Stereotactic Body Radiotherapy for Patients with Lung Oligometastatic Disease: A Five-Year Systematic Review

Simple Summary Stereotactic body radiation therapy (SBRT) is currently used to treat lung metastasis, particularly in the setting of oligometastatic disease. Objective of this treatment is lesion’s ablation. However, its techniques, efficacy, and safety remain to be evaluated. Several teams have published their experiences with different radiotherapy schedules. The aim of this review was to analyze these topics in regard to the last five years of publications. For this purpose, we carried out a systematic review of the literature using the PRISMA method. This review can help oncologists involved in lung oligometastasis to clarify their knowledge through the wealth of published data and may lead to improved management. Abstract For several years, oligometastatic disease has represented an intermediate state between localized disease accessible to local treatment and multimetastatic disease requiring systemic therapy. The lung represents one of the most common metastatic locations. Stereotactic body radiation therapy (SBRT) appears to be the treatment of choice for these patients. There are few data defining the place of radiotherapy and reporting outcome after SBRT in lung metastases. This 5-year review aimed to determine areas of SBRT usefulness and methods for the management of pulmonary metastasis in oligometastatic patients. A search for articles on PubMed allowed selection of the most relevant studies. Eighteen articles were selected according to pre-established criteria for this purpose. The analysis concludes that SBRT is an effective and safe treatment in selected patients when the disease remains localized from one to three organs.


Introduction
For several years, Hellman and Weichselbaum described oligometastatic disease evolution as involving a low number of metastases in only a few organs [1]. Depending on the authors, the definitions have varied from one to five metastases in one to three sites [2][3][4]. Some studies have suggested that life expectancy could be longer for these patients compared with those with more spread out disease [5,6]. Moreover, some studies have shown that local treatments can improve overall survival [2,7]. SABR-COMET, a recent phase II study, investigated the effect of stereotactic body radiation therapy (SBRT) on oligometastatic patients' survival. SBRT was associated with an improvement in overall survival compared with palliative standard of care treatments alone [2].
The lung is one of the most frequent sites of metastases for solid tumors [8]. Lung metastasectomy is the historical treatment, but it requires good general conditions and cardiovascular and respiratory functions [9][10][11]. Based on the favorable results of SBRT in non-small cell lung cancer (NSCLC) [12][13][14], it was proposed as a curative treatment for patients with lung metastases that were unsuitable for surgery. However, patient selection criteria remained blunt, and several variable irradiation schedules were used. No robust data allowed recommendations, although reported local control and survival rates
The number of irradiated metastases was 1705. The median number of metastases per patient ranged from one to two. The tumor diameters varied from 2 mm [17] to 124 mm [28]. Eight articles described local treatment prior to SBRT. Surgery was used at least once before SBRT in 164 patients, but no thermoablation before SBRT was identified. Only two authors specified surgery for metastases, which represented 44 patients [18,21,22,24,25,27,30,33]. The 2-year LC, PFS, and OS ranged from 31% to 93%, from 10% to 38%, and from 40% to 85%, respectively. The characteristics and outcomes of patients are reported in Table 1.
Only one study showed a relationship between the modalities of SBRT and toxicities. A dosimetric analysis revealed that the prescription maximum point (51.9 vs. 47.4 Gy) and higher bronchus maximum point dose (49.0 vs. 46.1 Gy) were statistically significant predictors of late grade 2 or higher atelectasis [31].

Prognostic Factors
Prognostic factors are reported in Table 2.

Tumor Size
LC was significantly better for the smallest metastases in three series [17][18][19]. Osti et al. and García-Cabezas et al. found that LC rates were significantly better for metastases smaller than 18 and 25 mm, respectively; the 5-year LC was 89.5% vs. 55.3%, (p = 0.001) in Osti et al. [17] and 100% vs. 78.9% (p = 0.023) in García-Cabezas et al. [18]. Helou et al. found that the largest lesions were independently associated with worse LC (HR 1.45; 95% CI: 1.01-2.10; p = 0.046). These authors did not specify a size threshold [19]. In contrast, Wang et al. did not find any LC differences between the volume of lesions more or less than 10 cc [29].

Number of Thoracic Metastasis Lesions
In univariate analysis, Wang et al. reported that the number of thoracic metastases was correlated with PFS. The 2-year PFS rate was 33.4% for patients with one lung metastasis compared with 0% for the other patients (p = 0.012) [29].

Fractionation of SBRT
In their univariate analysis, Sharma et al. found that a single-fraction irradiation was associated with lower LC than a multifractionated irradiation (HR 2.83; 95% CI: 1.43-5.59, p = 0.003). This result was not found in multivariate analysis [23]. the only LC prognostic factor in Berkovic et al. [24]. Figure 2 shows local control rates as a function of 10GyBED ( Figure 2).   [19].

