Pancreaticoduodenectomy (PD) is a standard treatment for malignant tumor of periampullary and pancreas head. As lymphatics (lymph node and lymph vessels) accompany the arteries and are distributed in the surrounding neural plexuses, complete clearance of peripancreatic tissue, including lymphatics and nerve plexus, is necessary for curative resection of the tumor.^1–4 As this operation is considered a complex procedure, a surgeon is required to be well trained in this specific surgical technique and to possess sufficient anatomic knowledge.

Despite a low mortality rate and improvements in perioperative care and operative management, there is still a relatively high complication rate following PD.^5,6 Several studies showed that intraoperative bleeding and red blood cell (RBC) transfusion are serious risk factors of postoperative complications in PD.^6,7 Recently, several procedures for artery-first approaches such as posterior, uncinated, and mesenteric approach have been introduced for improving perioperative outcomes such as curability and decreasing blood loss and morbidity.^8–11 Incidentally, it has been well known that early ligation of the inferior pancreaticoduodenal artery (IPDA)—one of the efferent arteries of the pancreas head—considerably reduces intraoperative bleeding and postoperative complications.^12–14 Owing to the various anatomic origins of IPDA, identification is difficult in some patients. Therefore, some groups have attempted to locate the origin of IPDA by preoperative enhanced multi-detector computed tomography (MD-CT) and 3-dimensional angiogram using MDCT data (3D-CT angiography).^12,13 In addition, an augmented reality technique using MD-CT data is being considered an innovative navigation system for PD.¹⁵ However, no simple intraoperative guidance system, which would greatly facilitate the complex procedure of vessel ligation and reduce intraoperative bleeding, has been tested for ligation of the IPDA during PD.

Intraoperative ultrasonography provides useful information for diagnosis and for guidance during the hepatobiliary-pancreatic surgery.^16,17 Color Doppler flow imaging facilitates to delineate the anatomy and to identify the vascular structures invading malignant tumors.^18–22 Recently, advanced navigation techniques have been introduced, such as 3D-CT angiography and intraoperative ultrasonography.^21–24 Doppler ultrasonography (Dop-US) has been used as an effective method for detecting the presence of potential bleeders.^25,26 However, Dop-US-assisted intraoperative identification and tracking of critical vessels for pancreatic surgery has not been reported to date.

The objective of the present study was to evaluate the potential of intraoperative Dop-US for detection of critical vessels relative to that of preoperative MD-CT, including MPR and 3D angiography, and to clarify the efficacy of vessel navigation surgery using Dop-US-guided tracking for the reduction of intraoperative bleeding.

Methods

Preoperative evaluation of the peripancreatic head vessels by MD-CT

A 16-detector-row MD-CT scanner (Aquilion; Toshiba Medical Systems, Tochigi, Japan) was used to obtain CT images. The scan delays from the administration of contrast material to the start of arterial, pancreatic parenchymal, and portal venous phases were fixed at 25, 43, and 70 seconds, respectively. Since 2005, 3D-CT angiography has been performed for preoperative evaluation of critical vessels around the SMA, such as the IPDA. The images were reformatted with 1-mm-thick multiplanar reconstruction (MPR) and 3D-CT angiograms and volume-rendered images of the pancreatic parenchyma on a workstation (Exavision; Ziosoft, Inc, Tokyo, Japan). The locations of the arteries were preoperatively analyzed using MPR and 3D imaging (Fig. 1a and 1b).

View Figure

• Download as Powerpoint
• Download Figure

Intraoperative vessel navigation using Dop-US

Intraoperative Dop-US was performed using the power Doppler mode with the ultrasonography probe connected to a high-end ultrasound scanner (Pro Sound SSD-5500; Hitachi Aloka Medical, Ltd, Tokyo, Japan) before resection of the neural plexuses of the pancreas head and those surrounding the SMA (Fig. 1c and 1d). First, we scanned the pancreas head along the SMA. Then, we identified all bifurcations of the pancreaticoduodenojejunal (PDJ) trunk and the IPDA and other branches. Next, early ligation of IPDA method was performed using intraoperative vessel navigation by Dop-US guidance (Dop-Navi). The procedures of this technique were as follows: (1) the operator identified the SMA and the PDJ trunk or the IPDA visualized by Dop-US; (2) real-time detection and tracking of the SMA and IPDA around the pancreas were performed with a bidirectional Dop-US blood flowmeter (ES-100V3; Hadeco, Kawasaki, Japan; Fig. 2a) and a thin probe (LRP08; Hadeco, Kawasaki, Japan; Fig. 2b); (3) the detected IPDA was exposed and ligated at the root of the bifurcation to reduce bleeding from the pancreatic parenchyma during the dissection of the connective tissue and nerve plexuses around the SMA. The blood flow data from the SMA and IPDA acquired with the Dop-US blood flowmeter were recorded for analysis (Fig. 3a and 3b).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Results

