Lung reirradiation with stereotactic body radiotherapy

Abstract

Purpose

We examined safety and efficacy of stereotactic body radiotherapy (SBRT) in reirradiation for lung recurrent lesions (LRLs).

Materials and Methods

Eighteen patients, 4 with lung local failure from primary non-small cell lung carcinomas and 14 with lung metastases, were reirradiated with SBRT for 29 LRLs. Doses were recalculated to an Equivalent Dose of 2 Gy per fraction (EQD₂) and α/β ratio was assumed to be Gy₁₀ for primary and metastatic lung tumors and Gy₃ for organ at risk. Cumulative administered doses were calculated adding doses of prior radiotherapy and reirradiation.

Results

Peripherally located lesions received 5 fractions of 8-10 Gy, while centrally ones lower doses (5 fractions of 5-8 Gy). Cumulative EQD₂ did not exceed 198 Gy₁₀ and reirradiated volumes were rather small (median 18 cc). Local control was obtained for all patients except one and lasted medially 43 months. Median overall survival was 40 months from reirradiation. Only acute grade 1 toxicity was recorded.

Conclusions

Reirradiation of LRLs with SBRT was feasible and effective. It is important to appropriately select patient and to adopt organ at risk constrains considering cumulative doses.

1. Introduction

Despite surgery, definitive radiotherapy and/or multimodal treatment of primary lung tumors, local recurrence is frequent within previously irradiated regions or in other lung areas (1-4). The lung is also a frequent site of metastasis from various tumors. The concept of oligometastases was recently introduced to describe an intermediate state of cancer spread between localized disease and widespread metastases (5,6). Evidence has showed that patients with limited metastatic disease, such as single lung metastasis, may be cured with local therapy, such as surgery or radiotherapy. Consequently, the treatment of recurrent lung lesions in patients who have undergone previous irradiation of same lung from primary or metastatic disease is a common dilemma facing oncologists today (5,6).

Jeremic et al. reviewed 11 studies of conventionally fractionated conformal radiotherapy (3D-CRT) for lung reirradiation, and shown an improvement in overall survival when high doses were used with respect to low-doses. Unfortunately, a high rate of grade 2-3 toxicity was also registered (3). Stereotactic body radiotherapy (SBRT) can delivery in single or 3 to 5 fractions high biologically effective doses to tumors while minimizing the dose to surrounding tissues (7-9). This approach has been adopted with good results for early-stage, inoperable lung cancer (NSCLC) and for lung metastases (4,7-9). However, there are few reports with limited number of patients which have evaluated the use of SBRT for reirradiation of lung recurrent lesions (LRLs) after a prior 3D-CRT (4,10,11) or SBRT (12-14).

In this retrospective study, we present the results of our institutional experience on reirradiation of LRLs with SBRT.

2. Material and methods

Patients with LRLs who were reirradiated with SBRT entered the study. Lung recurrent lesions were considered lung local failure from NSCLC and lung metastases from NSCLC or other primary tumors. Regarding NSCLC, recurrences of disease in the same pulmonary lobe were defined lung local failure otherwise lung metastases.

Lung recurrent lesions were defined as “in-field relapse” when more than 95% of the recurrence volume was within the original 50% isodose of radiotherapy, and as “out-field relapse” in the other cases, provided that it was in the same lung.

For reirradiation with SBRT the dose was prescribed to the isocenter and the minimal coverage accepted dose was 90% with a maximum dose not exceeding 110%. A stereotactic body frame with a rigid body fixation system and a fiducial box were used. The localization was obtained using a multi-slice computed tomography (CT) in free quiet respiration (15). The use of contrast was evaluated case by case. Gross tumor volume (GTV) was defined as the radiologically visible tumor using lung or mediastinal windowing for peripherally or centrally located lesions, respectively. Clinical target volume (CTV) was coincident with GTV and planned target volume (PTV) was designed as GTV/CTV plus an additional margin of 8-10 mm in the cranio-caudal and 4-5 mm in the axial directions. After dosimetric and physic calculations for each treatment, the GTV/CTV was re-contoured in the image slices obtained with a verification CT (i.e., PTV must include both GTVs/CTVs using the image fusion of the first CT and the verification CT) (15). Coplanar dynamic arcs were conformed around PTV, typical field-shape margin was 2 mm, and micro-multileaves (MLC) set-up changed every 10 degrees to follow the possible variations on target profile through the beam eye view system. Treatment was delivered with a 5-MV X-rays linear accelerator with external dynamic MLC. Before each SBRT fraction, accuracy of treatment was evaluated with daily MV portal imaging. For each reirradiation an image fusion with the first radiotherapy was performed, and previous target was recontoured in the new plan.

