Three-dimensional (3D) printing is the construction of a solid structure from a digital model using 3D printers which build the structure of a desired organ or anatomical segment by laying down many successive thin layers of building material. 3D-printed models are generated from digital images obtained by contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI). A software is required to obtain the 3D-printed model from the digital images. Several software programs have been developed for 3D printing, as the PLUTO software from the Nagoya University (Japan)¹, which is specific for the anatomic structure of the liver. The printed model provides information on the vascular and biliary structures, tumor arrangement, the area to be resected and the relationships between them, giving more accurate orientation on the anatomic landmarks for the surgical procedure.

Besides facilitating preoperative surgical planning, the model can be sterilized and transported to the operating room, thus improving the interpretation of the spatial relationship between the anatomic structures². The aim of this study is to report the initial experience with the use of 3D-printed models in 2 cases of liver surgery.

3D reconstruction: 3D reconstruction of CT images was performed using DICOM Osirix MD^® software program for image visualization and segmentation, and a contrast scan was performed to differentiate the structures of interest. The elements considered for surgical planning were the portal vein, hepatic veins, common hepatic artery and its branches, biliary ducts, tumor location and size, and the relationship between all these structures.

3D printing: different technologies were used for the 3D-printed model, depending on the surgical case and focused on the main point of interest or difficulty posed by the surgeon.

In the first case, a BCN3D Sigmax fused deposition model (FDM) printer (Spain) was used to construct a dual color model in 1:1 true scale. The liver parenchyma was printed in white and the structures of interest in a different color (Fig. 1 A).

Figure 1 A: Simulation and 3D-printed model. Case 1 B: Case 2. Reconstruction and 3D- printed model. 

In the second case, a FORM 2 (FormLabs, USA) printer with stereolithography technology was used to create a 3D-printed model with 80% of the liver full size; the structures were differentiated using different colors (Fig. 1 B). The material used in the models is called PLA, a thermoformable plastic that can be melted and recycled to obtain repeated models, which is beneficial for the care of the environment.

Use of the model: before the intervention, the surgical team examined the model, considering the main difficulties and the surgical approach they would use. In addition, the 3D-printed model was taken to the operating room during the intervention to provide the surgeon, who performed both interventions, with a better spatial representation of the structures. In both cases, intraoperative liver ultrasound was performed as an additional test to identify the vascular structures.

Case 1: The first case was a 35-year-old female patient with a history of inherited thrombocytopenia. She had undergone surgery for rectal adenocarcinoma with resection of the sigmoid colon and rectum followed by adjuvant chemotherapy. During follow-up, a CT scan identified a large liver metastasis in the right liver lobe and a small metastasis in segment III with stable disease. Therapy with irinotecan was indicated, with poor tolerance; the case was then evaluated by the Tumor Board where surgical treatment was decided. Right hepatectomy and limited resection of segment III was performed, achieving R0 resection (Fig. 2 A). The patient did not require intraoperative or postoperative transfusions of blood products and had favorable outcome. The CT scan and biomarkers were normal 6 months after surgery.

Figure 2 A: Computed tomography scan of the abdomen showing liver metastasis and surgical specimen B: Transection with indemnity of the right hepatic vein. Surgical specimen 

Case 2: The second case was a 60-year-old man with a history of hepatitis C, who complaint of long-lasting mild pain in the right hypochondriac region. The CT scan showed a mass in liver segments VI and VII. The laboratory tests showed increased alpha-fetoprotein levels, and, before performing endoscopic and imaging tests, a diagnosis of hepatocellular carcinoma was suspected. The patient was classified as Child class A, and surgical treatment was decided. Based on the 3D-printed model, resection of VI-VII segments was performed without requiring the use of blood products during surgery or in the postoperative period. The pathology report confirmed R0 resection (Fig. 2 B).

