Keywords

Introduction

In head and neck cancer (HNC) patients, locoregional recurrences occur in up to 50% in those initially presenting with locoregionally advanced disease [1]. Furthermore, 15% of patients present with a second primary tumor during their follow-up [2]. For these patients with recurrent or second primary HNC, surgery provides the greatest chance of long-term survival [3]. Unfortunately, only a minority of HNC patients with a locoregional recurrence or a second primary tumor is diagnosed with resectable disease. Therefore, these patients often end up receiving systemic therapy with a palliative intent [4]. To improve local control in this patient group, reirradiation has been proposed [5]. However, reirradiation, typically up to doses of 66–70 Gy in fractions of 2 Gy, comes with several challenges due to a high risk of severe toxicity [6]. Stereotactic body radiation therapy (SBRT) could represent a suitable RT technique in this situation [7]. Derived from intracranial stereotactic radiosurgery, the methodology was introduced to clinical practice by Lax and Blomgren at the Karolinska Hospital in Sweden in September 1991. Based on delivering precisely targeted high doses of radiation in one or several fractions, the concept of SBRT has rapidly gained acceptance and progressively spread around the world. SBRT has practical benefits in the reirradiation setting, but on the other hand, the high dose per fraction can be considered a risk factor associated with severe toxicity [8].

Besides the reirradiation setting, there is an increasing interest in the use of SBRT in the setting where HNC patients present with few metastases (oligometastatic disease). Apart from new systemic modalities leveraging the immune cells to target cancer, focus has been drawn towards local ablative approaches against metastases, in particular in oligometastatic patients [9]. Over the past few years, efforts have been undertaken to make a firm definition of oligometastatic disease and its different states [10, 11]. According to a recently published European Society for Radiotherapy and Oncology (ESTRO) and American.

Society for Radiation Oncology (ASTRO) consensus, the following two conditions must be met: a maximum of five metastases and all of them must be safely treatable [11]. Metastasectomy has been traditionally considered the gold standard in the oligometastatic disease setting when a local ablative approach was considered. However, in patients who are unwilling or unable to undergo an invasive procedure or deemed to be at high risk of postoperative complications due to underlying comorbidities, SBRT has emerged as a valid alternative to surgery.

This chapter aims to summarize the available literature regarding the use of SBRT in recurrent and second primary HNC as well as in oligometastatic HNC.

Reirradiation Using SBRT in Recurrent or Second Primary Head and Neck Cancer

Several studies have been published showing safety and efficacy of SBRT-based reirradiation. However, most of them included small patient groups with different inclusion and exclusion criteria and diverse fractionation schedules, making firm conclusions impossible and hampering the introduction of this technique in clinical practice in this setting. Recently, a meta-analysis was published comprising 10 papers [11,12,13,14,15,16,17,18,19,20] published between 2006 and 2016 [21]. The number of patients included in these studies ranged from 22 to 107. The majority had squamous HNC. The dose of SBRT-based reirradiation ranged from 24 to 44 Gy with a median value of 30 Gy, mostly delivered in 3–6 fractions. Median gross tumor volume (GTV) ranged from 19.1 to 103 cm3. Concerning the efficacy outcomes, median overall survival (OS) ranged from 8.6 to 16.2 months with a pooled median of 11.9 months. The pooled overall response rate was 61.7% (95% CI: 51.1–71.3). The 2-year local control (LC) rates ranged from 26 to 64%, and the pooled rate was 47.3% (95% CI: 33.1–62.1). The pooled grade ≥3 late toxicity rate was 9.6% (95% CI: 5.0–17.6) and grade 5 toxicity rate was 4.6% (95% CI: 2.4–8.6).

This meta-analysis demonstrated that, for patients with inoperable recurrent HNC or a second primary tumor in the HN region, SBRT-based reirradiation is a feasible therapeutic option with an acceptable severe toxicity rate below 10% and a good, pooled response rate of about 62%. Following SBRT-based reirradiation, the pooled 2-year OS rate was disappointingly only 30%. This is, however, in line with or even slightly better than standard reirradiation using longer schedules where a 2-year OS rate of 15–26% was observed [22] with a higher burden for the patient in terms of a treatment period reaching up to 7 weeks and severe late toxicity rates of more than 30%. On the other hand, SBRT is a much shorter treatment, typically administered over a period of 14 days or less.