Patient characteristics

Comparisons of clinical preoperative parameters and the histological results of the control and Dop-Navi groups are shown in Table 1. As all patients had a tumor that was diagnosed as malignant or low malignant preoperatively, lymph node dissection was performed for all of them. However, the tumors of 12 patients were diagnosed as benign in final pathologic analysis. There was no statistically significant difference between the 2 groups with respect to body mass index, diabetes mellitus, preoperative biliary drainage, or histologic type.

Table 1 Characteristics of the enrolled patients

View Fullscreen

Assessment of intraoperative Dop-US-guided vessel detection

The information obtained intraoperatively about the pancreatic vessels by Dop-US was compared to the preoperative MD-CT scans analyzed by MPR and 3D angiography (Fig. 4a and 4b). First, there was a significant difference in the number of vessels detected by the 3 imaging modalities (ANOVA, P < 0.05). Dop-US (Tukey test; P < 0.0001), and MPR (Tukey test; P = 0.0022) detected significantly more vessels in the pancreatic head than 3D angiography (Fig. 4a). In addition, these techniques also differed in terms of mean diameter (ANOVA, P < 0.01), with significantly lower values obtained with Dop-US (Tukey test; P < 0.0001) and MPR (Tukey test; P = 0.0069), indicating higher definition (Fig. 4b). These data are consistent with the higher success rate of Dop-US and MPR in the identification of the IPDA within the pancreatic tissue. These data show that intraoperative Dop-US is as efficient as preoperative 3D-CT scans analyzed by MPR for visualizing of critical vessels around the SMA.

View Figure

• Download as Powerpoint
• Download Figure

Assessment of intraoperative Dop-US-guided vessel tracking

During surgery, identification of the IPDA is complicated by the surrounding nerve plexus tissue in the proximity of the SMA. Intraoperative real-time Dop-US revealed that the 2 vessels exhibit distinct bloodstream waveforms (Fig. 3a and 3b). As the IPDA had a markedly smaller waveform than the SMA, the flowmeter was able to trace the IPDA, which flowed into the parenchyma of the pancreatic head during the operation. The maximum blood flow velocity in IPDA was 3 times lower than that in the SMA (P < 0.005; Fig. 4c). In addition, the mean blood flow velocity in the IPDA was 2.5 times lower than that in the SMA (P < 0.005; Fig. 4d). These data demonstrate the ability of intraoperative real-time Dop-US to identify and track the vessels targeted for ligation during the PD procedure.

We also compared the perioperative data of the 2 groups (Table 2). The 2 groups were subjected to the same surgical procedures, in terms of PPPD/SSPPD, lymph node dissection level, and SMA plexus dissection frequencies. Only 2% (1/36) of the patient monitored using Dop-Navi method had residual tumor (R1) tissue after surgery, compared to 14% (5/34) in the control group patients who were only assessed by preoperative CT. The R1 site was the surgical edge of the bile duct in the hepatic side or the distal main pancreatic duct (data not shown). The operation time and volumes of RBC transfusions were less in the Dop-Navi group, but there were no significant differences. In contrast, the Dop-Navi group had a significantly reduced bleeding volume (1120 mL versus 1590 mL; P < 0.01).

Table 2 Perioperative status and postoperative complications in Dop-Navi surgical patients and control patients

View Fullscreen

No severe postoperative complications or deaths were observed in either group. There were no significant differences in the frequency or type of postoperative complications (Table 2). A Grade B pancreatic fistula occurred in 2 patients, both of whom recovered with conservative treatment.

Discussion

Complete resection of pancreatic head tumors by PD is a very complex surgical procedure involving the ligation of numerous vessels. The imaging techniques used by surgeons to navigate this network of vessels and nerves are limited to preoperative 3D-CT, which do not accommodate tissue displacement during the operation. While IPDA ligation reduces intraoperative bleeding, the root of the IPDA is not always identified on preoperative 3D-CT.^12,13 The present study demonstrates the potential of intraoperative critical vessel detection by Dop-US and the efficacy of real-time vessel navigation by Dop-US-guided tracking technique to facilitate IPDA identification and early ligation, while reducing intraoperative bleeding.