Primary Colorectal Cancer
For Osti et al., the colorectal primary of lung metastasis (CRLM) was a significant predictive factor of poor LC and PFS. The 3-year LC rates were 64.8% for patients with CRLM compared with 86.3% for other primary tumors (p < 0.01), and the 3-year PFS rates were 12.1% and 41.9%, respectively (p = 0.03) [17]. Helou et al. reported a hazard ratio of 13.59 (95% CI 4.19-44.12; p < 0.001), with the poorest LC rates for CRLM [19].
Jung et al. showed that a pre-SBRT carcinoembryonic antigen was significantly predictive of the LC rate. Three-year LC rates were 0% and 78.1% for patients with CEA < 6 µg/L and ≥ to 6 µg/L, respectively (p < 0.01) [26].
Jung et al. found that the timing of SBRT (for a second or third lung metastasis) was a statistically significant prognostic factor for 3-year LC, with 63% for the first lung metastasis and 91% for the others (p = 0.04). The first session may have selected "good responder" patients for whom a second or even a third session was more easily indicated [26]. However, the authors did not discuss the possibility that the radiosensitivity of the first treated metastasis could be a marker for irradiating subsequent metastases that remained radiosensitive. For Li et al., patients treated for synchronous oligometastatic diseases or multiple lung metastases had a worse PFS than patients with a metachronous or single lesion (6% vs. 46%; p < 0.001 and 11% vs. 44%; p = 0.04, respectively) [25]. Li et al. found that oligoprogression was a worse predictor of regional control compared with oligometastatic disease (HR 2.78; 95CI: 1.04-7.48; p = 0.042) [22].

