Re-irradiation in lung disease by SBRT: a retrospective, single institutional study

Lung cancer is one of the most important causes of death for cancer in the world. Patients with early-stage lung cancer have a probability of 4–10% developing a second lung tumor in the first 5 years after treatment [1, 2]. Patients with stage III Non-Small Cell Lung Cancer (NSCLC) treated with concurrent chemo-radiotherapy (CRT) have a risk of 30% for isolated locoregional failure [3]. The lungs are a common site for metastases. A considerable percentage of patients have disease limited to the thorax, which represents an ideal subset for local therapy [4]. A considerable number of patients with primary or metastatic lung lesions need to repeat thorax treatment for recurrent or metachronous disease. For this reason, thoracic re-irradiation starts to be common in clinical practice. Thorax recurrences can be separated in local (lung parenchyma, bronchial stump, or chest wall) and regional (mediastinal lymph nodes). The aim of treatment can be palliative or curative [5]. The improvements of technologies with Stereotactic Body Radiation Therapy (SBRT) have introduced the concept of ablative radiotherapy with curative intent. The SBRT is able to give a high dose of radiation sparing the surrounding normal tissue in 1–5 fractions [6]. The purpose of this study was to evaluate clinical outcomes in terms of Local Control (LC) and toxicity in patients with lung disease (primary or metastatic) which have received re-irradiation by SBRT. Also we have correlated Local Control with tumor histology, status of patient, number of fraction, doses, chemotherapy after treatment, time from the first progression of disease after the first treatment, time between both treatments. Furthermore, we have correlated toxicity with site of treatment, doses and volume.

From April 2011 to December 2016, twenty-two patients received a re-irradiation by SBRT. Twenty-seven lesions were re-irradiated. There were 15 (68%) male and 7 (32%) female. Median age at time of enrolment was 70 years (range: 47–82 years). This study included patients with comorbidity or advanced age or patients with low cardio-respiratory reserve which were unsuitable for surgery or patients refusing surgery. The patients at time of re-irradiation had an Eastern Cooperative Oncology Group (ECOG), PS 0–2, and the intent of treatment was non-palliative. Patients and tumor characteristics are summarized in Table 1. The median time to the progression disease from first treatment was 10 months (range: 0–54 months). The median time between the first and the second treatment was 18 months (range: 6–66 months). For each patient we obtained the disease history and an objective examination. According to these informations we classified for ECOG, PS. The imaging study was performed using TBCT with contrast enhancement or with18FDG PET-CT to assess the status of systemic disease. The median follow-up time was 13 months (CI: 95%, range = 13–65 months).

The results in terms of Local Control and survival index are showed in Table 4, those of toxicity in Table 5. Finally, the correlation between the LC, survival index and variables are showed in Table 6.

