If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Correspondence: Kiyoshi Hasegawa, Hepato-Biliary-Pancreatic Surgery Division, Department of Surgery, Graduate School of Medicine, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-8655, Japan.
Real-time tissue elastography during open hepatectomy facilitates the differential diagnosis of liver tumors by providing information on elasticity. This study investigated the utility of intraoperative real-time tissue elastography (IORTE) during laparoscopic hepatectomy (LH).
Methods
Between 2012 and 2014, IORTE was performed during LH for 21 hepatocellular carcinomas (HCCs), 16 adenocarcinomas and 5 other tumors in 32 patients. The elasticity images were classified into six categories according to the modified criteria on the elasticity type of liver tumors, in which type 1 tumors show more strain than the surrounding liver and type 6 tumors no strain. The concordance of the IORTE findings with those of the pathological examination of the tumors was assessed (The registration no. 1418).
Results
Among the 21 HCCs, 20 were classified as “HCC pattern” (type 3, 4, or 5), resulting in a sensitivity of 95.2%, a specificity of 66.7% and an accuracy of 81.0%. Ten out of the 16 adenocarcinomas were classified as “adenocarcinoma pattern” (type 6), resulting in a sensitivity of 62.5%, a specificity of 92.3% and an accuracy of 81.0%.
Conclusion
IORTE is feasible and provides useful information on the elasticity of liver tumors in LH, in which conventional tumor palpation is difficult.
Introduction
Since the introduction of laparoscopic hepatectomy (LH) in the 1990s, this less invasive approach has gradually gained broad acceptance because of the development of advanced surgical technique and devices.
Previous meta-analyses have demonstrated that LH is a safe procedure. Its advantages over open surgery with respect to bleeding and postoperative hospital in appropriately selected patients has been reported.
However, the intraoperative examination of liver tumors is difficult in LH because conventional palpation is technically difficult, especially in the pure laparoscopic setting. Hata et al. reported that in 65 hepatectomies, 207 nodules of colorectal metastasis were newly found by classical palpation and/or inspection only.
Value of visual inspection, bimanual palpation, and intraoperative ultrasonography during hepatic resection for liver metastases of colorectal carcinoma.
During liver surgery, accurate information on the location and characteristics of tumors scheduled for removal is required. Fundamental B-mode intraoperative ultrasonography (IOUS) is a useful tool for the diagnosis of liver tumors.
and has increased the safety of laparoscopic liver resection, by allowing real-time identification of portal veins and hepatic vessels. However, both the limited mobility of the probe due to the trocar's position and the inflexibility of the probe are among the technical problems that remain to be solved.
Real-time tissue elastography provides visual information on tissue elasticity by estimating the strain modules based on the radio-frequency signals generated during external compression.
IORTE has also been used to characterize liver tumors and is able to distinguish adenocarcinomas from benign tumors based on the modified elasticity type of liver tumor (modETLT) criteria.
The aim of this prospective study was to evaluate the performances of IORTE using the newly devised mini-probes for the intra-operative diagnosis and characterization of liver tumors during LH.
Methods
Between January 2012 and December 2014, LH was performed in 37 patients at Tokyo University Hospital. Of these, 32 patients underwent IORTE during LH for liver tumors, concomitant with routine IOUS. Surgery and laparoscopic IORTE were performed after obtaining written informed consent from each patient. This prospective study was approved by the Research Ethics Committee of the Graduate School of Medicine and the Faculty of Medicine of The University of Tokyo (The registration no. 1418).
Preoperative imaging study, ultrasound system and surgical procedure
All patients were preoperatively diagnosed using contrast-enhanced computed tomography and ethoxybenzyl diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging.
IORTE was performed using the HIVISION Ascendus® ultrasound system and a 7.5-MHz laparoscopic probe (EUP-OL334; Hitachi, Ltd, Tokyo, Japan) (Fig. 1a). The elasticity images were color-scaled with respect to the number of strain modules, based on the radiofrequency signals during external compression. The tissue elasticity image was generated and analyzed in real time, yielding a color elastogram overlaid translucently on the B-mode image. The region of interest included both the tumor and the liver parenchyma. The images were obtained by cardiac beat. After insertion of the trocar and insufflation of the abdominal cavity, IORTE was performed concomitant with B-mode ultrasonography.
