Early-stage lung cancer is an important indication for pulmonary resection. To achieve adequate resection of a cancerous lesion, it is crucial that sufficient resection margins are respected. Multiple studies demonstrated that both lobectomy and segmentectomy are able to provide adequate resection margins, although the chance of local recurrence is increased in cases of segmentectomy with a margin to tumor size ratio <1 cm or if tumor size exceeded 2 cm (2, 3, 12, 13). In preoperative planning, securing an adequate resection margin is an important consideration when deciding on the operative approach and surgical planning. In case of segmentectomy, it is critical to identify the intersegmental veins as the boundary lines of the pulmonary segments and to identify the target segmental bronchi as the vertical surgical margin.

To date, many surgeons use traditional 2D CT imaging to estimate the resectability of a lesion and decide on operative approach. However, 3D-imaging using CT-based simulation techniques, sometimes assisted by artificial intelligence (14), can be used to accurately identify individual segments by tracing bronchovascular anatomy (15, 16), making it possible to estimate the probability that a cancerous lesion in a single segment has safe anatomical margins for segmentectomy, or needs change of the original surgical plan to another resection (for instance: bi-segmentectomy, lobectomy) if surgical margins are deemed insufficient (17–21). Comparative studies support this by describing a lower percentage of inadequate resection margins when preoperative assessment was conducted using 3D reconstruction instead of conventional 2D-imaging (22, 23). Evidence by Bakhuis et al. (24) supports this. They describe their experience with 3D reconstructions for preoperative planning in 50 patients with an indication for pulmonary segmentectomy and reported an adjustment of the preoperative plan that ensured radical resection in 52% of patients based on the acquired 3D reconstructions. They also found that in 14% of patients the tumor was localized in a different segment than assessed by conventional 2D-imaging. Nakamoto et al. (25) demonstrated that it was also possible to identify candidates for smaller, so-called “superselective” segmentectomy for diagnostic resection of small nodules located at a depth greater than 20 mm from the pleural surface. Using 3D-reconstruction, they were able to adequately identifying surgical planes based on arterial branching patterns up to the fifth-order that still satisfied the requirements for sufficient surgical margin in indeterminate small and deep pulmonary nodules, resulting in low volumes of resected parenchyma and thus preservation of pulmonary function.

The clinical efficacy and safety of preoperative 3D reconstruction in assessing resectability prior to anatomical sublobar resection is demonstrated by a recent study conducted by Hamada et al. (26) They showed a total 5-year overall survival, lung cancer-specific overall survival, and recurrence-free survival rates after VATS sublobar resection based on preoperative 3D reconstruction for surgical planning of 92.5%, 100% and 98.2% respectively in patients with stage IA non-small cell lung cancer. These results suggest that sublobar resection under 3D-reconstructive guidance achieves sufficient surgical margin for satisfactory long-term outcome.

Additionally, preoperative knowledge regarding the histologic origin of a specific tumor, if available, may also influence surgical decision-making. In most instances, preoperative histological specimens are required to tissue-type a tumor and decide on the most fitting surgical resection. In this regard, 3D-reconstructive techniques might also be of added value. A recent study used 3D reconstruction to identify hallmarks suggestive of malignancy, such as the presence of a solid component, lobulation, vascular convergence, or an air cavity density, as well as nodule size (27). These results highlight the potential benefit of preoperative 3D reconstruction in estimating the origin of ground glass nodules, thereby facilitating adequate surgical planning.

Pulmonary bronchovascular patterns are diverse and constitute complex spatial configuration. Surgeons must be familiar with both the general and patient-specific anatomy of pulmonary bronchovascular structures to conduct anatomical (sublobar) resections safely and adequately. Multiple studies showed that by using preoperative 3D reconstruction it was possible to accurately identify up to 98% of surgically relevant and potentially complicating pulmonary artery branches, including anomalous or uncommon variations in vascular branching patterns. Thereby, 3D reconstruction facilitates surgical resections by avoiding blind dissection of vessels and increases surgeons confidence in their knowledge of the spatial relationship between the lesion and adjacent anatomical structures in the individual patient (18, 19, 28–43).

Adequate resectability assessment and enhanced identification of potentially complicating anatomical variants suggest improved perioperative outcome when 3D-reconstructive techniques are used in the preoperative work-up. Results from multiple, primarily retrospective, comparative studies in patients operated on for a variety of pulmonary diseases, confirm this hypothesis, although the quality of evidence is debatable due to their retrospective nature (Table 1). In these studies, 3D-imaging is associated with shorter operating times (22, 36, 37, 44–46), more extensive mediastinal lymph node dissection (23, 36), a reduction in stapler reloads (44), reduced air loss on postoperative days 1–3 (44) and decreased intraoperative blood loss (36). No significant difference in other surgery-related parameters was found, such as length of hospital-stay, overall recovery, incidence of other postoperative complications and residual pulmonary function (22, 23, 36, 44, 45). Furthermore, there is evidence from multiple studies that preoperative 3D-imaging aided in selection of the appropriate surgical entry site tailored to the individual patient by taking their specific anatomy into account, resulting in less perioperative switching in surgical approach and lower conversion rates (36, 47, 48).

When comparing 3D-imaging to more other reconstructive techniques, such as 3D-printing, there is evidence that the use of 3D-printing has additional benefits regarding reduction of operative time, perioperative blood loss and conversion rates when compared to both 3D-imaging and conventional 2D-imaging, especially in more complex cases (45, 48).

Besides its role in preoperative assessment, the use of 3D-reconstructive techniques during surgery has also been reported. Examples include the use of a tablet during surgery to display preoperatively constructed 3D images, with specific focus on pulmonary bronchovascular anatomy related to the target lesion. Several authors report that the reconstructions displayed on the tablet helped to identify important anatomic structures and their spatial relationship, while being able to review and manipulate them during surgery. This significantly benefitted their intraoperative understanding of the relevant surgical anatomy and enhanced their ability to safely perform anatomical resections (49, 50).

Another example includes binocular stereo-navigation using 3D polarized glasses which displayed 3D-reconstructed images during surgery, to be used as an intraoperative navigation tool during VATS. This enabled easier and accurate identification of relevant bronchovascular structures, while achieving sufficient resection margins (51).

Chen et al. (48) used 3D-printed models to effectively aid surgeons in intraoperative identification of relevant surgical landmarks and concluded that 3D-printed pulmonary models enable easier localization of target lesions, selection of important anatomical landmarks as reference points and identification anatomical variations during the procedure from multiple angles. Qiu et al. (45) found that 3D-printed models displayed during surgery reduced perioperative blood loss by preventing vascular trauma in patients undergoing VATS anatomical partial-lobectomy for stage I lung cancer.

The high-quality 3D reconstructions rendered by modern CT-imaging techniques might also possess educational value in obtaining informed consent from patients regarding the surgical procedure they are subjected to. Also, evidence suggests it might improve surgical residency training (37, 45, 52, 53). The latter was investigated by Zhang et al. (46), who found evidence of faster learning of complex VATS procedures in surgical residents that used 3D reconstructions in their surgical preparation, compared to older surgeons that did not have these sophisticated techniques at their disposal, although age might have confounded these results. There was also a tendency for shorter operating times when preoperative resection simulation was applied using 3D-reconstructed anatomical models.