Discussion
For non-operable patients, to control lung metastases, SBRT seems to be the best option. Indeed, in recent years, technical progress has allowed delivery of an ablative dose to various locations [34]. Although no phase III trials are specifically available to definitively prove the efficiency of SBRT, several retrospective studies have demonstrated relevant results in lung oligometastasis [35,36].
Main objective of SBRT is the lesion's ablation. LC and survival data was good in our study, as in other reviews [37]. Reported LC rates are noteworthy, reaching up to 96% according to Navarria et al. [21]. However, the technical modalities are not yet standardized. Radiation oncologists have used different schedules according to the previously published results, mainly mimicking the treatment of stage I primitive lung cancer or following their own experiences. This heterogeneity has induced difficulties in comparing the studies and, consequently, proposing standard treatment protocols.
To our knowledge, this PRISMA literature review is the largest study regarding lung SBRT for oligometastatic lesion. More than one-third of patients reported in our analysis had a fiducial implantation. Neither of the two studies including patients treated with or without fiducial implantation reported an advantage to this technique [24,30]. Local control reported in studies with patients with fiducial implantation was not greater than that in studies with patients without fiducial implantation (57.4% to 90.6% vs. 30.9% to 96%) [23,28,29,31]. In addition, the pneumothorax risk was measured from 9% to 23%, depending on the study and this technique delays treatment [24,38]. It is therefore not appropriate to recommend systemic fiducial implantation as a tracking technique.
More than half of patients in our analysis had received a 4D scan. Although there is no demonstrated advantage, the use of a 4D scanner seems to be the best non-invasive technique to integrate respiratory movements to define the ITV [39].
The optimal physical dose to reach a high probability of LC is unknown, but this review suggests that a 10 GyBED of 100 Gy could be proposed to increase the LC [19,23,25,29]. This result is parallel with data obtained for NSCLC in which a 100GyBED affected LC [13,40]. CRC metastases appear to require a larger BED. The only prescription dose prognostic factor of LC found in the series concerned CRLM. Indeed, 60 Gy in 4 or 5 fractions could be a relevant threshold (corresponding to a BED of 132 or 150 Gy) [19,26]. Mazzola et al. found better result of SBRT after bevacizumab in lung oligometastases from colon cancer [41].
A multi-institutional phase I/II trial assessed SBRT for lung metastases. The delivered irradiation dose varied from 48 to 60 Gy in 3 fractions. This corresponded to a 10GyBED from 106 to 120 Gy. The actual 1-and 2-year LC rates were 100% and 96%, respectively [42].
Complication rate did not differ significantly by fractionation in studies reporting single and polyfractionated treatment [20,21,23,25,27,29]. However, care must be taken when irradiating lesions located in the no-fly zone or near the chest wall. Indeed, an adapted fractionation is necessary during the treatment of the most central lesions in order to obtain LC and toxicity similar to that of peripheral lesions [43,44]. Mazzola et al. proposed a simultaneous integrated protection (SIP) strategy for SBRT of central lung lesion with good results [45].
There was no prognostic factor to define the best margins in this review. Patients with oligometastatic disease are eligible for curative treatment. If a cure cannot be achieved, the treatment of each lesion can improve overall survival and PFS [46,47].
LC could be a survival prognostic factor. Indeed, a retrospective, multicenter European study showed that lung metastasis removal led to encouraging long-term survival, specifically after complete resection. The 5-year survival rates after complete and partial metastasectomy were 36% and 13%, respectively [10]. The good local results reported in this review and the absence of difference in terms of CL comparing SBRT and surgery argue to propose SBRT as an ablative treatment of lung metastases that can improve patient survival [28]. In the SABR-COMET phase II study, OS was significantly better in the experimental group. The median OS was 28 months in the control arm vs. 50 months in the SBRT arm (p = 0.006). In this study, the lung was the most common metastatic site and was involved in 53% of patients in the control arm and 43% in the SBRT arm. The treatment schedule varied according to lesion location in lung parenchyma. Doses varied from 54 to 60 Gy in 3 to 8 fractions every second day [2]. COMET 3 and COMET 10 phase III trials comparing SBRT and palliative care may help to confirm SBRT place in treatment oligometastatic patient's treatment.
Depending on the studies, the 2-year OS rates varied by 2-fold (40 to 85%) [21,31]. This difference reflects heterogeneity in the patient population included in the studies. Indeed, all studies were retrospective, and the first inclusion criterion was to have received SBRT for oligometastatic disease. A better selection of patients would be necessary to obtain better results. However, currently, there are no well-established criteria for this selection.
The median number of metastases per patient reported in this review ranged from one to two, consequently, for oligometastatic patients, it seems reasonable to propose SBRT up to two lung lesions. However, no data are available to advise repetition of SBRT with this number of metastases.
Clinical factors may split patient outcomes and could be used to adapt radiation treatment schedules. The responses of synchronous, metachronous, and progressive tumors were different. Patients with synchronous metastases have poorer prognoses [22,23,28]. Alongi et al. concluded the same results in a previous literature review [37]. Reported variables in the used series were so heterogeneous that it made our analysis difficult. The primary tumors were variable, and irradiation occurred at different times during the history of disease. In this context, EORT defines genuine oligometastatic disease with patients with no history of polymetastatic disease having low metastatic capacity. SBRT could delay the introduction of poorly tolerated systemic therapy. This would allow maintenance of a better quality of life. In contrast, some patients have an induced oligometastatic disease with a history of polymetastatic disease. These patients have a disease with metastatic potential. In this case, SBRT consolidates systemic therapy and aims to restore a status of stable disease or a status of complete response. It may also aim to restore a status of overall sensitivity to systemic therapy through eradication of oligometastases with resistance to the current line of systemic therapy [48]. A review of the literature has specifically studied the latter in the context of patients with oligoprogressive oncogene-addicted NSCLC treated with SBRT. The authors reported good results, with PFS of more than 10 months after local treatment and continued systemic therapy in patients with EGFR-mutated NSCLC considered resistant to TKIs [49].
The circulating micro-DNA analysis seems promising and could influence the oligometastatic or polymetastatic tumor profile phenotype through gene expression regulation [50,51]. This parameter could participate in the development of a predictive score for patients' metastatic history, with histology, size of the lesion, lesion numbers, metastatic site numbers, time between initial diagnosis, and metastases appearance and could better characterize the type of oligometastatic disease the patient has in order to optimize treatment.
It is still necessary to define scores and criteria to select patients for whom SBRT is impactful given their disease history. A data base for oligometastatic patients would be useful in the absence of phase III trials. Radiomics and genomics are interesting approaches to improve the characterization of the patient's prognosis [52].
It is reasonable to treat one or two lesions, while obtaining satisfactory toxicity and efficacy. This number could be increased if dosimetric conditions are respected to avoid any adverse effect; Lindsay et al. irradiated up to seven lesions per patient with a seemingly acceptable tolerance [33].
In a strategy to improve survival, toxicity must be evaluated. In this analysis of the literature, the most frequently reported complications were pneumonitis or lung fibrosis. Their incidences remained low, but several reasons can explain these results. First, given the reluctance to perform multiple exams in patients with metastatic disease, radiation complications can arise late and can thus be rarely diagnosed as a consequence of the relatively short life expectancy of the patients. However, the scarcity of these complications with SBRT had already been reported in NSCLC [53], and they are more rare than after conventional fractioned radiotherapy [54,55]. In NSCLC, the RP incidence (grade 2 or more) after SBRT varied from 2% to 28%, and this diagnosis was clinical and radiological [53]. None of the selected studies found a prognostic RP factor. A Danish study showed that comorbidity, older age, and average or lower location of the lesion influenced RP risk, but they also described tobacco as a protective factor. These data concern treatment with normofractioned radiotherapy [56]. Johansson et al. presented the same results about smokers, even with traditional splitting. They hypothesized that it could be explained by immunological reactions induced by cigarette smoke or by hypoxia, which smoking generates. However, the role of tobacco remains controversial [57]. Menoux et al. proposed recommendations concerning dosimetric parameters to be applied to reduce RP risk. They also point out that routine use of these parameters is difficult to apply, knowing that the definition of "healthy lung" varies depending on the study [53].

Conclusions
SBRT is an efficient and well-tolerated treatment for lung metastases in oligometastatic patients. However, the optimal treatment schedule is not definite. Protocols used for NSCLC, with a BED > 100 Gy, appear to be appropriate to obtain a LC comparable with that of surgery. It seems necessary to define the type of oligometastatic disease of the patients with the help of clinical, radiological, and biological markers in order to opt for the best strategy. SBRT is an ablative treatment to be integrated in this strategy, especially when surgery alone is unreasonable.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/cancers13143623/s1, Table S1: Study criteria of the 18 papers used for PRISMA method. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.