Background

Lung cancer is the leading cause of cancer death. Despite major progress in the understanding and management of this cancer, it is responsible for more than 33,000 deaths per year in France and 1.8 million deaths per year worldwide [1].

The standard treatment for localized non-small cell lung cancer (NSCLC) is parenchymal resection with lymph node dissection and there is increasing evidence in favor of minimally invasive surgical techniques to reduce operative risks and optimize functional recovery [2].

However, with higher average life expectancies and improved screening techniques (e.g. low dose thoracic computed tomography (CT)), there is an increased incidence of early stage I lung cancer diagnosed in the elderly or higher surgical risk patients [3]. Depending on the series, up to 25% of potentially resectable patients are not operable or refuse surgery [4, 5]. Notably, some patients will present with a second primary localized lung cancer after resection of their prior lung cancer, where operability can be challenging. These second cancers are defined as metachronous when they appear after a cancer free interval from the initial primary, as opposed to synchronous cancers, discovered concomitantly with the first. The cumulative incidence of metachronous lung cancer (MLC) is estimated at 8.3% at 5 years [6]. Although it is clearly accepted that MLC requires an ablative treatment (e.g. surgery, SABR, radiofrequency ablation), the choice of the therapeutic approach is debatable [7,

Table 1 Patient characteristics
Full size table

SABR treatment

The majority of patients were treated with 60 Gy in 3 fractions (n = 116, 84.7%), followed by five fractions regimens of 10 to 15 Gy (n = 21, 15.3%) (Table 2). The median radiation therapy duration was three days and only exceeded one week in 12 (8.8%) patients. The median interval between primary lung cancer to the second metachronous cancer was 39.4 months (13.9–121.4), which were initially treated with resection (n = 21, 75%) or SABR (n = 7, 25%) for their primary lung cancer. GTV and PTV values were significantly smaller (p = 0.007 and p = 0.001 respectively) for MLC when compared to the PLC cohort.

Table 2 Treatment characteristics
Full size table

Response and survival data

Overall, 8 (5.8%) patients had a complete response after SABR, 35 (25.5%) had a partial response, 55 (40.1%) had stable disease, 28 (20.4%) had progressive disease, and 11 (8%) patients could not be evaluated due to pneumonitis. Approximately half (n = 66, 48.2%) of patients recurred after SABR, with 30 (21.9%) having local recurrence, 56 (40.9%) having regional recurrence, and 54 (39.4%) having distant recurrence.

The 1-year/3-year LC and RC rate were 83.6%/72.6%, and 73.8%/58.2%, respectively. The 1-year/3-year MFS, PFS, and OS were 76.1%/58.9%, 68.7%/50.9%, and 78.6%/52.1%, respectively.

A total of 85 died (62%), with 7 (5.1%) from unknown cause, 51 (37.2%) experienced cancer specific mortality, 3 (2.2%) attributed to SABR toxicity, and 24 (17.5%) unrelated, of which consisted of 12 (50%) of respiratory failure, 8 (33.3%) of cardiac failure, 3 (12.5%) of cachexia, and 1 (4.2%) of fall trauma.

The median OS was 38 months in the PLC group versus 34 months for MLC (p = 0.9). There was a non-significant increase in cancer specific mortality in the PLC group (40.4% vs. 25%, p = 0.4), when compared to MLC. There was also no significant difference between PLC and MLC, in the 3-year PFS (68.7% vs. 50.9%, p = 0.9), MFS (76.1% vs. 58.9%, p = 0.3), LC (83.6% vs. 72.6%, p = 0.2) and RC (73.8% vs. 58.2%, p = 0.8) (Fig. 1).

Fig. 1
figure 1

Survival analyses (Kaplan–Meier). A Overall survival, B recurrence-free survival, C metastasis-free survival, D local control, E regional control. Solid lines correspond to the stage I primary lung tumor group. Dashed lines correspond to the metachronous tumor group. Cross marks correspond to censored data

Full size image

Tumor volume is significantly different between the 2 cohorts and could consequently influence tumor control. With a cut-off of 8 mL (median), GTV was not found to be correlated with local control, locoregional control, metastasis-free-survival or progression-free survival (p > 0.05).

We divided the whole group in a group with small GTV (≤ 8 mL) and another one with larger GTV (> 8 mL). The prognostic impact of the cohort (primary versus metachronous) on local control was then analyzed in these 2 groups. No significant correlation was found (p = 0.5 if GTV ≤ 8 mL; p = 0.7 if GTV > 8 mL).

Toxicities data

In the overall cohort, 71/137 (51.8%) patients experienced at least one toxicity attributed to SABR, with 66 (48.2%) grade 1–2 toxicity and 5 (3.6%) grade 3 + toxicity (Table 3).

Table 3 SABR toxicities
Full size table

There were 3 (2.2%) grade 5 toxicities, 2 in the PLC cohort (1 radiation pneumonitis, 1 acute respiratory failure) and 1in the MLC cohort (acute respiratory distress).

There was no significant difference between the PLC and MLC groups in terms of total (54.1% vs. 42.9%, p = 0.6), grade 1 to 2 (50.5% vs. 39.3%, p = 0.6) and grade 3 + toxicities (3.7% vs. 3.6%, p = 0.9).