In this retrospective study, the ventilation image derived from 4D-CT was applied for evaluation of radiotherapy treatment plans in functional lung sparing. One way to compare functional lung sparing is to compare the MLD for total lung and highly-functional lung. Based on the cases studied, for small target volume cases, the MLD of total lung was always higher than that of highly-functional lung volumes (20% ~ 40%). The reason for this is that the volume around tumor is usually of low ventilation, and for small targets, radiation beams cover smaller normal lung volume, thus affect less highly-functional lung volume. However, the consistent results were not found in this study for large target volume cases. The number of cases that the MLD in highly-functional lung volumes was higher than that of the total lung was about the same as the number of cases that the MLD in highly-functional lung volumes was lower, since the beams to cover large target volumes would go through large normal lung volume, which inevitably include highly-functional lung volume. The difference between the MLD values in highly-functional lung volumes and in the total lung was not great. While Case 7 to 10 are typical examples, Case 11 is an exception. The big tumor in Case 11 was in the left lung. The radiation beams were arranged to avoid the right lung completely where most the high ventilation lung volume was in. Thus in this case, the MLD of total lung was obviously higher than that of the highly-functional lung volumes.
In treatment planning, clinicians always try to minimize the radiation dose to normal lungs to reduce toxicity, usually without functional lung information . The dose in highly-functional lung regions determines the risk of complication and thus should be an important index of treatment planning quality. This retrospective study investigates the dose in highly-functional lung volumes in the treatment plans developed without incorporation of the functional lung information. For small tumors, V5, V20 and MLD of highly-functional lung regions were mostly lower than the values of total lung. This implies that even without incorporation of functional lung information in IMRT treatment planning, the dose in highly-functional lung regions is usually low. For large tumors, it varies case by case. But in general, radiation treatment affects functional lung more when the target volume is large. From this point of view, it is more critical to incorporate functional lung information in treatment planning for patients with large tumors. It may be more difficult to achieve certain goals in functional lung sparing for large target volumes due to anatomical limitations. When the target is small, functional lung sparing should be easier to accomplish because a more flexible beam arrangement is possible. This study has demonstrated this trend (Figure 3).
With the highly-functional lung sparing constraints, all cases showed better V5 in the functional plans (Figure 3). Most functional plans demonstrated significant improvement in V5, V20 and MLD in the top 20% high ventilation volumes. Based on the MLD to total lung, there is only one functional plan in which the MLD to total lung was higher in the functional plans by 0.05 Gy. The reason for this was that the tumor was oblong in shape and close to the esophagus. When the objectives of proper PTV coverage and highly-functional lung sparing in the functional plans were satisfied, the MLD to the total lung got slightly worse.
In general, all functional plans generated with incorporation of 4D-CT based ventilation image for all thoracic cancer patients were successfully within constraints (Table 2) and acceptable for clinical use. Among the 11 cases studied, when compared with the anatomic plans, both better CI and HI in the functional plan were found in one case, better CI and superior HI in the functional plan were found in 1 and 9 cases respectively. More cases in the anatomic plans had better CI (10 out of 11) while more in the functional plans had better HI (9 out of 11). However, the differences are within a small range with ΔCI < 0.15 and ΔHI < 0.03. The CI and HI differences showed no strong relationship with the tumor size but more related to the shape and location relative to critical organs. Since there are more constraints in the functional planning, one would expect worse CI and HI in functional plans when compared with the anatomic plans. This trend is not clear for HI in this study.
The dose to the other OARs, such as heart, spinal cord and esophagus, were relatively higher in the functional plans than that in the anatomic plans. Since the doses to the other OARs in the functional plans were within the planning constraints, they were not forced to be better than the anatomic plans. Tighter doses to the other OARs could be achieved if tighter dose constraints to other OARs were made. However, since there are more planning constraints in the functional planning, some compromise has to be made, which could be the doses to the other OARs. If tighter doses to the other OARs are not achievable, one needs to evaluate the risks of complications to the lung and other OARs and make decision which plan to use, the anatomic or functional plan.
A recent study showed that the poor initial ventilation close to the tumor may improve with treatment induced tumor shrinkage . Although lower V20 and MLD are recommended to limit the risk of radiation pneumonitis in general , many other studies showed that the pulmonary injury and radiation dose had a weak correlation [17, 27, 28]. Hence, the hypothesis that sparing of the best functioning lung volumes is beneficial for the patients still needs to be verified by clinical data.