All doses were recalculated to an Equivalent Dose of 2 Gy per fraction (EQD₂), with the formula: nd [(1+d/α/β)/(2+α/β)], with d = fraction dose (Gy), n = number of fractions, nd = D = total physical dose (Gy), and α/β = tissue repair capacity (Gy) with d the dose per fraction (Gy) and n the number of fractions. The α/β ratio was assumed to be 10 for an early reacting tissue such as the tumor (Gy₁₀), and 3 Gy for a late reacting tissue such as the lung (Gy₃).

For “in-field relapse” cumulative EQD₂ was calculated adding 100% of prior radiotherapy prescribed dose to 100% of reirradiation dose. For “out-field relapse” cumulative EQD₂ resulted from 100% of reirradiation dose plus prior radiotherapy doses resulting from isodose encompassing retreatment volume. For dose administered to organs at risk (i.e., spinal cord and esophagus) cumulative EQD₂ was calculated adding doses of prior radiotherapy and reirradiation.

Response was assessed with CT using Response Evaluation Criteria in Solid Tumors (RECIST; 16) and/or with positron emission tomography (PET)-CT with PET Response Criteria in Solid Tumors (PERCIST 1.0; 17). Local control was achieved if there was a lack of progression (i.e., any response and stable disease) and was defined as the time between retreatment and progression of the lesion.

Seven patients were followed by CT and 11 by PET-CT 3 months after the end of retreatment, and at 4 months interval thereafter with clinical evaluation and the same radiological exams.

Acute toxicity was scored according to the National Cancer Institute Common Terminology Criteria for Adverse Effects Version 4 (CTCAE v 4.0). Late toxicity was recorded according to the Radiation Therapy Oncology Group-RTOG and European Organization for Research and Treatment of Cancer Late Radiation Morbidity Scoring. Pulmonary function tests were performed before the start of treatment and repeated at first follow up and at 1 year.

A software package (MedCalc 11.1 Broekstraat 52, B-9030 Mariakerke Belgium) was used for statistical analysis. Overall survival and duration of local control were estimated for the entire population using the Kaplan-Meier product-limit method (18).

3. Results

Between January 2003 and April 2013, 18 patients already irradiated to the lung were reirradiated for LRLs, 4 had lung local failure from NSCLC and 14 lung metastases from NSCLC or other primary tumors (i.e., colon, breast, and endometrial carcinomas).

Patient characteristics are detailed in Table 1. The 18 patients included 14 men and 4 women with a median age of 68 years. The majority of cases had a good performance status, three quarter of lesions were in field relapses and time between first radiotherapy and reirradiation was quite long (median 18 months).

Table 1.

Patient characteristics

Characteristics	Number of patients	Number of lesions	Percent
Total cases	18	29
Sex male female	14 4		78 22
Age median range	68 years 53-84 years
Karnofsky performance status median range	100 70-100
Reirradiated lesions Local failure from NSCLC Metastases from NSCLC colon breast endometrium	4 14 9 3 1 1	4 25 16 7 1 1	14 86 64 28 4 4
Reirradiation field in-field out-field	14 4	23 6	79 21
Tumor location central peripheral		15 14	52 48
Time to reirradiation median range	18 months 6-90 months
Number of retreated lesions median range		1 1-4
Follow-up median range	57 months 6-132 months

Legend: NSCLC = non-small cell lung cancer

At the time of reirradiation with SBRT, 4 patients presented only LRLs, and remaining 14 had radiological documented extra-pulmonary metastatic disease (PET-CT and CT diagnosis in 11 and 7 cases, respectively) but were thought to have stable disease except for the foci targeted by SBRT. Patients with extra-pulmonary disease were treated also with systemic therapy. Of the 9 lung cancer patients, 5 received targeted therapy and 4 platinum-based chemotherapy. The 3 patients with colon cancer received 5FU-based chemotherapy, the patient with breast cancer taxane-based chemotherapy, and the woman with endometrial cancer a combination of carboplatin and taxol.