There are several factors that might have influenced the observed OS rates. Some papers demonstrated that radiation dose and tumor volumes can affect OS following SBRT-based reirradiation [15, 20]. It is believed that high-dose SBRT is essential to achieve prolonged OS especially for recurrent tumors since they might harbour radioresistant tumor cells that were not eradicated by previous chemoradiation [23]. Furthermore, Vargo et al. reported that gross tumor volume of less than 25 cm3 was associated with increasing OS in comparison to larger tumors [22].

In the above-mentioned meta-analysis [21], reirradiation with SBRT appears to be safe with a pooled event rate of grade ≥3 complications of 9.6% and only three trials reporting rates of 10% or higher. Among the included studies, Vargo et al. reported grade 3 toxicity in 6% of patients and no grade ≥4 toxicities following 8 fractions of 5–5.5 Gy, Furthermore, Lartigau et al. found that 30% of patients experienced grade 3 toxicities following 6 × 6 Gy [11, 16].

Many studies in this field used SBRT-based reirradiation together with systemic therapy, hoping for a synergic effect and a better OS, however the used schemes were very heterogeneous [21]. Vargo et al. concluded that combining cetuximab with SBRT resulted in a 1-year OS of 40% [11]. Lartigau demonstrated that SBRT-based reirradiation with cetuximab is a valuable alternative to salvage surgery with a 1-year OS of 48% [16]. Recently, immune-checkpoint inhibitors pembrolizumab and nivolumab demonstrated durable antitumor activity for recurrent and metastatic HNC ineligible for RT or surgery both in the first line (Keynote-048 [4]) and second line (Checkmate-141, Keynote-012, and Keynote-040 [24,25,26]). Therefore, also for cases of recurrent HNC, it is necessary to continue investigating the combined therapeutic efficacy of systemic agents and local modalities such as SBRT.

SBRT in HNC Patients with Oligometastatic Disease

Numerous retrospective and prospective studies showed that SBRT can improve disease-free and OS in the oligometastatic setting while maintaining good treatment tolerance [27,28,29]. However, randomized data comparing SBRT with metastasectomy in operable patients are lacking. Moreover, covering different primary tumor types and organ sites, mostly outside the head and neck region, the available evidence remains difficult to interprete.

In HNC oligometastatic disease, we will discuss retrospective and prospective studies separately. Until present, the former group has provided quantitatively more data albeit with several important limitations inherent for this study type. The definition of oligometastatic disease has not been uniformly defined yet. This comes forward especially in retrospective analyses where different author groups may use different diagnostic criteria; and sometimes the designation “oligometastatic” may even be attributed to a given case only after a retrospective review of his or her medical records. Hence, the results should be interpreted with caution, particularly if comparing different publications.

Pasalic et al. evaluated 82 patients with head and neck cancers of different histological types presenting with either synchronous or metachronous lung metastases. Forty-three of the 82 patients had oligometastatic squamous cell carcinomas (1–3 lesions). One and 2-year local control was 96% and 90%, respectively, and 1- and 2-year overall survival 74% and 66%, respectively [30]. In their primary analysis, Bates et al. focused solely on oligometastatic disease, reporting 1- and 2-year OS of 78% and 43%, respectively, in 27 squamous HNC patients (3 had nasopharyngeal cancer and 3 unknown primary) with up to five synchronous and metachronous metastases mostly affecting the lungs but also other organs encompassing the bones, liver, lymph nodes, and soft tissues. Local control of treated lung nodules was 74% and 52% at 1- and 2-years, respectively [31]. A similarly large cohort was described by Bonomo et al. who evaluated 27 squamous HNC patients with solitary lesions in the lungs. The investigators achieved an objective response rate at 3 months of 75% with 1- and 2-year time to progression of 56% and 35%, respectively [32]. Finally, Franzese et al. collected data of 48 consecutive HNC patients with a maximum of 5 oligometastases in up to 2 organs. Forty percent of primary tumors were salivary gland cancers and nasopharyngeal carcinomas. Efficacy results were available for the whole cohort with 1- and 2-year local control rates of 83% and 70%, respectively [33].

In addition, we have learnt from retrospective studies that human papillomavirus (HPV)-positive oropharyngeal cancer patients are probably better candidates for local ablation, including SBRT, compared with viral-unrelated HNC. This holds true particularly for cases presenting with slowly and late developing lung oligometastases. Although such clinical presentation is rare, it has been associated with long-term survival after metastasis-directed therapy and has implications for post-primary treatment follow-up. The latter does not usually comprise imaging methods. However, HPV-related oropharyngeal carcinoma might be one of the exceptions requiring a more comprehensive surveillance [9].

