Optimized radiotherapy to improve clinical outcomes for locally advanced lung cancer

Radiotherapy (RT), at a total dose of 60–66 Gy over 6 weeks, combined with platinum-based chemotherapy [1, 2], is the standard of care for stage IIIB or IIIA patients with unresectable or inoperable disease [3]. However, the long-term outcomes are poor, with a 5-year overall survival (OS) rate of 15–35% for stage IIIA and 5–10% for stage IIIB patients [4].

RT dose escalation has been tested in several phase I and II trials with promising results in terms of tumour control and toxicity [5]. However, the RTOG 0617 phase III randomized trial showed a better median OS (28.7 months) in the standard dose arm (60 Gy in 30 fractions) than in the high dose arm (74 Gy in 37 fractions, 20.3 months) [6]. This unintended and negative result may be attributed to the greater cardiopulmonary toxicity associated with dose escalation and prolonged overall treatment time (OTT). Indeed, a loss of 1.6% per day in the absolute 3-year survival rate is observed, if the OTT extends beyond 6 weeks, due to rapid tumour repopulation [7]. Thus, to improve tumour control, different modifications of fractionation have been tested, with a meta-analysis showing a modest but significant benefit for RT schedules with reduced OTT [8]. However, acceleration using ancient RT techniques is associated with prohibitive toxicities, especially in the case of concomitant chemotherapy, limiting the deployment of these protocols [9].

Compared to 2D and 3D-RT, intensity-modulated radiotherapy (IMRT) allows for improved organs-at-risk sparing, owing to the high dose conformation to the target volume, thus reducing toxicity rates [10].

In this context, the aim of this study was to evaluate the clinical outcomes (toxicity, loco-regional control (LRC) and OS) in patients with locally advanced non-small cell lung cancer (LA-NSCLC) treated with an accelerated RT approach using IMRT.

The median follow-up was 44 months. At the time of analysis, 43 patients were dead, 42 had systemic recurrences, and 22 had loco-regional recurrences. Fifty-nine patients were treated using the SMART schedule, and fourteen using the H-RT schedule. Tumour coverage and dose constraints to the OARs (organs at risk) were respected, and there was no relevant difference between the two schedules (Table 3).

In the present study, we described our experience using two accelerated schedules of IMRT in locally advanced lung cancer patients and demonstrated the presence of a low rate of severe oesophageal and pulmonary toxicity, along with encouraging LRC and OS. We hypothesize that our results are associated with the dosimetric advantages of IMRT, and the reduced OTT obtained with the use of accelerated RT schedules.

Accelerated RT, by decreasing the OTT allows for increases in the biologically effective dose (BED). This approach shows an overall survival benefit but there are concerns pertaining to the increased toxicity compared to non-accelerated RT [8, 14]. In our study, thanks to the use of IMRT, the whole dose constraints were respected, in particular for the lung and oesophagus (Table 3). This explains the low rate of grade 3 and 4 pneumonitis (5% and 1% respectively), grade 3 oesophagitis (1%) and the absence of grade 4 oesophagitis. The pulmonary toxicity was correlated with a worse OS in the RTOG 0617 trial [6] and in our study. Interestingly, grade ≥ 2 pneumonitis occurred despite the respect of the dose constraints, highlighting the need of new volume constraint, as the integration of lung perfusion scintigraphy [15].

Compared to 3D-RT, IMRT allows for better sparing of the surrounding normal tissues [16, 17]. Moreover, the use of a 4D-CT for each patient to incorporate the respiratory motion allows to minimize the PTV, leading to a reduction in the dose received by organs at risk [18]. This dosimetric benefit is likely to translate to a decrease in toxicity [19, 20]. Indeed, compared to 3D-RT, IMRT allows to lower rates of severe pneumonitis [21] and a better quality of life [22]. Moreover, the use of IMRT translates to a decrease in the radiation treatment interruption and correlates to a better OS [23].

Another advantage of IMRT relies on the possibility of accelerating the irradiation dose to the high tumour burden alone, while maintaining moderate doses to the anatomical areas considered being at risk of microscopic invasion and areas of overlap between the target and OAR. Retrospective studies in patients with LA-NSCLC reported their experiences with the use of SMART schedules [24‐27]. The prevalence of grade 3 oesophagitis and pneumonitis ranged from 0 to 25.3%, and 1% to 11%, respectively, which is in line with our results. The 2-year LRC and OS ranged from 62 to 66%, and from 45 to 56%, respectively. In our cohort, the median OS was 27 months, with a 2-year LRC and OS of 62% and 54%, respectively.

In the present study, a small group was not eligible for concomitant chemotherapy, and these patients were treated with a more accelerated schedule (24 × 2.75 Gy) based on the promising results of a phase II trial evaluating the effect of cetuximab with concurrent chemoradiotherapy for LA-NSCLC. In this trial, patients were treated with a hypofractionated and accelerated schedule (24 × 2.75 Gy), and a concurrent daily low dose of cisplatin (6 mg/m²). Despite the non-benefit of cetuximab, the OS was remarkably high, with a median OS of 31.5 months and 1- and 2- year OS rates of 74.5% and 59.4%, respectively. However this benefit is penalized by a high rate of acute and late oesophageal toxicity [28]. For the frail patients in our study we adopted the H-RT schedule to compensate for the inability to receive concomitant chemotherapy. Moreover, this group of patients comprised 43% stage IV (Table 1), with a single metastasis, and having received first-line chemotherapy alone and/or a local treatment of the single metastasis, followed by exclusive chest irradiation only. Despite the poor prognosis of these patients (elderly patients and/or with comorbidities, stage IV) unfit for chemotherapy, the use of H-RT led to an OS and LRC that were similar to those observed in the SMART and concomitant chemotherapy group (Figs. 1 and 2), maintaining a comparable low toxicity level (Table 4), as dosimetric characteristics (Table 3).

The recent results of the PACIFIC trial point to a consistent benefit in the progression free survival when treatment using immunotherapy (durvalumab) is started within 1 to 42 days after the end of the concomitant chemoradiotherapy [29]. With the changing landscape in the standard treatment of LA-NSCLC, the reduction in treatment-induced toxicity, while maintaining optimal tumour control, has become a priority, thereby warranting access to adjuvant immunotherapy for these patients. The benefits of concomitant chemotherapy could be re-evaluated in this context. The use of immunotherapy concomitant to radiotherapy may be the next step [30]. Nevertheless, as immune cells are highly sensitive to conventional RT doses, the paradigm of the standard irradiation volumes should be reconsidered [31, 32]. In this context, the introduction of IMRT to spare lymphatic tissues and bone marrow deserves evaluation in prospective trials [33].

Our study has some limitations. Due to its retrospective nature, the prevalence of toxicity may be underestimated. The population was heterogeneous but constituted mainly by locally advanced stages and some selected stages IV (11%) and II (11%). Stage IV include only patients with one lesion in one anatomical site treated with radical intent as these patients are recognized with a better prognosis than multimetastatic patients [34]. Stages II include patients with bulky centrally located disease, non-operable and non-eligible to stereotactic radiotherapy.

Moreover, although our results pertaining to clinical outcomes are promising, this data only serve as a hypothesis generating resource.

The use of accelerated IMRT opens a therapeutic window for radiation dose optimization for lung cancer, as an alternative to classical dose escalation strategies. The encouraging toxicity profile and efficacy of this technique warrants further investigation in a prospective clinical trial, especially in combination with immunotherapy.