Stereotactic body radiation therapy for post-pulmonary lobectomy isolated lung metastasis of thoracic tumor: survival and side effects

Tumor lung metastasis is one of the most common oncologic problems, and affects a large percentage of patients of cancer despite the histology of the primary tumor. In most cases, widespread metastases are observed. But in certain instances, lung metastasis may exist in isolation. Resection of isolated lung metastasis (ILM) has traditionally been practiced using methods, such as thoracotomy and video-assisted thoracoscopic surgery (VATS) [1‐5]. Such approaches have been proved to be effective, achieving a median survival of 35 months, and are associated with generally acceptable morbidity and mortality rates [6]. However, such pulmonary operations have been practiced in salvage treatments for colorectal cancer, breast cancer and other types of tumors, and rarely for thoracic tumor after pulmonary lobectomy.

Hypo-fractionated stereotactic body radiation therapy (SBRT) can deliver high, biologically effective doses to the tumors while minimizing the irradiation dose to the surrounding tissues [7]. Over the decades, SBRT has emerged as an alternative treatment for medically inoperable patients with early-stage non-small cell lung cancer (NSCLC), showing a 5-year survival rate of more than 80% with limited morbidity [8‐11]. Even among patients with multiple pulmonary metastases, SBRT has been reported as a safe and effective strategy [12‐14]. At present, SBRT is recommended by the National Comprehensive Cancer Network (NCCN) panel as a salvage treatment for patients with ILM [15].

For patients with ILM after pulmonary lobectomy, a few treatment outcomes have been reported to date, including for surgery and SBRT. Therefore, we retrospectively analyzed the clinical outcomes of patients at our institutions with post-lobectomy ILM who were treated with SBRT as a component of their overall treatment regimen.

The basic and clinical characteristics of the studied population are summarized in Table 2. The median age of the patients was 58 years (range: 45–74 years); most of them were male and with Eastern Cooperative Oncology Group (ECOG) performance status score 0–1 (21/23; 91.3%). Of the 23 patients, 10 (43.5%) had squamous-cell lung cancer, 10 had non-squamous cell lung cancer and three (13%) patients had, other pathological types respectively. According to our records, 9, 3, 5 and 4 patients had received resection of right upper lobe, right lower lobe, left upper lobe, and left lower lobe respectively. Two patients had undergone left pneumonectomy. The pathologic stage confirmed with surgery(Staging system, American Joint Committee on Cancer, 6^th edition) [22] in the present study were 6 (26.1%) stage I, 9 (39.1%) stage II, and 8 (34.8%) stage III respectively. The ILMs of the contralateral (12 patients) and ipsilateral (11 patients) lung were observed. The median time from surgery to ILMs and follow-up time was 16.0 months (range: 4.0-75.0 months, only 1 patient was diagnosed with ILM 75 months after pneumonectomy) and 14.0 months (range: 6.0-47.0 months), respectively.

The treatment of cancer patients with ILM is a common clinical problem. SBRT is an appealing treatment option, but little is known about its use in the post-lobectomy setting. Our initial experience of using hypo-fractionated SBRT for ILM after pulmonary lobectomy is presented here for the first time.

The 1-year LCR was 91.3% for all patients in the present study, the median PFS and OS were 10.0 and 21.0 months, respectively. Our data are consistent with those of other reports on SBRT for metastatic lung cancer [23‐25], especially the report from Norihisa et al. whose prescription dose was 48–60 Gy/4-5 fractions and LCR 90.0% [25]. These clinical outcomes were comparable with those achieved by surgical metastasectomy [6]. In 2009, Rusthoven et al. reported a prospective multi-institutional phase I/II trial of SBRT for metastatic lung tumor, and reported actuarial LCRs at 1 and 2 years after SBRT of 100 and 96%, respectively. After a median follow-up of 15.4 months, a median survival of 19 months was achieved using a prescription dose of 48–60 Gy in three fractions [26]. Our data also confirmed that the main pattern of failure after SBRT was distant metastasis, as was concluded in a systematic review by Chi et al. [27].

A few studies have evaluated the outcomes of SBRT among patients after pneumonectomy [28‐30]. Authors from the VU University Medical Center in the Netherlands reported on 15 patients with a second primary lung cancer who received SBRT after pneumonectomy in 2009 [28]. After a median follow-up time of 16.5 months, no local failures were observed and the 1-year actuarial disease-free survival rate was 92%. In 2013, the same investigatorsy updated their data and compared the outcomes between SBRT, hypo-fractionated radiotherapy, and conventional radiotherapy among such patients [29]. In this paper, they reported a median OS of 39 months during follow-up. Thompson et al. identified 13 patients with newly identified lung malignancy after surgery from 406 patients who received SBRT [30]. The doses delivered were 60 Gy/3 fractions (n = 1), 54 Gy/3 fractions (n = 1), 48 Gy/4 fractions (n = 7), 60 Gy/8 fractions (n = 2), and 50 Gy/10 fractions (n = 3). The Median survival was 29 months, and no local failures were observed. In our cohort, the targets were metastatic tumors; thus, even though the local control rate was similar to that the reported in the aforementioned studies discussed above, the OS (median 21 months) among our patients was shorter than in those patients with newly diagnosed lung cancer.

