It has been shown that systemic therapy can be expected to achieve meaningful response and thus reduction in tumor burden ranging from 10% to 30% of cases of advanced recurrent disease when patients have undergone significant prior therapy although the effect of therapy is temporary (median time to progression 6 months, survival <12 months) (15,16,17). Higher rates of disease response can be achieved with combination of more than one systemic agent compared to monotherapy, although such response typically occurs at the expense of higher toxicity and morbidity either without substantial improvement or only modest improvement in overall survival (15,16).
Such impact may not be desirable when the intent of therapy is symptom palliation particularly in situations where declining patient functional status may preclude aggressive measures. As such, it is not unreasonable to utilize systemic monotherapy in such situations, although multiagent therapy should be considered particularly when patients possess good performance status as there is some evidence that multiagent therapy can achieve improvement in specific parameters such as pain and swallowing without negatively impacting overall global QOL measures (14). Commonly utilized systemic agents used either as monotherapy or in combination include methotrexate, platinum compounds (cisplatinum, carboplatinum), 5-fluorouracil, taxanes (paclitaxel, docetaxel), as well as in recent years targeted therapy strategies using agents such as cetuximab which target the epidermal growth factor receptor pathway. Although symptomatic benefit is more likely in patients who achieve measurable disease response to therapy, interestingly, symptomatic improvement can also be observed despite absence of overt response to treatment albeit in a smaller proportion of patients (13).

In the setting of advanced recurrent locoregional disease, the use of radiotherapy in palliative format can be considered, although in most cases of advanced recurrent disease, prior radiotherapy in affected regions may preclude its use. The use of reirradiation has been utilized in situations in an effort to reduce tumor burden and achieve disease response with variable success using varying methodology involving standard fractionated, altered fractionated, stereotactic/conformal delivery, or brachytherapy delivery methods (18,19,20,21,22). It must be understood that even with altered delivery methods reirradiation alone is associated with low overall response rates and disease control in previously treated regions. Furthermore, despite demonstrated feasibility, the rate of severe treatment-related toxicity (i.e., grade 3 or greater) of reirradiation is significant (21,22). As such, in light of the low rate of observed disease response and significant treatment toxicity, the role of palliative reirradiation in achieving meaningful symptom relief in patients with otherwise incurable recurrent HNC remains unclear (21).

The use of reirradiation in combination with systemic chemotherapy (with or without surgery) has also been utilized in patients with advanced locoregional recurrence with sustained disease response documented in a small but significant proportion of patients (23,24,25). It must be understood, however, that long-term survival in this patient group remains low, and the morbidity and mortality associated with such measures are not small (23,24,25). Furthermore, the intent of such treatment should be considered definitive in nature, and as such, it would be difficult to categorize such treatment as palliative in intent. In situations of distant disease such as bony or central nervous system (CNS) metastases, it is reasonable to consider palliative radiotherapy to these sites in situations of severe symptoms or high potential for debility from disease (e.g., bone pain, impingement upon spinal cord/CNS) (26).

The use of surgery to reduce tumor burden in palliative fashion when meaningful chance of cure does not exist is controversial. In the situation of advanced recurrent head and neck malignancy, surgery is often not feasible or practical due to disease extent (i.e., unresectable) or patient factors (medical debility, poor nutritional/performance status). Even if feasible, surgery often carries substantial morbidity with high risk of postsurgical wound complications and substantial potential for negative impact upon functions including speech, swallowing, and respiration and as such is frequently associated with need for adjunctive devices such as tracheostomy and/or feeding tube which may be either at least temporary or may be long term or permanent in a significant proportion of patients (27). Furthermore, in patients previously treated for advanced disease, surgery for recurrent disease is associated with low chance of long-term disease control (27,28). As such, the benefit of surgical measures must always be carefully weighed against the potential substantial negative impact upon function and QOL particularly when odds of meaningful disease control are low and treatment intent is palliative. It is not unreasonable however to consider surgical efforts—even aggressive efforts—in situations where symptoms, QOL, and longevity are primarily impacted by locoregional disease provided that disease is felt to be reasonably addressed using surgery (with or without further adjuvant measures) and potential for recovery of function is substantial (i.e., via reconstruction and/or rehabilitative methods). It is also reasonable to consider palliative surgical measures in situations where distant disease may exist yet have indolent behavior provided that locoregional disease can be controlled where other methods such as chemotherapy and/or radiotherapy carry poor odds of reducing locoregional disease burden (e.g., adenoid cystic carcinoma, well-differentiated thyroid cancer). Such considerations must be taken in concert with careful surgeon assessment of the feasibility and functional impact of surgical treatment and honest discussion with patients and families regarding realistic expectations of any such proposed measures particularly with regard to symptom control and QOL.