Eighteen patients [12 males and 6 females, with an average age of 46 (51 ± 12.9) years] were enrolled in this study. The inclusion criteria were as follows: (1) acute pancreatitis with pancreatic or peripancreatic necrosis accompanied by infection but without any invasive treatment; and (2) patients or their families accepting the evaluation of 3D visual reconstruction. The exclusion criteria were as follows: patients with acute pancreatitis without local complications or liquefied necrotic lesions without infection in a stable condition. According to the 2012 Atlanta classification of acute pancreatitis[9], 9 patients had moderately severe acute pancreatitis and the other 9 patients had severe acute pancreatitis. This study was approved by the ethics committee of the PLA General Hospital.

Data acquisition and 3D reconstruction: The 128-slice spiral computed tomography (CT) scanner (GE Corporation, Stamford, CT, United States) was used for obtaining abdominal enhanced CT scans of arterial and portal venous phases. The layer thickness was 1.5 mm, and the layer distance was 1.5 mm. Lipiodol was injected as the contrast medium (Beilu Pharmaceutical Co., Beijing, China); the concentration was 350 mg/mL. A high-pressure syringe was used for elbow vein injection. The dose was 60 mL, and the injection rate was 4 mL/s. The collected image data were stored in the form of Digital Imaging and Communications in Medicine and introduced into 3D visualization system (Mimics 17.0; Materialise Co., Leuven, Belgium) for reconstruction.

Anatomical evaluation and determination of position and paths of optimal puncture points: The reconstructed model could be viewed from any direction (magnified, contracted, rotated, or transparent) through the 3D visualization system. Therefore, the size, shape, position and number of the lesions and surrounding structure were intuitively demonstrated. Moreover, virtual operations could be done on the 3D image to observe different debridement ranges through various puncture points. At last, the optimal puncture number and paths were determined based on the following principles. First, retroperitoneal access was preferred to transabdominal access, which meant less intraperitoneal contamination. Second, the paths were established along the longitudinal axis of the necrotic cavity, and the puncture point should be as close to the necrotic cavity as possible, facilitating the maximum removal of necrosis. Third, multiple drains should be placed during the same procedure, if necessary, to avoid the visual blind area or operational blind area, which were also estimated using the 3D virtual system.

PCD and percutaneous nephroscopic necrosectomy for removing necrotic tissues: PCD was performed under the guidance of CT in strict accordance with the position and direction of puncture points designed by the 3D visualization system. Subsequently, an antiinflammatory drug was administered according to the drainage culture and the results of drug sensitivity test. The changes in the disease were assessed with a reexamination of abdominal CT scan weekly. If the condition did not improve obviously or continued to aggravate, the percutaneous nephroscopic necrosectomy was performed under general anesthesia. Briefly, a 1.2-cm skin incision was made that was centered on the PCD. Amplatz renal dilators (Cook Urological Incorporated, Bloomington, IN, United States) were used to serially dilate to create a 30F tract, following which a 12-mm trocar was inserted.

An operating nephroscope (Hopkins Telescopes; Karl Storz-Endoskope, Tuttingen, Germany) with an 8-mm working channel was then passed through the trocar into the necrotic cavity. Subsequently, the piecemeal removal of solid necrosis was performed repeatedly using a fenestrated grasper through the working channel. Finally, a 10F catheter sutured to a tube drain of 28F was placed into the distal end of the necrotic cavity to allow continuous lavage after surgery. All patients underwent CT scanning reexamination weekly to evaluate the results of necrosectomy and drain placement. The operation might be performed again, if necessary according to the changes in the condition. The position and range of the lesion, the time and number of surgery, postoperative complications, and the time of hospitalization of each patient were recorded.

The SPSS 17.0 software (SPSS Inc., Chicago, IL, United States) was used for statistical analysis. Measurement data were expressed as χ ± s.

The 3D reconstruction was successfully completed for all of the 18 patients. The results clearly showed the shape, size and number of necrotic lesions, as well as the anatomical relationship with surrounding blood vessels and organs. Thus, the stereoscopic visual observation of the lesions from any angle and virtual surgery using the aforementioned software system were performed. Further, the removal range and residual blind area of different puncture points were defined. Also, the individualized optimal position, direction and number of puncture paths suitable for patients were determined.

