26 Systemic Therapies in the Management of Head and Neck Cancer

Several agents have demonstrated activity against HNSCC, including platinum compounds (cisplatin, carboplatin), taxanes (docetaxel, paclitaxel), and antimetabolites (5-fluorouracil [5-FU], methotrexate). Cisplatin, a potent radiosensitizer, is established as a standard agent in combination with radiation or coupled with other chemotherapeutic agents. It remains a preferred option over other platinum agents such as carboplatin, which has demonstrated improved tolerability at the expense of less activity in HNSCC. In 2009, a large meta-analysis was updated including 87 trials and more than 17,000 patients, evaluating the benefit of chemotherapy in HNSCC given as concurrent chemoradiotherapy (CRT), induction chemotherapy (ICT), or adjuvant treatment. The included trials compared locoregional treatment plus chemotherapy versus locoregional treatment alone. The total observed benefit from chemotherapy was an absolute 4.5% higher 5-year survival, confirming the benefits of chemotherapy in locally advanced HNSCC. A more pronounced absolute benefit of 6.5% was observed in trials of concurrent CRT, whereas there was no clear survival benefit seen for ICT or adjuvant chemotherapy.¹

The current treatment options utilizing systemic therapy will be reviewed here, as well as future directions including the utilization of targeted agents and immune therapy.

26.2 Induction Chemotherapy

The rationale behind the use of ICT prior to definitive therapy includes several considerations. One aim is to improve locoregional control, though in theory the additive systemic therapy may obviate the progression of metastatic disease and prevent distant relapses. It is indeed estimated that at least 50% of patients treated for locally advanced HNSCC will develop locoregional or distant relapse within 2 years of treatment despite receiving adequate local control with surgery and/or radiotherapy.² It is also pertinent to consider that increased tumor volume and hypoxic tumor volume may have a deleterious effect on local therapies, particularly radiotherapy, considering its dependence on oxygen-derived free radicals. This is supported by several studies of laryngeal cancer, which have demonstrated an inverse relationship between tumor size and local control with radiotherapy.^3,4,5 As such, in cases with relatively large tumor burden, ICT can be expected to reduce tumor volume and afford more adequate responses to locoregional therapy by enhancing radiosensitivity or resectability. Others have adopted the use of ICT as a basis for organ preservation, by inducing better responses to locoregional control with nonsurgical methods. By the same token, inadequate responses may permit salvage surgery to occur in a nonirradiated tissue environment and prevent complications such as fistula formation or poor wound healing.

As mentioned previously, several agents have demonstrated activity against HNSCC. Platinum-based regimens have generally shown the most anticancer activity and continue to be the most commonly used agents in the first-line setting. Many of the initial investigations in the 1980s and 1990s evaluating ICT followed by locoregional therapy could show a decrease in the rate of distant metastases, but were unable to consistently demonstrate a survival benefit. Cisplatin plus 5-FU became a standard regimen for ICT based on the observed high response rates and its ability to eliminate the need for surgical resection in some trials, though only a limited number of these were able to show a benefit in overall survival (OS).^6,7,8 Though the previously described meta-analysis of chemotherapy in head and neck cancer (MACH-NC) revealed only a modest and statistically insignificant OS benefit from ICT (2.4%; p = 0.18), it should be considered that these trials included a heterogeneous collection of chemotherapy regimens, and when limited to those utilizing platinum and 5-FU, a statistically significant hazard ratio (HR) of 0.90 was obtained.¹ Importantly, the analysis also demonstrated a meaningful reduction in the rate of distant metastasis, with an absolute difference of 4.3% (HR, 0.73; 95% confidence interval [CI], 0.61–0.88) at 5 years. Improvement in this parameter provided critical support for the idea that ICT can improve distant control, and implied a clear benefit for its addition to locoregional treatment. Though controversial and not universally accepted, these findings have continued to justify the investigation of ICT in HNSCC.

In the 1990s, taxane therapy (docetaxel or paclitaxel) began to garner immense interest as a new drug in the treatment for HNSCC. This was spurred in large part by a phase II trial which enrolled patients with recurrent, metastatic, or locally advanced incurable HNSCC to receive paclitaxel, noting an impressive response rate of 40%.⁹ The early success of taxane therapy inspired efforts to investigate its role in the setting of ICT. Ultimately, a benefit from the addition of a taxane to PF (cisplatin, 5-FU) regimens was demonstrated in three landmark phase III studies. The TAX323 study compared TPF (docetaxel, cisplatin, 5-FU) with PF as ICT in patients with locoregionally advanced, unresectable disease. In this study, 358 patients were randomized to receive TPF or PF followed by concurrent chemoradiation. At a median follow-up of 32.5 months, median progression-free survival PFS was 11.0 months in the TPF group compared to 8.2 months in the PF group (HR, 0.72; p = 0.007). Median survival was 18.8 months in TPF compared to 14.5 months with PF (p = 0.02).¹⁰ In the TAX324 study, 501 patients with locoregionally advanced disease were randomized to receive either PF or TPF. The TPF group achieved better locoregional control than in the PF group (p = 0.04). Median OS in the TPF group was 71 months, compared to 30 months in the PF group (p = 0.006). No significant difference was seen in the rate of distant metastases.¹¹ Lastly, the GORTEC group for organ preservation randomized 213 patients with larynx and hypopharynx cancer requiring total laryngectomy, to receive three cycles of TPF or PF. Those who responded to chemotherapy received radiotherapy with or without additional chemotherapy, and those who did not respond to chemotherapy underwent total laryngectomy followed by radiotherapy with or without additional chemotherapy. The 3-year larynx preservation rate was 70.3% with TPF compared to 57.5% with PF (p = 0.03), showing a superi- or response rate and higher likelihood of larynx preservation with the TPF regimen.¹² Together, these trials created a new standard for ICT by demonstrating that the inclusion of a taxane (TPF) was superior to the previously established PF regimen. A subsequent meta-analysis including 1,772 patients comparing PF with TPF supported this notion, showing an absolute survival benefit at 5 years of 7.4% (HR, 0.79; 95% CI, 0.70–0.89; p < 0.001), as well as a significant reduction in progression, locoregional failure, and distant failure with the use of TPH compared to its PF counterpart as ICT.¹³ These findings came with an important criticism that they were comparing two experimental arms, as ICT was not established as a standard treatment. Moreover, questions remained based on the aforementioned MACH-NC meta-analysis, which failed to demonstrate a meaningful survival benefit from the use of ICT. It is worth noting, however, that these trials did not include TPF regimens.

