Clinical Investigation
Quantification of Motion of Different Thoracic Locations Using Four-Dimensional Computed Tomography: Implications for Radiotherapy Planning

https://doi.org/10.1016/j.ijrobp.2007.05.016Get rights and content

Purpose

To assess the respiratory motion of different thoracic nodal locations and its dependence on the presence of enlarged nodes; to assess the respiratory motion of different parenchymal tumor locations; and to determine the appropriate margins to cover the respiratory motion of targets at these locations.

Methods and Materials

We reviewed the four-dimensional computed tomography scans of 20 patients with thoracic tumors treated at our institution. The motion of four central thoracic locations (aortic arch, carina, and bilateral hila), parenchymal tumor locations (upper vs. lower, and anterior vs. middle vs. posterior thorax), and bilateral diaphragmatic domes was measured.

Results

For the central thoracic locations, the largest motion was in the superoinferior (SI) dimension (>5 mm for bilateral hila and carina, but <4 mm for aortic arch). No significant difference was found in the motion of these locations in the absence or presence of enlarged nodes. For parenchymal tumors, upper tumors exhibited smaller SI motion than did lower tumors (3.7 vs. 10.4 mm, p = 0.029). Similarly, anterior tumors exhibited smaller motion than did posterior tumors in both the SI (4.0 vs. 8.0 mm, p = 0.013) and lateral (2.8 vs. 4.6 mm, p = 0.045) directions. The margins that would be needed to encompass the respiratory motion of each of the evaluated locations in 95% of patients were tabulated and range from 3.4 to 37.2 mm, depending on the location and direction.

Conclusions

The results of our study have provided data for appropriate site-specific internal target volume expansion that could be useful in the absence of four-dimensional computed tomography-based treatment planning. However, generalizing the results from a small patient population requires discretion.

Introduction

Lung cancer remains the leading cause of cancer death despite improvements in diagnosis and more aggressive therapy, with a 5-year survival rate of 49% for localized disease and a dismal 16% for regional disease (1). In a Phase I radiation dose-escalation study using three-dimensional conformal radiotherapy, Rosenzweig et al. (2) reported a significant difference in local control and overall survival (23% vs. 52% at 2 years for Stage III) when stratified by radiation dose (<80 Gy vs. ≥80 Gy). This suggests that greater radiation doses, achieved through more conformal treatment delivery, could lead to improved outcomes. Ultimately, increasing the conformality of the radiation dose distributions will require accurately accounting for tumor motion.

In contrast to tumors at other anatomic sites, lung tumors generally move (0.4–3 cm) with breathing 3, 4, 5. However, helical computed tomography (CT) scans conventionally used for radiotherapy planning do not accurately represent this motion and, in fact, result in substantial artifacts when scanning a moving object 6, 7. In conventional treatment planning, adding a margin is used to compensate for tumor motion. In general, the same margins are used around the parenchymal tumor and the targeted nodal stations without knowledge of the site-specific motion characteristics. Excessive margins can lead to treatment complications and limit the dose that can be delivered to the target 8, 9, 10, 11.

Fluoroscopy has been used to evaluate organ motion; however, only selected structures are readily visualized with fluoroscopy 12, 13, 14. Several CT methods are available to evaluate organ motion more accurately, including four-dimensional CT (4D-CT) 15, 16, 17, breath-hold CT (18), and prospectively gated CT (19). These techniques make possible patient-specific respiratory motion-compensated radiotherapy, either by designing customized margins for target and organ motion or by respiratory-gated radiotherapy in which the treatment beam is activated only when the target is in a predetermined location.

Currently, few published data are available on the motion of different thoracic structures often included in the radiation portal (20), and none are available on those corresponding to nodal stations. Also, little is known about the influence that enlarged nodes might have on the respiratory-induced motion of these regions. One could hypothesize that enlarged nodes cause a mass effect that alters the motion of these regions. Most reports on lung tumor motion have focused on the movement of parenchymal tumors. Although some reports have agreed that tumors in the lower lobes tend to be most mobile, other studies have found no relationship between the degree of mobility and the anatomic tumor location 3, 5, 13, 21, 22.

In this study, we analyzed our recent experience with 4D-CT in radiotherapy planning for patients with thoracic malignancies to assess the respiratory motion of different thoracic nodal locations and its dependence on the presence of enlarged nodes; to assess the respiratory motion of different parenchymal tumor locations; and to determine the appropriate margins to cover the respiratory motion of targets at these locations.