Only 6 of 29 (21%) LRLs were “out-field relapses” in the same previously irradiated lung, whereas remaining 23 (79%) LRLs were “in-field relapses”. Table 2 summarizes administered doses by number, site and volume of reirradiated LRLs. Eleven (61%) patients were submitted to a curative prior SBRT at 40-50 Gy in 5 fractions and seven (39%) a prior 3D-CRT which was done with palliative intent at various doses in 3 cases, radical intent in 3 cases with 60 Gy in 30 fractions and 54 Gy in 27 fractions, in 1 case an adjuvant treatment with 50 Gy in 25 fractions. It is worth to note that 11 (61%) patients were reirradiated for only one lesion, whereas in 5 (28%) and 2 (11%) patients reirradiation was done in 2 and 4 lesions, respectively.

Table 2.

Administered doses by number, site and volume of reirradiated lesions

N. of reirradiated patients (%)	N. of reirradiated lesions	First RT doses	N. of patients	Reirradiation doses	N. of patients	EQD2 (Gy10)	Site of lesions	Reirradiated PTV (cc)
11 (61)	1	5x8Gy °       5x10Gy °     10x3Gy* 30x2Gy* 5x4Gy * 2x8Gy*	5       1     1 2 1 1	5x8Gy 5x6Gy 5x5Gy   5x8Gy     5x8Gy 5x8Gy 5x6Gy 5x7Gy	1 2 2   1     1 2 1 1	120 100 91   144     93 90 64 74	C C/P C/P   C     P C/C C C	51 13/22 38/55   12     36 17/18 39 51
5 (28)	2	5x8Gy °           5x10Gy °     27X2Gy*     25X2Gy*	2           1     1     1	5x8Gy 5x8Gy#   5x10Gy 5x10Gy #   5x8Gy 5x6Gy #   5X10Gy 5x8Gy #   5X9Gy 5X9Gy #	1     1     1     1     1	71^ 60^   84^ 84^   110^ 124   138 198   121 71^	C P   P P   C P   P P   P p	15 18   17 17   10 43   26 11   27 14
2 (11)	4	5x8Gy °         5x10Gy °	1         1	5x5Gy 5x5Gy # 5x5Gy # 5x4Gy #   5x10Gy 5x8Gy # 5x6Gy # 5x6Gy #		91 91 91 84   168 144 124 124	C C C C   P C P C	50 16 14 28   17 8 22 19

Legend: stereotactic body radiotherapy °, conformal radiotherapy *, out field lesion ^, two or more reirradiation cycles #. RT = radiotherapy, EQD₂ = total equivalent dose in 2 Gy fraction, C = central, P = peripheral, PTV = planning target volume.

Reirradiation doses were decided according to tumor location, peripherally located lesions received 5 fractions of 8 Gy to 10 Gy, while centrally ones received lower doses (i.e., 5 fractions of 5 Gy to 8 Gy). Cumulative EQD₂ did not exceed 198 Gy₁₀ and PTVs of reirradiated lesions were quite limited (median PTV = 18 cc, range 8-55).

From the calculation of summed plans, the median cumulative maximum EQD₂ for spinal cord was 23 Gy₃ (range, 2-48) and for esophagus 45 Gy₃ (range, 4-138).

At time of the analysis, all 18 reirradiated patients were evaluable for response and survival. Ten patients were alive (2 with complete response, 2 with stable disease, and 6 with progression), the remaining 8 were death (6 for distant progression, 2 for both local and distant progression).

Complete response was registered in 11 (38%), partial response in 7 (24%), stable disease in 10 (34%), and progression in 1 (4%) LRLs. Local control was obtained in all but one patient (96%). After a median follow-up of 57 months (range, 6-132), the median duration of local control was 43 months (range, 2-97) with 1-year and 2-year local control rates of 82% and 66%, respectively. Median overall survivals were 40 months (range, 10-51) from reirradiation, and 64 months (range, 25-143) from first radiotherapy (figures1 and 2).

Acute grade (G) 1 toxicities (i.e., dysphagia, thoracic pain, and asymptomatic pneumonia) were recorded in only 3 patients. There were neither acute G 2–5 toxicities, nor late toxicity.