Primary and metastatic lung diseases often relapse, creating difficulties in the type of therapeutic approach to be use. The relapses may occur in different sites: in the same site or another site of lung or in the lymph node. It can relapse outside of lungs. Generally the first approach for lung disease recurrence is surgery or chemotherapy. Actually progression in technologies, especially SBRT, gives a new treatment approach, not only as palliative intent [9, 10]. There are few data about the efficacy, the doses, the number of fraction and the ideal time between the two treatments. Usually, the target of the first treatment is larger than the target of retreatment. More attention is focused about toxicity, involving more reasonable choice of doses and volume of retreated lesions [11]. The aim of this study was to evaluate the efficacy and tolerability of SBRT re-irradiations treatment of curative intent. We had a LC of 66%, with rates of 67% at 1- year and 54% at 2- years. The others studies showed rates of LC from 59 to 95% at 1- year and from 50 to 92% at 2- years [12]. Liu et al. at MD Anderson in a retrospective study have analysed 386 patients with primitive or metastatic lung disease treated by SBRT for a single lung recurrence, with a maximum diameter ≤ 4 cm. The dose was 50 Gy in 4 fractions, the local control was 95% at 1-year, the median follow-up was 16 months [13]. Kilburn et al. studied a group of 33 patients with primitive and metastatic lung disease treated by SBRT for in-field and out-field lesions. The dose was 50 Gy in 10 fractions. The local control at 2 years was 67%. The median follow-up was 17 months. They found greater efficacy in terms of local control with multiple fractions vs. single fractionation (Log Rank, p = 0.006) [14]. As know, the tumor radio resistance depends on the degree of hypoxia. Therefore, if we have more fractions we have more time to reduce the degree of hypoxia and improve the tumor response [12]. In our study, we observed a statistically significant correlation of single fraction vs. multiple fractions (p = 0.05), median cumulative EQD2(10) of the two treatments ≥167 Gy vs. < 167 Gy (p = 0.03) and median cumulative BED (10) of the two treatments ≥200 Gy vs. < 200 Gy (p = 0.02). Comparing these data with those of other studies, we observed discordance about local control results and number of fraction. A possible reason is that the lesions treated with single fraction are smaller than lesions treated with multiple fractions. We did not find any statistical correlation with volume of lesions. Regarding the dose, in terms of BED(10) and EQD2(10), our data confirm the data of Reyngold et al. They found a statistical correlation between LC and BED(10) ≥ 100Gy for the re-irradiation (p = 0.04) [4]. Trakul et al. showed good results on correlation between LC and BED when the time between the two treatments was greater than 16 months (p = 0.042) [15]. Reyngold et al. confirmed these good data with an interval greater than 36 months (HR = 0.25, p = 0.05). In addition, the same study showed good results on LC for a PTV volume < 75 cm³ (p = 0.03) and a KPS ≥ 80 (p = 0.03) [4]. Kilburn et al. found a correlation between PTV and LC (p = 0.03) [14]. The lesion volume is very important to define the radio resistance, because in a bigger volume there were a lot of hypoxic cells [7]. We did not find the same correlations in our study. In our study no major toxicity findings or fatal events were identified. The tolerance was good, with 59% of cases with no acute toxicity and the same percentage with no late toxicity. The maximum acute toxicity was one case of dyspnea G3. The maximum late toxicities were a case of dyspnea G3, a case of chest pain G3 and two case of pulmonary fibrosis G3. There were no correlation with dose, localization and volume of lesions. Kelly et al. showed a correlation between out-field lesions and Grade 3 pneumonitis (p = 0.03). However, they did not find correlation with volume, lesion localization (central or peripheral), dose or with history of lung disease. Patients treated for in-field lesions had more frequently chest wall pain and low rates of pneumonitis, while patients treated for out-field lesions had a higher frequency of pneumonitis and less probability of chest pain. Radiation pneumonitis is an inflammatory process. One hypothesis is that previously irradiated areas can be fibrotic, and less susceptible to the inflammatory process caused by a new radiotherapy. No clear explanation can be found for chest pain. One hypothesis is that the nervous damage, fractures and myositis were caused by the re- irradiation [16]. Trovò et al. found an association between the dose received from the heart and the incidence of radiation pneumonitis. Patients who developed grade 3 to 5 pneumonitis had a higher heart Dmax (mean Dmax = 27 Gy), p < 0.05 [17]. Liu et al. described pre-SBRT parameters that could influence the incidence of Grade 3–5: ECOG PS 2–3 (p = 0.008), a pre-SABR FEV1 ≤ 65% (p = 0.012), a V20 ≥ 30% in the composite plan (p = 0.020), and previous bilateral mediastinal PTV (p = 0.024) [13]. The limits of our study were a small number of patients, a lack of homogeneity in the first treatment and re-irradiation, and a missing algorithm able to sum the doses for OARs and target between the two treatments. We have not reported the results about the index of survival, the correlation with variables because patients and tumor characteristics were too heterogeneous and the study sample was too small to find statistically significant results.

In this study we showed that SBRT may represent a new chance of treatment of curative intent in the recurrence lung disease. Toxicity was acceptable. However, more knowledge and studies with a higher number of patients, and greater homogeneity of treatment and longer follow-up could help to understand better who is the best candidate to obtain optimal results in terms of LC and toxicity.