Figure 1(a) 7.5-MHz probe (Hitachi, Tokyo, Japan) using in IORTE during LH. (b) Modified elasticity type of liver tumor classification: type 1, more strain in most areas (the lesion had a mosaic pattern with a dominant red or yellow area); type 2, the lesion was homogeneously green; type 3, the lesion had a mosaic pattern with a dominant green area; type 4, the lesion had a mosaic pattern with a dominant blue area; type 5, the lesion was homogenously light blue; and type 6, the entire or partial lesion had no strain of uniform quality (the lesion was completely or partially deep blue, with no small areas of green.) (c) The elasticity image of HCC obtained during laparoscopic hepatectomy (LH) according to modified elasticity type of liver tumor (modETLT) criteria, the lesion had a mosaic pattern with dominant green area (white arrow head). Yellow arrow head showed the lesion of B-mode image. (d) The elasticity image of HCC, obtained during LH according to modETLT criteria, the lesion had a mosaic pattern with dominant blue area (white arrow head). Yellow arrow head showed the lesion of B-mode image. (e) The elasticity image of adenocarcinoma: the entire or partial lesion was deep blue (white arrow head). The margin of the tumor was unclear by B-mode intraoperative ultrasonography (IOUS). Yellow arrow head showed the lesion of B-mode image. (f) The elasticity image of adenocarcinoma overlaid on contrast enhanced-IOUS (CE-IOUS) (white arrow head). The margin was identified more clearly than by B-mode IOUS (yellow arrow head). (g) The elasticity image of small adenocarcinoma less than 1 cm, the entire lesion was deep blue (white arrow head). Yellow arrow head showed the lesion of B-mode image
(Fig. 1b). The ETLT criteria classify hepatic tumors into four categories, ranging from type A (normal strain) to typed D (no strain), according to the degree of strain compared to the normal surrounding liver. We specifically focused on the elasticity images generated from hepatocellular carcinomas (HCCs) and adenocarcinoma. The modETLT criteria assign hepatic tumors to one of six categories, ranging from type 1 (more strain than in the normal liver) to type 6 (no strain). A benign solid tumor is classified as type 1 or 2, HCC as type 3, 4 or 5 and adenocarcinoma as type 6.
Typical elasticity images of HCC are shown in Fig. 1c and d. The lesion in the figure has a mosaic pattern with a dominant blue or green area, indicative of HCC. Typical elasticity images of adenocarcinoma are shown in Fig. 1e–g. Either all or a portion of the lesion is deep blue, indicative of adenocarcinoma. The tumors were classified according to the ETLT and modETLT criteria independently by two investigators (YK and KO) blinded to the patient's clinical details and to the other imaging findings. Discordances between the investigators were resolved by consultation with the other co-authors (YK and JA). The concordance of the findings from the IORTE images with those from the pathological examination of the tumor was assessed.
Statistical analysis
Continuous values are expressed as median (range) and compared using the generalized Wilcoxon test and ANOVA. Categorical variables are expressed as number (%) and compared using Fisher exact test or the chi-square test, as appropriate. A P value less than 0.05 was considered statistically significant. All statistical analyses were conducted using JMP software (version 12.2.0; SAS Institute Inc., Cary, NC).
Results
The characteristics of the patients and their liver tumors are summarized in Table 1. In addition to HCC and adenocarcinoma, other malignant tumors were a metastatic tumor from a uterine carcinosarcoma and intrahepatic cholangiocarcinoma. The benign tumors consisted of lymphoma (n = 1), focal nodular hyperplasia (n = 1) and a regenerative nodule (n = 1), all of which had been diagnosed preoperatively as malignant. No patients with adenocarcinoma underwent preoperative chemotherapy.
Table 1Characteristics of solid lesions according to final diagnosis
The distributions of ETLT and modETLT elasticity types for all tumors are detailed in Table 2. Elasticity findings were obtained from all tumors included in this study. The sensitivity, specificity and accuracy of the IORTE findings based on the ETLT and modETLT criteria are shown in Table 3. Using the ETLT criteria, 12 of 21 HCCs were correctly classified as type B, and 20 of 21 tumors other than HCC as type A, C or D. According to these results, the sensitivity, specificity and accuracy of IORTE for the diagnosis of HCC were 57.1%, 95.2% and 76.2%, respectively. All 16 adenocarcinomas were correctly classified as type C or D and 16 of the 26 tumors that were not adenocarcinoma as type A or B. The sensitivity, specificity and accuracy of the ETLT criteria for the diagnosis of adenocarcinoma using the ETLT criteria were 100%, 61.5% and 76.2%, respectively.