Recently, Dop-US has been successfully used to obtain detailed blood flow images for various organs and organ systems such as the cardiovascular system, kidney, hepatobiliary-pancreatic system, and spleen.^18,30 The Dop-US appearances of peripancreatic vessels have been well documented and are well visualized even if the examination is performed through the abdominal wall.²² As we were able to place the ultrasound probe on the surface of the pancreatic parenchyma directly during the operation, the acquired view of pancreatic tissue around the SMA was quite clear. Each artery detected by intraoperative Dop-US completely correlated with the MPR imaging and 3D-CT angiogram. Moreover, our method could also visualize small arteries with diameters of less than 2 mm. It was difficult to recognize these small arteries on 3D-CT angiogram. By using Dop-US, we were able to clearly visualize the PDJ trunk and the bifurcation of the IPDA in most patients without exposing the patient to ionizing radiation or contrast media. We determined that intraoperative Dop-US, in the power Doppler mode, was better than MD-CT for precise visualization of vessels.

It is well known that the IPDA usually branches on the dorsal side of the SMA and runs along the inside of the pancreatic nerve plexus tissue anatomically. It was not easy to use the conventional method to locate the IPDA in the operative field when the operation was performed with an anterior approach, despite having acquired preoperative imaging. The Doppler blood flowmeter can measure blood velocity during the operation.²² Ultrasonography-guided operations are performed in the fields of neurosurgery and urology.^26,31 Usually, it is not easy to recognize the arteries that run into the pancreatic nerve plexus around the SMA intraoperatively without palpation. A thin, rod-shaped probe enabled real-time detection of the precise location of the arteries, which were located in the neural plexus of the pancreas head. As our results show, the waveform and flow velocity data acquired using the Doppler blood flowmeter were quite different between the IPDA and SMA, which enabled the surgeon to quickly and easily recognize the location of both arteries.

To perform safe and effective resection of cancer of the pancreas head, it is very important to identify the vascular ramification patterns around the pancreas head in each patient. Kawai et al¹³ reported that early ligation of the IPDA using preoperative CT imaging before isolation of the pancreas head was useful for reducing intraoperative bleeding in PD. Horiguchi et al¹² showed that locating the IPDA preoperatively and intraoperatively with MD-CT imaging is a similarly effective approach to safely perform PD. However, MD-CT images acquired preoperatively cannot be easily transposed on the real-time view of the operative field. On the other hand, ultrasonography is a very useful and non-invasive technique that provides real-time information during dissection. As our procedures controlled the critical vessels such as IPDA and/or PDJ trunk prior to complete dissection of the connective tissue between pancreas and SMA safely, we considered that insidious bleeding from dissected pancreatic tissue was avoided during the operation. As PD consists of many complicated procedures, we tried to evaluate our technique simply by elimination of the patients who underwent additional invasive procedures such as portal vein resection or combined resection of other organs. However, the patients who enrolled in this study were not randomized in the Dop-Navi group or control group. In addition, it was not sufficient to elucidate the predominance of our technique because of the primary lesion diversity and the sample size. Unfortunately, it seems difficult to clarify the effect of this procedure, because this study did not take an accurate measurement of bleeding volume in each procedure particularly. It is necessary to demonstrate how the Dop-Navi method contributes to achieving a reduction of intraoperative bleeding volume by a prospective study.

It is well known that a strong advantage of ultrasonographic examination is noninvasiveness. Nevertheless, intraoperative Dop-US has been used only for the diagnosis of tumor spread or assessment of blood flow volume in the pancreas to date.^32,33 The Dop-Navi procedure is also a very simple method where the surgeon or surgical assistant only needs to put the probe on the pancreas tissue to directly hear the Doppler sound during the operation. Inflammation due to biliary drainage sometime causes chronic tissue change, which interferes with continuing the operation around bile duct. As there was no significant difference of biliary drainage rate between both groups, it was also thought to be equal to the influence of inflammation from tumor in both groups. The Dop-Navi group had shortened total operation time in spite of additional procedures. This examination did not contribute to an increase in the length of the procedure or increase operative complications. We demonstrated that this vessel navigation strategy is quite easy and useful for avoiding injury of critical arteries in the pancreatic field and for reducing operative bleeding volume without requiring palpation. Therefore, this technique might be utilized by less experienced surgeons or in advanced laparoscopic surgery.

In conclusion, the present study demonstrates that intraoperative navigation using Dop-US guided vessel detection and real-time tracking technique allows surgeons to clearly identify the IPDA during PD while avoiding injuries to major arteries. This simple and easy imaging technique contributes to improving the overall efficiency of pancreatic surgery by reducing bleeding volume without extension of the operative time.