Neoadjuvant (induction) therapy prior to surgery is sometimes referred to as sequential chemoradiotherapy when chemotherapy is followed by definitive radiotherapy. This latter approach specifically excludes or limits the extent of surgery, except in treatment failures, and is growing in importance as an “organ preservation” approach for locally advanced resectable disease. The neoadjuvant concept has been expanded to include novel studies of neoadjuvant chemotherapy followed by concomitant chemoradiotherapy.
The use of chemotherapy as induction treatment prior to local therapy for advanced HNSCC has been evaluated for about 20 years. The strong rationale for this approach comes from mathematical models of cell kinetics and acquired drug resistance and from preclinical in vivo data. The principal objectives of induction are as follows: to promote regression and thereby enhance subsequent locoregional therapy; to identify patients with responding lesions that might be controlled by more conservative local treatment (organ preservation) rather than by extensive surgical procedures; to use chemotherapy before tumor and normal vasculature is altered by surgery or radiation; to identify responding tumors that may benefit from adjuvant chemotherapy following surgery or radiotherapy; and to provide early treatment for micrometastatic disease.
Early nonrandomized neoadjuvant trials based their selections of agents on trials in recurrent disease. Single-agent early trials achieved response rates of 30% to 40%. Gradually, neoadjuvant studies expanded to include multiagent therapy including cisplatin plus infusional 5-FU, which produced response rates of 70% to 80% (CR rates of 30% to 40%) but no long-term benefit. More recently, attempts to develop more active therapy include the integration of taxanes and regional therapy.
Over 30 randomized multiagent trials of neoadjuvant therapy with or without adjuvant chemotherapy compared with local therapy control groups have been reported to date. Although none have shown that neoadjuvant therapy significantly improves survival, many of these trials had major design limitations. These limitations include suboptimal patient numbers, relatively inactive induction regimens (median complete response rate < 20%), variations in locoregional therapies (surgery alone, surgery and radiotherapy, radiotherapy alone), heterogeneous study populations with multiple primary sites, and resectable and unresectable cases (
Table 90-27). None of the neoadjuvant regimens produced a CR rate of greater than 50%. Most of the randomized studies had such small study numbers (fewer than 75 patients per arm) that they could not detect significant survival differences below 25%.
Organ preservation is an important goal of primary chemotherapy. Those patients fortunate enough to survive their cancer often face a lifetime of significant morbidity because of cosmetic and functional debilities from surgical resection. Despite marked advances in reconstructive surgery and rehabilitation, patients who have undergone laryngectomy, glossectomy, or composite resection still have major debility. Unfortunately, the importance of organ and function preservation often is ignored. Many opt to receive the less effective therapy of radiation alone to avoid curative surgery. They risk shorter survival rather than face increased survival with severe surgical morbidity. An active area of HNSCC research involves chemoradiotherapy (sequential or concomitant) to facilitate organ-preserving approaches for advanced disease. This research is designed to preserve anatomic structures without compromising survival. In such studies, extensive surgery is reserved to salvage patients who fail to respond to initial therapy or whose disease recurs.
Laryngeal preservation is the most advanced application of organ-preservation chemotherapy. In advanced laryngeal SCC (T3 or T4, and/ or nodal metastasis), total laryngectomy improves survival at the price of impaired speech, and up-front irradiation preserves the voice at the price of a decreased chance of survival. Survival remains low with strategies to combine radiotherapy and surgery because of the high frequency of locoregional and distant recurrence.
Loss of natural speech, of course, is the chief forfeit of surgery for laryngeal cancer. Laryngectomy also can result in the need of a neck stoma (for breathing), an inability to smell or sneeze, a diminished sense of taste, and problems with swallowing. Cosmetic and psychological consequences can be great.
Although preservation of the voice has long been an important goal of the treatment of these patients, interest in the issue has heightened over recent years. A trend of decreasing use of radical surgery so as to preserve normal function and improve quality of life is developing in a variety of other solid tumor settings as well. Two examples are limb preservation in osteosarcoma and breast preservation in breast cancer.
Results of several randomized trials for laryngeal preservation in advanced operable cancers of the larynx or hypopharynx have been reported (Table 90-28). Based on promising pilot data in 1985, the Veterans Affairs Cooperative Studies Program (VACSP) initiated a multi-institutional randomized prospective trial that attempted laryngeal preservation in previously untreated patients with locally advanced resectable laryngeal cancer (
Figure 90-9). Patterns of treatment failure differed. At a median follow-up of 8 years, an increased locoregional recurrence rate and a decreased distant metastasis rate were seen in the sequential chemoradiotherapy group in comparison with the surgery/radiotherapy group. The findings also indicated that a delay of definitive local therapy in chemotherapy nonresponders was not detrimental. Chemotherapy responders and nonresponders in the experimental arm had about the same rate of survival. Eight-year rates were similar (approximately 30%) in the two study arms, and laryngeal preservation was achieved in over 60% of the chemotherapy recipients. Sequential chemoradiotherapy was least effective in advanced local (T4) and regional disease (N2, N3), and over 50% of patients in these subgroups required salvage laryngectomy.