The SBRT treatment was well tolerated in our patient population. The most common toxicity was cough (60.9%), and acute RP of grade 3 or worse was observed in three patients (13.0%). Other treatment-related toxicities included dyspnea, chest wall pain, and acute esophagitis. These findings are consistent with the reports evaluating SBRT in newly identified lung cancer after pneumonectomy [28, 30] and in medically inoperable or operable NSCLC [26, 31‐37]. In the post-pneumonectomy settings, Haasbeek et al. reported that only two2 patients experienced toxicity of grade 3 or higher toxicity [28]. In a report by Thompson et al., two2 patients in a 13-patient cohort had grade 3 RP [30]. For medically inoperable or operable NSCLC, in the RTOG trial 0236, with a prescription dose of 54 Gy in three fractions, treatment-related grade 3 and 4 toxicities of pulmonary or upper respiratory tract were observed in 14.5 and 1.8% of patients, respectively [31]. In a phase II study of SBRT, Baumann et al. reported that grade 3 pulmonary toxicities were seen in 11.8 and 12.5% of patients in the cardio-vascular disease group and chronic obstructive pulmonary disease group, respectively [32]. In a study from Japan (JCOG 0403), Nagata et al. reported grade 3 toxicity in 6.2% of their patients who received SBRT treatment [33]. For metastatic lung cancer, Rusthoven et al. reported that grade 2 RP occurred in only one patient (2.6%) in their multi-institutional phase I/II trial [26]. The investigators suggested that the low rate of pneumonitis observed might have contributed to the dose constraint used (V₁₅ < 35%) in their patient population. However this needs to be confirmed in a larger cohort of patients because in our study, one patient developed grade 4 RP, and the V₁₅ was less than 17% according to the DVH analysis.

Several studies have evaluated the potential value of the DVH-based lung parameters in predicting acute “symptomatic” RP after SBRT [34‐38]. The RP rates were reported within a range of 9.4% to 28.0%, and the possible predictive factors for RP differed among these studies. In 2007, Yamashita et al. reported that 29% of their patients had developed grade 2 or worse RP after SBRT (48 Gy in four fractions), and that the conformity index was the only factor associated with incidence of RP [34]. Ricardi et al. observed a good correlation between mean lung dose (MLD) and grade 2–3 pulmonary toxicity (p = 0.008, odds ratio 1.5) in a 60-patients cohort after SBRT of 15 Gy per fraction × 3 fractions [35]. Moreover, reports by Borst et al., Guckenberger et al. and Barriger et al. indicated that MLD (ipsilateral or total lung) was correlated with incidences of symptomatic RP after pulmonary SBRT [36‐38]. Onestudy indicated that V₅ of total lung >37% and V₅ of contralateral lung > 26% were suitable predictors of pneumonitis in a cohort of patients treated with SBRT [39]. Additionally, Guckenberger et al. reported that the V_2.5-V₅₀ were correlated with incidences of RP with a continuous decrease of the goodness of fit for higher doses [37]. In a Japanese study, Matsuo et al. concluded that the symptomatic RP rate was significantly lower in the group with PTV < 37.7 mL compared with the group with larger PTV (11.1 vs. 34.5%, p = 0.02) [40]. In the present study, we also identified two factors that might significantly be associated with RP of grade 3 or worse after SBRT in the post-lobectomy situation: PTV and IpV₅. Like other parameters already mentioned, the value of these two factors as the thresholds in SBRT for ILM warrants further clinical investigations.

To the best of our knowledge, there is little information regarding the correlation between various DVH-based factors and lung toxicity in radiotherapy among patients after pulmonary lobectomy. Uno et al. reported that higher in V_13/20 and MLD values could be a surrogate for RP in NSCLC patients after lobectomy [41]. While the treatment was concurrent chemo-radiotherapy for recurrent NSCLC, these parameters could not be easily followed in an SBRT setting.

Some limitations of the present study justify mention. First, this analysis was retrospective and the number of patients evaluated was limited, thus leading to a bias of selection. Second, being a multicenter study, there was no central data review, and the determination of RP can be subjective and challenging. Third, there is an obvious difference between the RTOG system, CTC AE v2.0, and CTC AE v3.0 regarding steroid use for RP. Tucker et al. reported 442 patients who received definitive radiotherapy using these three toxicity grading systems: RP of grade 2 or worse was observed in 29, 25, and 44% of patients according to RTOG, CTC AE v2.0 and CTC AE v3.0, respectively [42]. Therefore, attention should be paid to the toxicity grading systems when interpreting the results discussed herein.

In conclusion, our results indicate that SBRT is a promising tool for the salvage treatment of ILM in patients who had previously received pulmonary lobectomy. PTV and IpV₅ are possible predictive factors for the development of symptomatic RP. Prospective studies are needed to verify these findings.