Background
Stereotactic body radiotherapy (SBRT) results in a high local control rate for relatively small lung tumors and has low treatment-related toxicity, and it thus has many benefits for patients, especially patients who cannot undergo surgery [
1‐
3]. SBRT may also be beneficial for patients who choose not to undergo surgery. For example, elderly patients, even elderly patients with no or only a few comorbidities and moderate lung function, often hesitate to undergo surgery because of concerns about postoperative complications, decline in activities of daily living, and progression of dementia, even in short-term admission. A time-trend analysis in the Netherlands showed that the number of elderly patients treated with radiotherapy for stage I non-small-cell lung cancer (NSCLC) has been increasing over time. This has occurred in parallel with increased availability of SBRT and has led to a decrease in untreated patients and no increase in surgically treated patients, despite operative and perioperative advances such as video-assisted thoracic surgery [
4]. Furthermore, the outcome for stage I NSCLC treated with SBRT is now close to that after lobectomy, based on the results of a recent propensity-score matched analysis [
5].
Retrospective analyses and one prospective analysis of SBRT for oligometastatic lung tumors have shown 2- to 3-year survival rates with good local control (LC) that compare favorably with surgical results [
2,
3]. If SBRT has an efficacy comparable to that of tumor resection, this technique may provide patients with a better quality of life, shorter time away from work, and minimal interruption of other treatment. This third issue may be a major advantage because systemic chemotherapy is often performed for targeting oligometastatic lung tumors, treating another lesion, or as maintenance or consolidation therapy for potential metastases.
Recently, the new notion of oligo-recurrence was proposed because the initial concept of oligometastases did not eliminate the uncontrolled primary site with several distant metastases. Oligo-recurrence has been suggested as a state of metachronous limited recurrence or metastases possibly cured with local therapy [
6]. SBRT for lung oligo-recurrence with good LC rate and survival rate has also been reported [
7].
Many prognostic factors for LC after SBRT have been reported, including tumor diameter, standardized uptake value (SUV) on [18F]fluorodeoxyglucose positron emission tomography (FDG-PET), low dose distribution, metastatic lung tumors, and colorectal lung metastases [
8‐
14]. In contrast, there have been only a few studies in which prognostic factors for overall survival (OS) were examined, though clarification of such factors is important to maximize the benefit/toxicity ratio [
14‐
17]. In this study, we retrospectively reviewed our results for lung tumors treated with SBRT with the goal of identifying prognostic factors associated with LC and OS and thus establishing a strategy for balancing the benefits and risks in use of SBRT.
Discussion
This study was performed as a review of a single center experience of SBRT for lung tumors. In multivariate analysis, metastatic lung tumors, increased tumor size, early date of treatment and prescription dose of BED10 > 105 Gy were significant factors for unfavorable LC. LC itself was associated with OS in univariate analysis (HR: 2.02, 95% CI: 1.33-3.01, p = 0.001), indicating that a strategy for improving LC is important. For this purpose, dose escalation may be reasonable for metastatic lung tumors or larger tumors, given the relatively low toxicity in our patients, despite the poor pulmonary background and the presence of complications. However, this strategy may not be applicable in all cases and may not always give the expected outcome because critical structures can prevent delivery and distribution of a sufficient dose. This problem is also likely to be increased in dose escalation for larger tumors or multiple metastatic tumors.
Prescription dose of BED10, which was divided into two categories at the sample median (> 105 Gy, ≤ 105 Gy), was an independent predictor. The results confirmed previous findings [
8,
14]. Prescription dose ≤ 105 Gy was often used when the target was close to a critical structure. Thus, this strategy was relatively safe but might contribute to relatively low LC rate because of the low prescription dose.
The date of treatment was a significant factor for LC, possibly due to improvements in contouring (including spiculation of the GTV) and radiation planning (introduction of the concept of the homogeneity index). Furthermore, LC rates naturally became lower as the follow-up period became longer. The median follow-up periods in 2001–2005 and 2006–2011 were 50.2 months and 41.6 months, respectively, and this difference would also therefore affect the results.
We found that a metastatic lung tumor was an independent poor prognostic factor for LC and OS. For LC, this result confirms a previous finding of Hamamoto et al. [
13] and is based on a much longer follow-up period in the current study.
The poor LC rate of a metastatic lung tumor may be the result of formation of a metastasis. Metastatic tumor cells are the particularly potent malignancy of tumor, which are the most aggressive cells in the neoplasm because most malignant cells entering the metastatic process are killed, especially in the blood circulatory system [
19]. Also, in some cases, adjuvant chemotherapy or hormonal therapy is performed after surgery. Cells with high metastatic potential are resistant to this treatment based on genetic instability due to a several-fold increase in the rate of mutation compared to tumor cells with lower metastatic potential [
20]. This genetic instability increases the angiogenic and invasive potential of tumor cells, and the tumor cells escape from immune surveillance, thus increasing the chance of metastasis. Through these steps, and after formation of a metastasis, tumors in advanced stages develop biological heterogeneity and increased radioresistance before and during SBRT.