Table 2Distribution of original and modified elasticity type of liver tumor classification types for all solid tumors in relation to final diagnosis
Table 3Sensitivity, specificity and accuracy of elastography for the diagnosis of benign solid tumors, hepatocellular carcinoma and adenocarcinoma, according to two classification systems
Sensitivity (%)
Specificity (%)
Accuracy (%)
ETLT
Type A as benign solid tumor
66.7 (2 of 3)
20.5 (8 of 39)
95.2 (40 of 42)
Type B as HCC
57.1 (12 of 21)
95.2 (20 of 21)
76.2 (32 of 42)
Type C or D as adenocarcinoma
100 (16 of 16)
61.5 (16 of 26)
76.2 (32 of 42)
ModETLT
Type 1 or 2 as benign solid tumor
66.7 (2 of 3)
97.4 (38 of 39)
95.2 (40 of 42)
Type 3, 4 or 5 as HCC
95.2 (20 of 21)
66.7 (14 of 21)
81.0 (34 of 42)
Type 6 as adenocarcinoma
62.5 (10 of 16)
92.3 (24 of 26)
81.0 (34 of 42)
Abbreviations: HCC, hepatocellular carcinoma; ETLT, elasticity type of liver tumor; ModETLT, modified elasticity type of liver tumor.
The elastographic patterns of 20 of the 21 HCCs were classified according to the modETLT as types 3–5, and those of 14 of the 21 non-HCC tumors as type 1, 2 or 6. Based on these results, the sensitivity, specificity and accuracy of the modETLT for the diagnosis of HCC were 95.2%, 66.7% and 81.0%. Ten out of 16 adenocarcinomas were classified as type 6, and 24 of the 26 tumors other than adenocarcinoma as types 1–5. The sensitivity, specificity and accuracy for the diagnosis of adenocarcinoma were 62.5%, 92.3% and 81.0%.
Discussion
Laparoscopic IORTE together with the modETLT criteria was able to correctly characterize liver tumors and the results compared favorably with those obtained from open hepatectomy IORTE.
This result suggested that laparoscopic IORTE has the same feasibility as open hepatectomy IORTE in the differential diagnosis of liver tumors. When used together with the modETLT criteria, laparoscopic IORTE provides surgeons with the same qualitative information on liver tumors obtained with manual palpation.
Manual palpation describes the elasticity of tumors and the non-tumoral liver parenchyma. Such manual palpation and tactile feedback are helpful for the surgeons to identify invisible tumors during open hepatectomy (OH). They also contribute to tailor the transection line to avoid tumor exposure and achieve R0 resection.
Value of visual inspection, bimanual palpation, and intraoperative ultrasonography during hepatic resection for liver metastases of colorectal carcinoma.
Several attempts have been made to obtain the same visual information on tumor elasticity during laparoscopic surgery. Ottermo et al. described a laparoscopic grasper to measure the pressure distribution generated by the deformation of tissues held between the jaws of the instrument.
Without palpation and tactile feedback, an identification of hepatic tumors during LH is more difficult than during OH. The diagnostic feasibility of IORTE during OH has been reported by detecting even non-visible and non-palpable tumors.
In this study, elasticity images could be obtained from all lesions using IORTE during LH and then classifying the tumors according to the modETLT criteria. Information on elasticity could therefore also be obtained from impalpable and invisible lesions. Correct placement of the trocar for use in IORTE provided an adequate view of the lesions scheduled for removal, including those in Couinaud's segment 1.
This advantage of IORTE would be demonstrated more clearly in LH, which lacks the manual palpation and tactile information.