Despite these criticisms, the enthusiasm for ICT and the success of these trials ultimately provided the basis for its measurement against concurrent CRT alone as definitive therapy for HNSCC. In a phase III trial conducted by Hitt et al, patients with locally advanced HNSCC were assigned to ICT with either docetaxel, cisplatin, and 5-FU (TPF) or cisplatin and 5-FU (PF) followed by concurrent chemoradiation (cisplatin + RT) or chemoradiation alone. The results showed a median PFS of 14.6, 14.3, and 13.8 months in the TPF + CRT, PF + CRT, and CRT groups, respectively (95% CI, 11–17.5; p = 0.56).¹⁴ They were unable to demonstrate statistically significant differences in time to treatment failure or in OS. In the PARADIGM trial, a phase III study, 145 patients with untreated, nonmetastatic head and neck cancer were randomized to receive ICT with TPF followed by concurrent CRT (weekly carboplatin + RT) or concurrent CRT alone (with standard two doses of cisplatin + RT). Three-year OS was 73% in the ICT + CRT group and 78% in the CRT alone group (HR, 1.09; 95% CI, 0.59–2.03; p = 0.77).¹⁵ The DeCIDE trial was a phase II trial comparing ICT with TPF versus CRT in patients with locally advanced HNSCC and high nodal stage disease (N2 or N3). At 3 years, no significant difference in OS was found, noting 72% in the induction arm and 69% in CRT arm (p = 0.69). There was a difference in rate of distant failure at 3 years with ICT (10 vs. 19%) but this difference was not statistically significant (p = 0.11).¹⁶ A meta-analysis in 2014 evaluating ICT followed by CRT concluded that ICT with TPF before CRT does not improve OS (HR, 1.008; 95% CI, 0.816–1.246; p = 0.94) and suggested only a mild, insignificant benefit in PFS (HR, 0.881; 95% CI, 0.723–1.073; p = 0.207).¹⁷ Of note, the slow rate of accrual in both studies led to premature termination after far less of the expected patients were enrolled (285 of expected 400 in PARADIGM, and 145 of expected 300 in DeCIDE). Therefore, though no difference in survival was noted between those treated with ICT followed by CRT and those receiving CRT alone, the deficiency of statistical power made it inherently difficult to detect any benefit to ICT.

In summary, the role of ICT prior to locoregional therapy in HNSCC remains controversial, and no consensus guidelines are available to guide its use. Despite the negative findings from these pertinent trials and meta-analyses, flaws in methodological design and heterogeneity in chemotherapy regimens employed limit any conclusive answer to the question of whether ICT has a definite role. Until stronger evidence is available, its use in locally advanced disease should be limited to unique scenarios and clinical trials, while concurrent CRT remains the standard of care. Some experts suggest considering ICT when delays in CRT therapy are expected, as a larynx-preserving approach, and in patients who are at high risk for distant relapse (such as those with N2c or low-neck disease). It is important to consider that in addition to the lack of efficacy achieved in the DeCIDE and PARADIGM trials, sequential treatment also caused more toxicity than CRT alone. In DeCIDE, severe adverse events were significantly higher in the ICT arm (47 vs. 28%, p = 0.002).¹⁶ In PARADIGM, the incidence of grade III–IV febrile neutropenia was 23 versus. 1%, and grade III–IV mucositis was 47 versus 16% in the ICT arm.¹⁵

26.3 Concurrent Chemoradiotherapy

The primary benefits of chemotherapy in the setting of nonmetastatic HNSCC stem from its utility as a radiosensitizing agent. Therefore, the focus has been on agents with known activity in head and neck cancer, as well as established radiosensitizing properties including cisplatin, cetuximab, and 5-FU. The use of radiotherapy has been established with two main strategies, including concomitant chemotherapy (single or multiagent) with continuous radiotherapy or multiagent chemotherapy with split-course radiotherapy. Other acceptable approaches now include radiotherapy with altered fractionation. As mentioned, historically, chemotherapy was used exclusively for unresectable disease. The demonstration that chemotherapy could improve locoregional control compared to radiotherapy alone ultimately led to its study in other groups of patients including those with resectable disease and with organ preservation intent, and as adjuvant therapy after surgical resection in those with high-risk features. In 1992, Merlano et al randomized 157 patients with untreated, unresectable HNSCC to receive either cisplatin plus 5-FU plus radiotherapy or radiotherapy alone, and showed an increase in the frequency of complete response (43 vs. 22%, respectively; p = 0.037), increased PFS at 5 years (21 vs. 9%, respectively; p = 0.008), and increased 5-year survival benefit (24 vs. 10%, respectively; p = 0.01) in the combined treatment group.¹⁸

Stay updated, free articles. Join our Telegram channel

Join