Section snippets

Patient population

We identified 34 consecutive patients with thoracic tumors who underwent 4D-CT scanning for treatment planning in our department during a 1-year period. Of these, 20 patients who had regular breathing patterns were included in the analysis. The remaining 14 patients had irregular breathing patterns, despite the use of audio and/or video coaching to help regularize the breathing, resulting in 4D-CT scans with a high degree of motion artifact that prohibited accurate measurement of organ

Parenchymal tumor location

Two patients were excluded from the analysis of parenchymal tumor motion because one had a metastatic osteosarcoma and no parenchymal tumor, and one had a large invasive thymoma and a collapsed lung. One patient had two separate parenchymal tumors. Thus, 19 parenchymal tumors were analyzed. Of these 19, 13 were located in the upper thorax and 6 in the lower thorax; 5 were in the anterior thorax, 10 were in the coronal midplane, and 4 were in the posterior thorax; and 11 were in the left lung

Discussion

As radiotherapy becomes more conformal, the importance of understanding organ motion becomes increasingly important. Conventionally, adding margin is used to compensate for breathing-induced target motion. However, little information is available to guide appropriate margin design in a site-specific fashion. A study from The Netherlands demonstrated that adding a 1-cm uniform margin to Stage I non–small-cell lung cancer according to the findings from a single pretreatment CT scan is

Conclusions

The results of this study have added to our understanding of site-specific respiratory motion in the thorax and provided pertinent information to guide the design of the internal target volumes for different nodal and parenchymal locations for conventional radiotherapy for lung cancer in the absence of 4D-CT–based treatment planning. However, generalizing the results from a small patient population requires discretion.

References (24)

Cited by (36)

  • Feasibility of markerless 3D position monitoring of the central airways using kilovoltage projection images: Managing the risks of central lung stereotactic radiotherapy

    2018, Radiotherapy and Oncology
    Citation Excerpt :

    A discrepancy between the RPM signal and internal motion has been described before [20]. In some studies, airway motion and/or whether the carina position can be used as a surrogate for the tumor position has been investigated using 4D CT scans [21–26]. In addition, airway motion has been monitored in 2D using electronic portal imaging [27–29].

  • Anatomic optimization of lung tumor stereotactic ablative radiation therapy

    2015, Practical Radiation Oncology
    Citation Excerpt :

    A subset of patients received deep inspiration breath-hold (DIBH) computed tomography (CT) at simulation in addition to 4-dimensional CT (4DCT). To limit the number of scans to which patients were exposed, only patients clinically judged likely to have potentially large respiratory tumor motion and who demonstrated on physical examination the ability to perform a breath-hold maneuver received DIBH CT; usually, these patients had an inferior tumor location.6 Consecutive patients who had DIBH scans in addition to 4DCT acquired at simulation were to be included in this planned secondary analysis until the required sample size of treatment plans was accumulated.

  • Aortic arch vessel geometries and deformations in patients with thoracic aortic aneurysms and dissections

    2014, Journal of Vascular and Interventional Radiology
    Citation Excerpt :

    One reason our results reflect a greater range may be because we tracked translations of the BA, LCCA, and LSA ostia, whereas Weber et al tracked only the aortic lumen centroid just distal to the LSA. We report average respiratory-induced 3D arch translations of 1.5–17.9 mm, whereas Maxim et al (26) reported translations of 1.2–14.7 mm. We tracked three points on the aortic arch, whereas Weber et al (22) tracked only a single point just distal to the LSA.

  • Volumetric image guidance using carina vs spine as registration landmarks for conventionally fractionated lung radiotherapy

    2012, International Journal of Radiation Oncology Biology Physics
    Citation Excerpt :

    Primary lung tumor and mediastinal nodes are known to move during respiration (8). Furthermore, the carina is intimately related to the lung parenchyma and moves with respiration (8-10), as opposed to the spine, which is rigid. The carina is also in close proximity to nodal disease and centrally located tumors; and if the tumor involves a major airway, any change in tumor position (eg, new atelectasis or resolution of lung collapse) will likely be reflected by a change in carina position.

View all citing articles on Scopus

Presented at the 48th Annual Meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO), November 5–9, 2006, Philadelphia, PA.

Conflict of interest: none.

View full text