Video-assisted thoracic surgery versus stereotactic ablative radiotherapy: is the game over?

To date, available evidence assumes clinical equipoise between surgery and stereotactic ablative radiotherapy (SABR) in the management of early-stage non-small-cell lung cancer (NSCLC). However, there are very few long-term prospective data and no head-on randomized comparison between these 2 therapeutic options. Several randomized trials, including RTOG 1021 (NCT01336894), ROSEL (NCT00687986), STARS (NCT00840749) and SABRTooth (NCT02629458), have failed to accrue because of physician and/or patient preference to favour either lobectomy or SABR, according to the context of the study [1]. A pooled analysis of the STARS and ROSEL trials showed provocative higher survival after SABR than with lobectomy for operable early-stage NSCLC [2], but that analysis had some lethal limitations, one of them being the scarcity of contemporary surgical techniques such as the use of minimally invasive approaches. Current ongoing randomized controlled trials are designed to declare a winner, boldly powered for overall survival superiority of SABR versus sublobar resection for patients with high-risk operable NSCLC (STABLEMATES, NCT02468024), and for SABR versus lobectomy (VALOR, NCT02984761), their results being expected from 2024 to 2027.

Randomized clinical trials are postulated to be the golden standard in medicine, especially when assessing the safety/efficacy profile of a new therapeutic in comparison with the benchmark treatment. They are designed to prevent study bias and obtain accurate data relevant to the study outcome. Given that they operate under controlled conditions, they answer the questions only within a trial-defined framework and, as such, do not necessarily provide clinical relevance for the whole spectrum of those patients that we have to deal with daily. Therefore, real-world data are playing a distinct and increasing role in clinical research and healthcare decision-making.

Real-world data are an overarching term for observational data generated during routine clinical practice. Patient registries are organized systems that use observational study methods to collect uniform real-world data to evaluate specified outcomes for a population defined by a particular disease. The current global movement towards innovative and individualized healthcare promotes these patient registries as unique tools for collecting prospectively information on clinical practices, patient outcomes, safety assessment and clinical efficacy. This was the approach taken by de Ruiter et al. [3] in their article published in the present issue of the European Journal of Cardio-Thoracic Surgery.

They have built a cohort of 597 consecutive patients who underwent mini-invasive lobectomy or SABR for a clinical stage I NSCLC during a 2-year period, by merging databases from 12 Dutch hospitals. The primary goal of the study was achieved successfully. It consisted of assessing the feasibility to perform detailed data collection on a nationwide scale, as the first step of a more ambitious project aiming at involving more than half of all Dutch hospitals performing lung cancer surgery and/or SABR, during a larger period of investigation. As expected, the imbalances in baseline prognostic factors between the 2 treatment groups in this feasibility study cohort were obvious: patients treated with SABR were older, had a higher comorbidity index, a poorer performance status and poorer pulmonary function. Because these imbalances could contribute to explain the observed difference in unadjusted overall survival between mini-invasive lobectomy and SABR to the benefit of surgery, propensity score analyses were performed to overcome confounding factors and selection bias. In so doing, propensity score analyses turned underpowered to detect significant variation in outcomes, and this finding led to estimate that a three-fold sample will be required to assess relevant differences.

These methodologies, prospective cohorts versus real-world data, are not mutually exclusive and they can even be combined, as illustrated by the STARS revised study [4]. Noting their inability to recruit in the formal STARS study, the investigators devised a hybrid design mixing a prospective accrual of patients in the experimental SABR arm within the strict frame of the initial study, and a propensity score matching with data from a surgical cohort as the control arm. Thereby, they provided the current high-level evidence that for patients with operable stage IA NSCLC, long-term survival after SABR is non-inferior to mini-invasive lobectomy. But is the game over?

Despite the considerable progress made by the adoption of minimally invasive techniques and enhanced rehabilitation after surgery programmes, it will come as no surprise that patients will prefer during decision-making discussions a truly ambulatory procedure that is delivered in 3 or 4 fractions with a low impact on their quality of life [5]. Considering oncological effectiveness, surgery has the theoretical advantage of removing the anatomical territory in which the tumour has grown, with its lymphatic drainage and wide safety margins, thus leading to a complete resection, which is the best surrogate of long-term survival. Occult nodal disease, for instance, will only be detected in patients undergoing systematic lymphadenectomy. Large contemporaneous national databases, even in the era of modern pre-treatment staging methods, show a 10–20% incidence of nodal upstaging in patients with clinical stage I NSCLC and treated with lobectomy [6]. The golden rules of surgical radicality, however, are poorly adhered to in routine clinical practice. Real-world data show that a completeness of lymph node dissection as quality metrics is fulfilled only in a minority of patients with clinically staged N0 NSCLC, irrespective of the surgical approach, both in Europe and in the USA [7, 8].

During the past decades, surgeons have focused their time and energy to the legitimate quest for less invasive surgery, while wrongly neglecting the imperative needs to standardize surgical techniques and to stick to fundamental principles of lung cancer treatment. They have now to admit with humility that in patients presenting with stage IA NSCLC, it seems that there are 2 outstanding treatment options available for long-term survival. The promise of real-world data, as strongly suggested by de Ruiter et al. study, is that cumulating experience collected prospectively in large nation-based data would help at the identification of specific patient groups that are more eligible for either surgery or SABR, by performing comparative subgroup analyses for different disease and patient characteristics.

REFERENCES