No examined variables (i.e., gender, age, performance status, primitive versus metastatic lung tumor, central versus peripheral recurrence, administered doses, 3D-CRT versus SBRT radiotherapy, “in-field relapse” versus “out-field relapse”, number of reirradiated lesions, time between first irradiation and reirradiation) significantly conditioned response, survival and/or toxicity.

4. Discussion

In the lung reirradiation, SBRT can be a useful treatment to give high precision dose to the tumor limiting radiation-induced side effects. This was recently confirmed by a review of Mantel et al. summarizing the available studies on lung reirradiation with 3D-CRT or SBRT (19). From this review, SBRT resulted superior to 3D-CRT permitting a higher control rate with lesser toxicity.

An important consideration emerging from our experience is that, although reirradiation assured a good local control rate (96%), complete response was registered in only one third of cases (38%). This result could be explained by the relatively low prescribed doses (40-50 Gy in 5 fractions equal to 60-84 Gy₁₀). This hypothesis seems to be confirmed by Kelly et al. (4) who obtained a higher complete response rate using higher doses (50 Gy in 4 fractions, equal to 94 Gy₁₀).

In Trakul et al. paper (20), local control was significantly worse for tumors reirradiated before 16 months (46.7%) from first treatment, compared with those reirradiated after a longer interval (87.5%). Time to reirradiation as other examined variables (i.e., gender, age, performance status, primitive versus metastatic lung tumor, central versus peripheral LRLs, “in-field relapse” versus “out-field relapse”, number of reirradiated lesions) did not condition our results. Probably we have not registered statistically significant differences in outcome between aforesaid variables because of the quite limited number of patients recruited.

We registered a median overall survival of 40 months (range, 10-51) from reirradiation, and 64 months (range, 25-143) from first radiotherapy, values which are better than others reported in literature (13,14). These results can be due to the more accurate patient selection. In fact, at the time of reirradiation with SBRT, 4 patients presented only LRLs and 14 had LRLs plus stable extrapulmonary disease. Additionally, all patients had a high performance status and the majority of cases underwent hormonal-therapy and/or chemotherapy.

Apart from local control and survival, the main topic on lung reirradiation remains the risk of radiation-induced toxicity. After lung reirradiation a high grade of toxicity for centrally located lesions was reported by Peulen et al. and Trovò et al. (12, 21). Peulen et al. registered G3-4 toxicities in 27% of patients and G5 toxicity (i.e., massive bleeding with death) in other 3 (10%). Patients with severe toxicities had central large tumors with a median PTV exceeding 100 cc, and some of those were retreated 3 or 4 times to a very high cumulative radiotherapy doses. So, in the Peulen’s series, severe lung toxicity was probably due to reirradiation of large central volume with very high dose (12). In the Trovò’s experience, 4 patients (23%) experienced G3 radiation pneumonitis, and 2 (12%) G5 lung toxicity (fatal pneumonitis and hemoptysis) after reirradiation with SBRT of previous primary central tumors treated with standard fractionated radiotherapy at 50-60 Gy in 20-30 fractions. Probably, the relatively high incidence of iatrogenic deaths registered can be related to the central location of all relapsing lesions, the large target volume of the first treatment (i.e., tumor and involved nodal irradiation), and the relatively high cumulative doses administered.

Figure 1.

Overall survival probability from reirradiation.

Figure 2.

Cumulative survival probability from first radiotherapy.

In our study, LRLs were all in the same previously irradiated lung, the majority (79%) of LRLs were in-field, and half of them (59%) were centrally located. Nevertheless, only acute G1 toxicities were registered (1 dysphagia, 1 thoracic pain, and 1 asymptomatic pneumonia). Our findings could be explained by the relatively small volume of reirradiated PTV (median 18 cc), the limited cumulative administered doses which did not exceed 144 Gy₁₀ for central lesions and 198 Gy₁₀ for peripheral ones, and limited median cumulative maximum doses administered to spinal cord 23 Gy₃ and esophagus 45 Gy₃. Aforesaid doses are coherent with constrains reported by Meijneke et al. and Kilburn et al. to limit iatrogenic toxicity (13,14).

In conclusion, reirradiation of LRLs with SBRT was feasible and effective. It is important to appropriately select patient and to adopt adequate organ at risk constrains which account the sum of radiotherapy plans for evaluation of cumulative doses.