Overview
The role of chemotherapy in the primary treatment of locally advanced head and neck cancer has become established in certain clinical settings. These include the definitive treatment of nasopharyngeal carcinoma and in organ preservation protocols for laryngeal and hypopharyngeal tumors. It also plays a role in the palliation of recurrent disease.

Research into breast, lung, prostate, and colon cancers benefits from large, relatively homogeneous study groups. HNSCC research, on the other hand, suffers from great disease heterogeneity and a relatively low incidence in the United States, factors contributing to confusion and conflicting reports regarding chemotherapy outcome (response and survival) and underscoring the importance of large multicenter randomized chemotherapy trials. Large-scale studies of laryngeal and nasopharyngeal cancers are demonstrating the importance of multicenter studies of specific HNSCC sites.

Prognostic Factors
Prognostic factors for chemotherapy response generally accepted as standard include performance status, nutritional status, degree of tumor differentiation, tumor burden (size of primary tumor, resectability, and extent of regional nodal metastases), disease stage, and primary cancer site. In head and neck cancer, however, correlative results are conflicting, and it is difficult in most cases to say which of these standard factors has significant effects on response. One clear exception relates to site; undifferentiated nasopharyngeal cancer is the most responsive and adenoid cystic salivary gland cancer the least responsive to chemotherapy. Recent work looking for additional prognostic factors in HNSCC has focused on immunologic studies, tumor DNA content, p53, thymidylate synthase, glutathione S-transferase, and GADD 153.

Chemotherapy Response
Comparing the results of HNSCC chemotherapy trials can be difficult because of inconsistent standards for response determination. Confusion regarding tumor measurements can arise as a result of superinfection from ulcerative lesions or, in studies in recurrent disease, from scarring after local therapy. The timing of response determination also varies. Clinical response determinations 4 to 6 weeks after therapy often describe persistent nonmalignant abnormalities (eg, fibrosis, edema) that resolve after longer follow-up. This is a problem with bulky tumors treated with chemoradiotherapy.

Complete response to chemotherapy is of overwhelming importance to a patient’s survival outlook. Response reporting in primary HNSCC trials differentiate between responses at primary sites and those at regional sites. Pathologic documentation of primary-site clinical complete response is critical, and the number and depth of biopsies must be rigorously defined. There is disagreement, however, as to the comparative importance of histologic and clinical complete responses. Much of the discrepancy in this area of research may relate to a lack of thoroughness in clinical staging and a lack of uniform specifications for the number and depth of postchemotherapy biopsies. Although critical in screening for new drug activity, partial responses generally do not increase survival in primary or recurrent disease.

Toxicities
Chemotherapy-related toxicities frequently encountered in HNSCC are reviewed briefly in the following discussions of each agent. They are classic toxicities seen with the use of these agents in other cancers. Toxicities that are not well established or that are unique to drugs not currently in common use are not mentioned here but are mentioned only in the applicable discussion of each drug.

The most frequent and dose-limiting toxicities with current chemotherapeutic approaches in HNSCC have been reviewed recently and fall into three areas: myelosuppression, mucositis, and the legendary toxicities associated with cisplatin. Myelosuppression is a well-known side effect of chemotherapy, and its mitigation by hematopoietic growth factors is well established.

Mucositis is a major limiting acute toxicity in the chemotherapy of HNSCC. This is very relevant to two approaches for controlling advanced HNSCC; concomitant chemoradiotherapy and chemical/pharmacologic modulation of 5-FU are severely limited by this toxic effect. How best to mitigate mucositis is not clear. Several strategies to limit chemotherapy-induced mucositis in HNSCC are under study, but no proved standard approach currently exists.

Problems with cisplatin toxicities (renal, otologic, neurologic, and gastrointestinal) may be solved largely by substituting carboplatin, a platinum analogue. Nevertheless, cisplatin will remain a leading agent against HNSCC when management is hampered by myelosuppression. The availability of specific serotonin receptor antagonists (eg, ondansetron and gramsetron) greatly enhances control of the major acute problem of cisplatin-induced nausea and vomiting.