An Enhance Recovery After Surgery (ERAS) protocol was initiated and the patient was discharged without complications on postoperative day 3. At 9 months, alpha-fetoprotein levels and CT scan were normal, and the is currently developing his usual working activities. The implementation of 3D printing technology was very useful, as it facilitated the understanding of the spatial relationships among the anatomic structures in both cases. Understanding this spatial relationship allowed to effectively and safely solve situations that could not be predicted before. Igami et al. used 3D-printed models with intraoperative ultrasound in right minor hepatectomies with liver partition between the right anterior and posterior sectors and concluded that the procedure was easy and suitable¹. We also associated both methods in our cases, bearing in mind that they are non-mutually exclusive, as the sum of both methods helps in the decision-making process. These authors also applied this technology to liver surgery in two patients with liver metastases from colorectal cancer who underwent chemotherapy before surgery. The size of the metastases reduced so much that were not visible by preoperative and intraoperative ultrasonography. Based on the data provided by the 3D-printed model, resection was successful, with histologically negative surgical margins, demonstrating the feasibility of the procedure².

This technology has been implemented in the field of liver transplantation. Nizer et al. used 3D-printed models in three pairs of patients for living donor liver transplantation, emphasizing it is a valuable tool for understanding the spatial relationships between vascular and biliary anatomic structures, facilitating surgery, and minimizing intraoperative complications³. Laparoscopic liver surgery is another field of implementation, as this approach has experienced significant development and expansion over the past decade. The implementation of 3D-printing on decisionmaking in the context of laparoscopic liver resections has been reported by Witowski et al., among others, who concluded that these models are helpful for planning the extent of complex and major laparoscopic liver resections and for detecting candidates who may suffer from post-hepatectomy liver failure⁴.

The development of this tool is based on biomedical images acquired mainly by CT scan, although MRI is also useful. Despite the benefits demonstrated, there are some concerns, as the accuracy of 3D-printed models in relation with the actual anatomic structure, costs (which are generally high) and the time required to create and obtain the model. The environmental impact generated by the material used to make these models should be mentioned; however, in our case, 3D-printed models can be ground, melted and recycled to build new models, thus making the material environmentally friendly. In 2018, Witowski et al. evaluated the accuracy of 3D printed liver models developed by a cost-effective approach. They compared the tomographic images of the liver of 15 patients with the tomographic images of the 3D-printed models obtained from the initial CT scan and demonstrated the accuracy and similarity of the models developed by a cost-effective approach⁵. Other investigators have developed models that can be rapidly constructed with reduced costs and which can relate the vascular and biliary structures with the liver surfaces, facilitating a more accurate definition of the anatomic landmarks⁶.

In a review of 14 articles, the authors described inexpensive, low-quality 3D-printed models and other more expensive complex, multimaterial liver models with limited availability, suggesting that low-quality models might be useful for educational purposes for colleagues, patients and their families⁷.

Although we have used the models in only 2 patients, we consider that they were useful for planning and facilitating surgery, since they provided us with a very approximate map of the vascular landmarks that were displaced by the large lesions, which were also corroborated by intraoperative ultrasound. The time taken to obtain the model was about 7 days with a cost of 380 dollars including the digital model and the printed model. Although the time required was rapid, the costs in the current context of our activity must be considered high. The 3D-printed model had an accurate correlation with the liver structures, allowing for a reduction in the operative time, accurate identification of the anatomy and a reduction of bleeding events, especially in the case of the patient with inherited thrombocytopenia, which was evident by the absence of intraoperative and postoperative transfusion of blood products.

For the reasons mentioned above, and mainly because of what has been reported in the literature, we believe that the use of 3D-printing in liver surgery could be considered, mainly in complex and selected cases, evaluating its cost-benefit, the understanding it would provide for the preoperative period, surgical approach and accuracy during the intraoperative period, and as an educational tool for patients and colleagues. Finally, we must mention that we did not find national scientific publications on the use of 3D printing in liver surgery.