Prospective evidence on SBRT in oligometastatic HNC is scarce. Only a few randomized phase II trials were conducted, and most reports are single arm studies with marginal number of HNC patients [34, 35]. Sutera et al. recruited 147 patients with up to five metachronous, biopsy-proven metastases visualized on FDG-PET/CT in at most three organs, comprising the lungs (>50%), lymph nodes, bones, and other sites. There was a large variety of primary tumors with more than half of them compromised of lung cancer (22%), colorectal cancer (21%), and HNC (11%). Because of an excess of early deaths, median OS of 17.6 months in 16 patients with HNC, out of which only 11 had squamous HNC, was inferior to that observed in other primary tumor subgroups. However, the 42% 5-year OS yielded in this cohort of 16 patients compares favorably to outcomes in surgical studies but can be biased by the very small patient number [36].

The only randomized trial exploring the addition of SBRT to a standard systemic palliative treatment according to primary cancer was the SABR-COMET phase II study. It was the first trial in oligometastatic disease to explore OS as the primary endpoint while previous randomized trials aimed at proving benefit in PFS. In SABR-COMET, oligometastatic state was defined by a maximum of five metachronous lesions with not more than three of them per organ (meaning that a maximum of 3 organs could be involved). Patients included were not considered candidates for metastasectomy. The three most frequently included primary tumors were breast, colorectal, and lung cancers (each about 18%), which were not balanced between the two study arms. The number of HNC patients was not specified except for a short comment in the supplementary materials on a case of oropharyngeal cancer treated for a lung metastasis. In the whole cohort of 99 patients, SBRT improved 5-year overall survival from 18 to 42%, however, at the cost of increased grade 2 or worse treatment-related toxicity (9% versus 29%) including grade 5 adverse events (0% versus 5%). Another noteworthy observation was the similar proportion of patients presenting with distant failure at untreated sites in the two study arms [37]. Therefore, despite the overall promising outcomes, the results of SABR-COMET testify the need for a proper patient selection to limit unnecessary toxicities and prevent indiscriminate use of local ablation in patients with few metastases, as some of them may in fact present with a widespread microscopic dissemination. A novel approach to this issue is to consider the speed of cancer development in the first place. It is characterized by rates of tumor-cell shedding and proliferation. Distant lesions with a slow speed of development, dubbed argometastases, may comply better to the need of identifying suitable patients for local ablation than if we rely solely on a low number of lesions [38].

Conclusion and Recommendations

SBRT reirradiation seems a promising modality for patients with inoperable recurrent HNC or second primary HNC providing acceptable safety along with short overall treatment times. OS following SBRT reirradiation remains moderate, which might be due to insufficient doses used in published studies. There is a need for well-designed trials of SBRT-based reirradiation in terms of dose escalation and combined treatment strategies with systemic agents. Currently, the largest body of evidence supporting SBRT-based reirradiation is furnished by the meta-analysis by Lee et al. evaluating this treatment in local and regional recurrences and second primary tumors [21]. Further studies in well-defined patient groups are necessary.

Based on the available literature, we can make the following recommendations for the use of SBRT in the reirradiation setting:

  • Only patients with small local or regional recurrences are good candidates; ideally the GTV should be below 25 cm3 [11].

  • 5 × 7 Gy or 6 × 6 Gy are commonly used schemes [21].

  • Following an R1 resection, 5 × 6 Gy should be prioritized [21].

  • Three fractions per week with the time interval between fractions of 48 h is recommended [16].

  • Overall treatment time should not exceed 14 days [11].

Another emerging role of SBRT is in the management of patients with oligometastases. Here, we need to understand which patients benefit the most, when the right moment is to intervene, how SBRT compares with surgery in operable patients and with other modalities (e.g. radiofrequency ablation), and what the impact of combination strategies is (e.g. with immune checkpoint inhibitors). Therefore, we advocate conducting dedicated studies for oligometastatic squamous HNC patients. One of the steps forward is the ongoing EORTC 1945 OligoRARE trial (NCT04498767) that investigates SBRT in addition to standard of care treatment in patients with oligometastatic rare cancers including HNC.