous trans-hepatic biopsy was used. However, CT does not allow for real-time surveillance of the puncture and a puncture route from the bottom up depends on the operator's experience because it needs a manual angle evaluation in regular CT-guided procedures (5, 15).
In recent years, MRI is considered to be a perfect alternative to CT and US for guidance. The potential and actual excellent clinical results of MRI guidance have been successfully established in percutaneous biopsy and also shown some advantages in various lesions (8, 15-18). Among the different modalities of image guidance, MR imaging offers several advantages as monitoring of pinpoint attack: excellent soft-tissue contrast without contrast medium, good spatial resolution, multi-planar imaging capabilities, arbitrary slice selection, intrinsic flow sensitivity, near real-time imaging and no ionizing radiation involved. MR can display the lesions clearly due to the excellent soft-tissue contrast, which is very important because confident visualization and clear demarcation of the target lesion are necessary (19, 20). Biopsy of a lesion with poor conspicuity will render the procedure at-risk of mis-targeting or obtaining a too small sample volume for making a pathologic diagnosis, which leads to repeated biopsies of an invasive nature and additional costs in time and hospitalization expenses (20). Better still, the open device allows good access to the patients and has adequate space for interventional procedure. The open system makes puncture possible with the biopsy needle, while the near-real-time MR fluoroscopy goes on, which the conventional closed-bore MRI system cannot achieve because of the limited patient access and times required for removing the patients in and out of the MR operation table making it, thus, impractical for needle guidance under MR fluoroscopy (21).
In the MR-guided biopsy procedure, the images can be acquired with short scan times. In our study, images were acquired with a frame rate of approximately 1 image per 1.6 sec during the free-hand and MR fluoroscopy-guided puncture procedure as this approach provided the near-real-time guidance for the interventional radiologists to target the lesions. Therefore, our average procedure time for each pass of the needle from skin entry to the lesion was only 1.1 min, while the total time for MR-guided biopsy procedure observed was only on average 28.5 min. This improves the expenditure of time compared with reports for CT-guided liver lesions' biopsy procedures by approximately 30.4 min (22). Although the results show only a small difference between the two groups, we believe that the results are misrepresented because our MR procedures were performed in the lesions that are all located in the hepatic dome that are more difficult to puncture than those located in non-special positions. It is worth of mentioning that the average size of nodules in our group were smaller (22). The free-hand and MR fluoroscopy technology offered a more precise navigation beyond the decreased procedure time. With our technique, we could visually follow the needle tip while puncturing based on the traverse and coronal orientation and adjusting the direction and angle of the needle in both orientations. Benefitting from this technology, we could accomplish the biopsy procedure excellently within a short procedure time.
To the best of our knowledge, this study is the first evaluation on the diagnostic accuracy and feasibility of 1.0T open-MRI-guided percutaneous trans-hepatic biopsy in focal liver lesions located in the hepatic dome. In this study, the diagnostic accuracy of percutaneous liver biopsy was 97.2%. This is comparable to previous studies of US- and CT-guided percutaneous liver biopsy in which diagnostic accuracy ranged from 61-99% (1-3, 12-13, 22). The target site must be selected in the suitable location to ensure obtaining representative and sufficient sample for diagnosis. The needle path of sampling should not pass through the necrotic zone of the lesion, which can lead to false-negative results (23). Small hepatic lesions (≤1.5 cm) are more challenging to target and biopsy as one might expect higher miss rates in small versus large lesions (1-2). However, in this study, diagnostic accuracy was satisfactory both in big hepatic lesions (>1.5 cm) and smaller nodules with a diameter of ≤1.5 cm (92.9%), while there was no significant difference in accuracy between the two groups. This satisfactory result may be due to the precise navigation of the MR free-hand and fluoroscopy-guided puncture procedure. First of all, the capability of visualizing the target nodule in both transverse and coronal image planes makes it easier and more precise to target the nodule and analyze the relationship of the needle tip and the lesion during the biopsy procedure. This benefit is most relevant in the coronal orientation to visualize respiratory motion. Therefore, variation in depth of respiration had little effect because both the lesion and the needle trajectory could be depicted continuously in a single plane, largely independent of the craniocaudal direction of diaphragmatic motion (24). Secondly, the procedure visualization under MR fluoroscopy was similar to the US-guided biopsy process making it particularly advantageous when compared to the need for manual angle evaluation in regular CT-guided procedures. The MR free-hand and fluoroscopy can be considered in an equal manner of performing US-guided or CT fluoroscopy-guided liver lesion biopsy without the use of ionizing radiation while providing more clear image quality than US and CT fluoroscopy.
Only three patients exhibited a minor bleeding complication in the liver induced by biopsy; however, all were self-healed without need of additional treatment. No tumor seeding or diaphragmatic injury occurred. None of the patients developed clinically important hemorrhage, which may be attributed to our strict adherence to pre-established coagulation criteria and the intrinsic flow sensitivity of MR to display the vessels clearly for the operators to avoid puncturing, along with the experience of our operators. There was no tumor seeding that could be due to the few needle passes for biopsy specimens (range=2-4 times, average=2.6). Of note, there was no diaphragmatic injury due to the arbitrary slice selection ability of MR. Therefore, the imaging plane, including the whole needle trajectory, can be adjusted so that the distance from the needle tip to the diaphragm, in this imaging plane, can be measured (Figure1G and H). Only in the situation that the distance from the needle tip to the diaphragm was farther than the length of specimen to be obtained, the needle biopsy system was fired. This occasion called for no diaphragmatic injury. Besides, there was no complication of pneumothorax because all r biopsy trajectories shrank away from puncturing the lung.
There are, however, several limitations to this study. First, the number of patients is limited; therefore, statistical analysis might have been biased. Second, the study had a relatively short overall follow-up period and not all results were confirmed by surgical histopathology (there were 10 lesions that required clinical follow-ups), which limits assessment of accuracy. Another disadvantage is the strong susceptibility artifact of the biopsy needle that might obscure the small target lesion and, thus, induce an inaccurate biopsy. Such a possibility may explain that, in our study, there was one nodule that biopsy considered as cirrhotic but, finally, diagnosed as hepatocellular carcinoma in the follow-up.
In summary, for biopsies of focal hepatic lesions located in the hepatic dome, MR-guided with free-hand and fluoroscopy technology appears to be an accurate and feasible approach and considered to be an excellent alternative to CT and US both in patients with lesions >1.5 cm and/or ≤1.5 cm in diameter.
Acknowledgements
The data analyzed in this study were provided by the Department of Interventional MRI, Shandong Medical Imaging Research Institute affiliated to Shandong University, People's Republic of China. The scientific guarantor of this publication is Professor Chengli Li. This study received funding from the Shandong Science and Technology development plan (2014GGH218005). All Authors kindly provided constructive advice and help for this manuscript. Written informed consent was obtained from all patients who participated in this study.
Footnotes
↵* These Authors contributed equally to this study.
- Received May 3, 2017.
- Revision received May 25, 2017.
- Accepted May 26, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.