Predicting the Risk of Secondary Lung Malignancies Associated with Whole Breast Radiation Therapy

Introduction

The risk of secondary lung malignancy (SLM) is a significant concern for women treated with whole breast radiation therapy after breast conserving surgery for early stage breast cancer. This concern has gained prominence as the use of adjuvant breast radiation has increased and as prognosis has steadily improved over the past decade (1-4). Furthermore, the latency period of radiation-induced second cancers is often greater than 10 years, and the risk persists 30 to 40 years after therapy (5-8). As prospective trials may not be feasible and as epidemiologic studies would require extremely long follow-up, modeling approaches may provide more immediate insight into the magnitude of the potential risks. It would be particularly valuable to comparatively estimate the risk of secondary lung malignancies between two clinically acceptable whole breast irradiation techniques such as supine whole breast irradiation, the current standard adjuvant radiotherapy technique, and the clinically accepted alternative techniques of prone breast irradiation and hypofractionated breast irradiation (9-11). A comparative analysis of the risk estimates between these techniques may impact treatment decision-making.

In this study, a novel biologically-based mathematical model of spontaneous and radiation-induced carcinogenesis (12, 13) was used to quantitatively predict the lifetime absolute and relative risks of secondary lung malignancy for women who were planned for breast radiation treatment in both the standard supine position and the alternative prone position, using both the standard fractionation schedule and the alternative hypofractionated schedule.

Methods and Materials

Results

There was statistically significantly less radiation dose delivered to the ipsilateral lung for each of the fifteen patients in the prone position when compared to the supine position. For the standard fractionation schedule of 5000 cGy in 25 fractions, the average mean lung dose (MLD) was measured to be 54.2 cGy +/− 5.2 cGy (standard error (SE)) in the prone position vs. 645.5 cGy +/− 43.5 cGy in the supine position (p < 0.001, paired t-test). A similar difference in average MLD was noted between the supine and prone hypofractionated plans (548.7 cGy +/− 37.3 cGy and 46.1 cGy +/− 4.5 cGy, respectively, p<0.001).

The predicted lifetime absolute risks of lung cancer for each patient are shown as a scatter plot for treatment with standard fractionation in both the prone and supine positions in Figure 3. The mean predicted lifetime absolute risk of lung cancer for standard fractionation was 4.86% +/− 0.43% in the supine position and 1.99% +/− 0.18% in the prone position for a mean difference of 2.87% (95% confidence interval (CI) 2.28 to 3.47%; p<0.001, paired t-test). An analysis of hypofractionated plans revealed that the mean predicted lifetime absolute risk of lung cancer was 4.78% +/− 0.43% in the supine position and 1.88% +/− 0.17% in the prone position (p<0.001).

There was no significant difference in lifetime absolute lung cancer risk when comparing the two fractionation schedules for the supine position (p=0.18). The mean difference in absolute risk between the fractionation schedules was 0.10% (95% CI 0.08 to 0.13%; p<0.001, paired t-test) in the prone position. However, there was no difference in the relative risk (RR) of SLMs between standard fractionation and hypofractionated schedules in either the supine (RR 1.05, 95% CI 0.97 to 1.14) or prone positions (RR 1.01, 95% CI 0.88 to 1.15). A higher risk of SLMs was predicted for supine breast irradiation when compared with prone breast irradiation for both the standard fractionation and hypofractionation schedules (Relative Risk (RR) 2.59, 95% confidence interval (CI) 2.30 to 2.88, and RR 2.68, 95% CI 2.39 to 2.98, respectively).

Finally, the relative risks of SLMs for the various treatment techniques compared to background risk were as follows: supine standard fractionation, RR 4.04, 95% CI 3.62 to 4.46; supine hypofractionation, RR 3.98, 95% CI 3.56 to 4.40; prone standard fractionation, RR 1.56, 95% CI 1.46 to 1.66; and prone hypofractionation, RR 1.48, 95% CI 1.40 to 1.57. The relative risks compared to background risk by schedule and position are summarized in Table 2.

The mean expected lifetime absolute risks of lung cancer associated with all four treatment techniques and the background risk are shown in Figure 4.

Discussion

Several epidemiological studies with long-term follow-up have documented an increased risk of SLMs in women who have been treated with post-mastectomy radiotherapy (16-20) or supine whole breast radiotherapy (21-30). These studies have typically included patients treated with standard whole breast irradiation to a total dose of 4500 to 5000 cGy in the supine position (31). This standard of care technique, however, has been known to result in substantial levels of radiation delivered to nearby organs including the ipsilateral lung and the heart (32, 33).