A typical case is given here to illustrate the use of 3D visualization technology: Patient No. 4, a 43-year-old male with alcoholic pancreatitis. The patient suffered from a high fever and abdominal pain for more than 2 wk till he was transferred to the hospital. The CT examination showed extensive peripancreatic necrosis in the retroperitoneal space. The results of 3D reconstruction clearly showed horseshoe-shaped necrosis, as the purple lesion shown in Figure 1A, involving the head of pancreas, uncinate process, root of mesentery, body and tail of pancreas, splenic hilus and left paracolic sulcus, which was closely related to surrounding organs and blood vessels. Virtual percutaneous nephroscopic necrosectomy was performed through the 3D virtual system to determine the three optimal puncture paths. As shown in Figure 1B, the necrosis in the blue area could be debrided from the right puncture path. As shown in Figure 1C, the necrosis in the green area could be debrided from the upper left puncture path. As shown in Figure 1D, the necrosis in the orange area could be debrided from the lower left puncture path. Residual necrosis was found without any single puncture path. Therefore, the three puncture paths were used simultaneously to avoid the operational blind area or residual necrosis. Figures 23 and 4 are the cross-section, coronal-section and 3D reconstruction images of puncture points on the right side, upper left side, and lower left side, respectively. These were critical to ensure the accuracy of PCD under the guidance of CT.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Three-dimensional visualized reconstruction image and virtual surgery for Patient No. 4. A: Horseshoe-shaped infected necrotic lesions (purple) adjacent to important organs and blood vessels; B: Debrided area of percutaneous nephroscopic necrosectomy through the point of puncture on the right puncture point (blue area); C: Debrided area of percutaneous nephroscopic necrosectomy through the point of puncture on the upper left puncture point (green area); D: Debrided area of percutaneous nephroscopic necrosectomy through the point of puncture on the lower left puncture point (orange area).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Cross-section, coronal-section and three-dimensional reconstruction images of the right-side puncture path for percutaneous catheter drainage in Patient No. 4. The red arrow represent the fictional direction and path of puncture.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Cross-section, coronal section and three-dimensional reconstruction images of the upper left-side puncture path for percutaneous catheter drainage in Patient No. 4. The red arrow represents the fictional direction and path of puncture.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Cross-section, coronal section and three-dimensional reconstruction images of the lower left-side puncture path for percutaneous catheter drainage in Patient No. 4. The red arrow represents the fictional direction and path of puncture.

The CT-guided PCD of the 18 patients was successfully performed strictly according to 3D visualization design, and subsequent percutaneous nephroscopic necrosectomy was carried out according to the changes in the condition, debriding maximum necrosis. Twelve patients were cured by conducting one-time surgery. Five patients were cured after two-time surgery and only one patient underwent the surgery three times. Patient No. 14 underwent additionally endoscopic necrosectomy for removing residual small lesions due to the winding sinus. The median operation time was 102 (102 ± 20.7) min, and the postoperative hospitalization time was 35 (35 ± 8.1) d. No major surgical complications occurred. All the patients were cured and discharged from the hospital eventually (Table 1).

The patient in the aforementioned typical case underwent percutaneous nephroscopic necrosectomy for removing necrotic tissues 14 d after PCD. A large number of infected necrotic tissues were removed from the three puncture points. The clinical condition of the patient improved obviously after surgery. Figure 5 shows the results of abdominal CT re-examination on the 15^th day after surgery. Finally, he was discharged from the hospital on the 29^th day postoperatively.

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Image of abdominal CT reexamination of Patient No. 4. Abdominal CT reexamination 15 d after percutaneous nephroscopic necrosectomy showed that the abscess cavity disappeared and the drainage tube was unobstructed. CT: Computed tomography.

Infected necrotizing pancreatitis (INP) is a severe disease with high mortality, which generally requires percutaneous catheter drainage (PCD) and following surgical debridement if necessary. Whether the puncture paths of PCD are appropriate or not is related to not only the effect of drainage but also the efficiency of subsequent minimally invasive removal of infected necrosis. However, a number of patients’ PCDs were insufficient or even the catheter passed through the hollow organs due to the doctors’ lack of experience, which was a disaster for INP patients.

Three-dimensional (3D) visualization technology has been proved to be of great help for precise intervention or surgery, which also might be useful to optimize the puncture paths of multiple PCDs for INP patients.

To explore the value of 3D visualization technology for PCDs in INP patients.

Preoperative computed tomography images were converted into 3D modellings through a software and the lesions were well displayed. PCD number and puncture paths were designed through virtual surgery (percutaneous nephroscopic necrosectomy) based on the principle of maximum removal of infected necrosis conveniently. We retrospectively analyzed 18 INP patients’ clinical data and present a typical case in detail.

All the patients’ 3D modellings was well reconstructed, through which the optimal PCD paths were designed. As a result, infected necrosis was conveniently removed in abundance using a nephroscope during the following surgery and two-thirds of the patients were cured after only one-time operation. Postoperative hospitalization time was 35 d on average, no major surgical complications occurred, and no one died.

3D visualization technology was useful for INP patients to maximize the PCD effect. Moreover, it significantly improved the efficiency of subsequent percutaneous nephroscopic necrosectomy, which was critically important for improving the prognosis.

Although the case number of this study was limited, the initial result indicated the value of 3D visualization technology for this terrible disease. Of course, analysis of more cases from multiple centers is needed to support the findings.