These long-term results from the large VACSP trial indicate that induction chemotherapy followed by definitive radiotherapy can be an effective strategy for achieving laryngeal preservation in a high percentage of patients without compromising overall survival. The trial’s success argues strongly for adoption of this new treatment strategy in order to spare patients the functional, psychological, and cosmetic deformities resulting from total laryngectomy.
While the VACSP study shows that induction chemotherapy plus radiotherapy is as effective as the previous standard treatment of surgery and postoperative radiotherapy, the question as to what role chemotherapy actually plays has been raised; does it add efficacy or is it merely predictive? For certain laryngeal cancer subsets such as T3N0 glottic tumors, it appears that definitive radiotherapy alone with surgery held in reserve as salvage gives essentially the same results as either arm of the VACSP study and is much more cost-effective. A Head and Neck Intergroup randomized trial for stage T2-3N0-3 tumors of the supraglottic and glottic larynx addresses this question. The precise role of induction chemotherapy was evaluated in a phase III trial in the United States. This trial compared three cycles of induction chemotherapy using the VACSP cisplatin plus 5-FU regimen followed by radiotherapy to concomitant radiotherapy and cisplatin versus radiotherapy alone in patients with locally advanced stage III or IV SCC (T1 and T4 [with tumor penetrating through cartilage or invading > 1 cm into the base of the tongue]) of the glottic or supraglottic larynx. Preliminary data from this study have been presented. The 2-year larynx preservation rate is highest for concomitant cisplatin and radiotherapy. There was a significant advantage in the radiotherapy plus concurrent cisplatin arm (88% of patients preserving their larynx at 2 years) versus the arm with induction chemotherapy followed by radiotherapy (74%) (p = .005) and the radiotherapy-alone arm (69%). These results represent a reduction of over 50% in the rate of total laryngectomy with concurrent chemotherapy/radiotherapy, compared with the previous standard of care. No significant difference has been observed in laryngectomy-free or overall survival among the three study arms. However, the time to laryngectomy was significantly longer for the concomitant cisplatin and radiotherapy group compared with the induction chemotherapy arm (p = .0094). Although there was a statistically significant effect and a 50% reduction in the relative rate of laryngectomy, the absolute improvement in larynx preservation at 2 years was 15%, and there were no survival advantages or results of larynx function, (eg, no data regarding voice quality, quality of life, or swallowing functions). These results established chemotherapy/radiotherapy as an alternative standard of care for organ preservation in locally resectable laryngeal and hypopharyngeal cancer.
Two recent European phase III trials add further support to this organ-preservation approach. A French trial had a design similar to the VA trial, except that it used carboplatin instead of cisplatin and included all head and neck cancer subsites. The locoregional therapy was surgery and/or radiotherapy, determined by the treating physician. Patients were assigned randomly to chemotherapy followed by standard local therapy (in 152 patients) or standard local therapy alone (154 patients). The neoadjuvant carboplatin plus 5-FU produced overall and complete response rates of 57% and 31%, respectively. The chemotherapy/radiotherapy arm produced an increased 4-year survival rate (49% vs 38%, P 5 .053) and organ-preservation rate (48% vs 20%, P 5 .001), compared with the standard local therapy arm.
The EORTC reported a phase III trial that randomized 197 patients with locally advanced hypopharyngeal cancer; 100 patients were randomized to chemoradiotherapy with cisplatin plus 5-FU, and 97 patients were randomized to standard local therapy for total laryngectomy, partial pharyngectomy, and radical neck dissection and postoperative radiotherapy. In this study, patients in the neoadjuvant arm who achieved less than a complete response after two cycles of cisplatin plus 5-FU underwent immediate surgical therapy as outlined in the standard arm. At a follow-up of 3 years, survival rates were similar in the two arms and 28% of patients in the neoadjuvant arm were alive, were disease-free, and had preserved their larynxes.
Future trials aimed at laryngeal preservation will need to establish chemotherapy regimens that achieve higher complete response rates, investigate innovative fractionated-radiation schemes, and assess concomitant chemoradiation programs.
Promising new combinations are being tried to overcome the major disappointment of neoadjuvant chemotherapeutic approaches; that is, despite achieving high response rates of 70% to 90% (20% to 40% complete response rates) in advanced untreated patients, neoadjuvant chemotherapy has not translated into improved survival. Data suggest that neoadjuvant regimens capable of significantly improving survival must be able to produce reproducible complete response rates of over 50%. So far, complete response rates in phase II trials have appeared to be inflated and always drop in phase III trials. Although developed most often in recurrent disease, promising new dose-intensive regimens are designed for later integration into primary chemotherapy. The following discussion includes trials of new combinations in both the recurrent-disease and neoadjuvant settings.
Preclinical data suggested a promising new line of work to potentiate 5-fluorouracil cytotoxicity. Laboratory studies indicate that 5-fluorouracil resistance can be overcome by adding reduced folates (eg, leucovorin) before 5-fluorouracil exposure, due to enhanced stability of the active ternary complex with thymidylate synthase. Combined leucovorin with cisplatin and 5-fluorouracil (PFL regimen) has undergone extensive study. Although the data are mixed, overall the data suggest a modest increase in acivity countered by a major increase in severe toxicity.
Revision date: July 4, 2011
Last revised: by David A. Scott, M.D.