In recent years, alternative techniques of breast irradiation have emerged for patients with early stage breast cancer. Prone breast irradiation was initially developed to improve dose homogeneity in woman with larger breasts (34). The technique also typically results in decreased dose to normal tissues, including the heart and lungs, as the breast tissue tends to fall away from the chest wall. Many studies have confirmed the improved dosimetric results of prone breast irradiation by showing consistently decreased radiation dose to the lungs (35-39). However, there have been no studies published to date that quantify, or have attempted to quantify, the excess relative risk of secondary lung malignancy when a breast cancer patient is treated with radiotherapy in the supine position compared with the prone position.

Another alternative technique, using a hypofractionated schedule, has been shown to be equivalent to standard fractionated whole breast RT in a recently updated randomized controlled trial with 10-year median follow-up (11). Due to the relatively recent acceptance of this fractionation schedule, there have been no studies published to date that have quantified, or attempted to quantify, the relative risk of secondary lung malignancy when a breast cancer patient is treated with hypofractionated breast radiotherapy compared with standard fractionated breast radiotherapy.

The relative risk estimates for SLMs from the present study are comparable with the relative risks of secondary lung malignancies estimated from several retrospective series on patients treated with supine whole breast irradiation (21-25). A study from the Institut Curie examined the risk for different secondary malignancies after breast radiotherapy and noted a statistically significant increase for lung cancers (relative risk 3.1), a risk that persisted after 10 years of follow-up from radiation treatment (21). In a large cohort study of patients from the Connecticut Tumor Registry, the estimated relative risk of secondary lung malignancy after breast radiotherapy was 2.8 after 15 years of follow-up (22). Another study using the Connecticut Tumor Registry by Neugut and colleagues found a relative risk of 3.2 for secondary lung malignancies after more than 10 years from the initial primary breast cancer (23). A large-scale retrospective SEER analysis on more than 180,000 breast cancer patients treated with radiotherapy estimated the relative risk of SLMs to be 1.93 at 10 years and increasing with longer follow-up (24). A study using The Swedish Cancer Registry to identify 141,000 women with breast cancer estimated the relative risk of SLMs to be 3.2 at 10 years (25). Finally, a separate SEER analysis focusing on mortality showed a long-term increased risk of lung cancer mortality in patients treated with whole breast RT compared to the general population with a relative risk of 2.7 after 15 years (40).

The current study suggests that the predicted lifetime risk of radiotherapy-induced lung cancer is significantly lower when women are treated in the prone position with whole breast radiotherapy compared to the more commonly utilized supine position. According to these estimates, the relative risk of secondary lung malignancies may be decreased more than two-fold using the prone breast technique. On the other hand, hypofractionated breast radiotherapy did not appear to clinically significantly affect the secondary lung malignancy risk when compared with standard fractionation for a given treatment position.

There are several limitations that are important to note with respect to the current study. First, this study was performed on a limited number of patients with risk factors that may not be representative of those of the general population. Second, these results are based on a model that, while improved over older secondary malignancy models as it incorporates both short-term and long-term carcinogenic processes, excludes potential confounding factors that may affect secondary lung malignancy risk. These potential factors include family history, chemotherapy use (17), and, most importantly, smoking status (23, 29, 41-43). Third, this study did not investigate the potential influence of either a boost to the tumor bed or the breath-holding technique on the risk of SLMs. The breath-holding technique, which is occasionally employed in supine breast radiation therapy and may result in decreased ipsilateral lung dose (44), was not utilized in any of these patients. Fourth, the amount of lung irradiated in a typical supine breast radiotherapy plan may vary from physician to physician, which may limit the application of these findings. Fifth, prone breast radiation therapy itself has several potential limitations including less reliable setup and patient tolerability as compared to supine positioning (45). Furthermore, prone positioning may not be applicable in patients requiring regional nodal irradiation such as patients with locally advanced or node positive breast cancer (33). Finally, partial breast irradiation, a newer technique currently the subject of a large-scale randomized trial, NSABP B-39 (46), may result in an even lower risk of secondary lung malignancy than prone breast radiation. This topic is currently being studied at our institution in a separate modeling analysis.

Conclusions

Patient position during whole breast radiation treatment is an important factor in determining the associated risk of secondary lung malignancy. Treatment in the prone position produces a substantially lower risk of secondary lung malignancy than treatment in the more common supine position. Quantifying this risk may be useful for clinicians as they counsel women with early stage breast cancer about their treatment options.