Another reason for the poor LC rate of a metastatic lung tumor might be the state of the patient. The effects of disease treatment and disease progression may lead to cachexia and anemia temporarily or throughout the course of the disease, making the tumor more hypoxemic. Tumor hypoxemia has been hypothesized to lead to tumor growth and resistance to therapy because of induction of angiogenesis, genetic mutations, resistance to apoptosis, and resistance to free radicals from radiotherapy [
21]. These changes may be one of the reasons for the poor LC of metastatic lung tumors, but both the “seed” aspect and the “soil” aspect may be important. According to the “seed and soil” hypothesis, metastasis is the product of interactions between selected cancer cells (the “seeds”) and specific organ microenvironments (the “soil”) [
22]. In the lung environment, stromal products cause upregulation of P-glycoprotein in the cells, which prompts excretion of a variety of toxic compounds, including chemotherapeutic drugs, resulting in enhanced resistance to drugs [
23]. Similarly, through tumor-stroma interactions, some stromal products may cause increased radioresistance of metastatic cells. If a microenvironment is changed to a rich soil due to interactions and molecular factors, circulating tumor cells may be reseeded, as at the primary site [
24]. Therefore, modulation of a tumor microenvironment using approaches such as antiangiogenic therapy may be effective in some metastatic cases.
In subgroup analysis, metastases from colorectal cancer showed a lower local control curve than did metastases from non-colorectal cancer, consistent with previous findings [
12]. However, we cannot conclude that colorectal metastases are more radioresistant than other metastases because of the small sample size and the variety of non-colorectal tumors, given that malignant diseases have various natural histories. For example, our cases included two cases of metastasis from thyroid cancer, which generally shows slow growth. Our findings do suggest that metastases from colorectal cancer are aggressive and are likely to cause short-term local relapse.
Although our inclusion criteria for metastatic lung tumors included both oligometastasis and oligo-recurrence, SBRT as a local therapy might be effective in both situations. According to the Norton-Simon hypothesis, the efficacy of chemotherapy is proportional to the growth rate of the tumor, and the growth is faster when the tumor is not bulky. Therefore, treatment of a bulky tumor with a slower growth rate and thus lower sensitivity to chemotherapy might make the remaining tumor cells more sensitive to chemotherapy [
25,
26]. Improvement in LC by reducing metastases should lead to increased OS for patients including both patients with oligometastasis and patients with oligo-recurrence.
Tumor diameter was identified as an independent factor for OS in multivariate analysis. This was probably due to the poor LC of a large tumor and the increased tendency for metastasis. In pulmonary resection with node dissection for clinical stage I NSCLC, 19.4% of the patients were found to have pathologically positive nodes, with a particularly high rate of 31.8% for unexpected positive nodes in clinical stage IB cases [
27]. Thus, this analysis showed that tumor size and solid consistency were independent predictors for node metastasis but did not show that unexpected pathologically positive nodes affected OS. However, in our analysis of stage I NSCLC cases, clinical hilar lymph node or mediastinal lymph node failure was a significant factor for unfavorable OS (HR: 1.73, 95% CI: 1.04-2.78, p = 0.034). This finding suggests that staging must be performed thoroughly before SBRT. In an operable case, assessment of suspicious lymph nodes by endoscopic ultrasound-guided fine-needle aspiration may be necessary. Adjunct therapy for a larger tumor should also be considered in cases in which this is possible.
Female gender was also found to be a favorable factor for OS but not for LC. This may be because smoking was less common in females, and this may reduce smoking-related complications and development of another malignancy. Also, epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) therapy for primary lung cancer with EGFR mutation was more common in females, with 6 of 16 female patients (37%) receiving an EGFR-TKI in the course of treatment, in contrast to only 2 of 66 male patients (3%).
Analysis of tumor consistency was not included in the study because there was no local failure of ground-grass opacity (GGO) tumors and there was only one case in which disease progression occurred. Most tumors with mainly a GGO component were atypical adenomatous hyperplasia, adenocarcinoma in situ (formerly referred to as bronchioloalveolar carcinoma) or minimally invasive adenocarcinoma, all of which have a relatively good prognosis [
28]. Thus, assessment of GGO tumors will require more cases and longer follow-up periods. However, it may be more useful to examine the safety of treating multiple sites by SBRT. Tumors with mainly a GGO component are often multifocal, and for all such tumors (not limited to GGO tumors), the lifetime incidence of a second primary tumor after surgical resection is > 10% [
29]. More cases of multi-site treatment should emerge as the outcome of SBRT improves, and this will allow an evaluation of the safety of repetitive treatment with SBRT.
There were several limitations to this study. First, this study was a retrospective single institute analysis with a limited sample size, the number of metastatic lung tumors being particularly small. Second, various treatment protocols were included in the analysis. There were a variety of total doses, methods of dose prescription and fractionation schema.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
TY designed the analysis, reviewed the clinical data, performed statistical analysis and drafted the manuscript. KJ treated the patients, reviewed the clinical data and revised the manuscript. YS and MK treated the patients and reviewed the clinical data. HM, TS, MK, RU, KA, NK, YI, MK, NT and KT treated patients and collected clinical data. YT conceived the study, reviewed the clinical data, and treated patients. All authors read and approved the final manuscript.