The classification of adenocarcinoma was slightly worse in LH than in open hepatectomy. Six adenocarcinomas (37.5%) were miss-classified as type 4 or 5 (HCC pattern) but were identified as malignant according to the modETLT criteria. A small HCC has a homogenous internal pressure due to capsule formation whereas an adenocarcinoma, which lacks a capsule, shows a heterogeneous elasticity of its internal component and thus a relatively soft elastographic pattern. Some adenocarcinomas had a necrotic component with soft elasticity due to their very large size. However, the small number of tumors in this series and the different ratios of tumor size might have influenced the results. In addition, 8 out of 16 adenocarcinomas (50%) were located in the posterior sector of the liver, where it was difficult for the laparoscopic probe to maintain a stable pressure on the liver surface. Thus, good-quality IORTE images that allow tumor classification rely on techniques that allow imaging of the tumor directly under the probe and image acquisition during the cardiac beat. The latter may be more difficult in posterior sectors of the liver than in other sectors. Moreover, fibrosis of background liver has been reported to influence the color elastogram obtained during surgery.
When the liver parenchyma is cirrhotic, it is difficult to get the required color contrast between the hardness of the tumor and surrounding liver parenchyma.
Contrast-enhanced-IOUS (CE-IOUS) with Sonazoid® can be used to identify new lesions during open hepatectomy
and may also have applications in LH. Although the utility of CE-IOUS in this setting has been described in previous reports, the diagnostic capabilities of this technique were not examined in detail. In our study, CE-IOUS clearly visualized the edge of the tumor while IORTE provided additional elastic information on the features of the tumor (Fig. 1e,f).
A drawback of laparoscopic IORTE is that new hepatic lesions cannot be identified intraoperatively because the restricted flexibility of the laparoscopic probe makes it difficult to examine the whole liver.
Thus, elastography is unsuitable for total liver screening. However, in our cases, the small tumors (<1 cm) which B-mode IOUS unclearly detected were clearly visualized using IORTE. We could not detect a new lesion in this study. However, we have the cases which a new lesion was detected in open hepatectomy.
Laparoscopic IORTE might have the possibility small new lesions can be detected which preoperative image findings cannot visualize. In addition, to obtain an appropriate IORTE image, the region of interest should be set to include both the tumor and liver parenchyma. During OH, the flat shape of the linear probe made it easy to place effective pressure on the liver, leading to a more accurate acquisition of tissue elasticity images. The laparoscopic probe is not flat and releases an ultrasound wave from the lateral side of the tip. Therefore, it is more difficult using laparoscopic probe to acquire an accurate IORTE images than using linear probe. The range of motion of the laparoscopic probe is also limited by the trocar position. A smaller and flexible laparoscopic probe is therefore needed to resolve these technical difficulties.
A limitation of this study was that modETLT provides only a qualitative assessment whereas other intraoperative diagnostic devices used in LH, such as shear-wave elastography, are able to quantify the elasticity of liver tissue; however, reports on the intra-operative differential diagnosis of liver tumors are limited.
Moreover, the probe for shear-wave elastography is large, as it must resist the pressure of the shear wave. A smaller and flexible laparoscopic probe that provides a quantitative determination of the elasticity of liver tumors is therefore needed. Finally, no patients with adenocarcinoma underwent preoperative chemotherapy in this study. Therefore, our results could not demonstrate the influence of chemotherapy-associated-liver injury (e.g. sinusoidal obstruction syndrome, steatohepatitis). IORTE could potentially measure the elasticity of background liver and evaluate the degree of sinusoidal obstruction syndrome and steatosis in the future.
This is the first report to describe the capabilities of IORTE during LH. Because elastic information on liver tumors cannot be obtained during laparoscopic surgery, IORTE enables the “virtual palpation” of impalpable tumors and therefore provides clinically important information during LH. Although whole-liver screening and the detection of new lesions are still not possible with IORTE and better color contrast of the liver tissue is needed for the identification of small lesions, IORTE performs well in examining any area of the whole liver.
Acknowledgment
The authors thank Dr. Hiroki Kudo and Dr. Hiroko Okinaga (Hepato-Biliary-Pancreatic Surgery Division, Department of Surgery, Graduate School of Medicine, University of Tokyo) for acquisition of data and helpful insights.
Conflicts of interest
None to declare.
Appendix A. Supplementary data
The following is the supplementary data related to this article:
Supplementary Figure 1The color elastograms of the tumor and each type of ETLT and modified ETLT classification.
Value of visual inspection, bimanual palpation, and intraoperative ultrasonography during hepatic resection for liver metastases of colorectal carcinoma.
30921-8&id=gr1.jpg){kind=link}
30921-8&id=figs1.jpg){kind=link}