Docetaxel induction therapy in locally advanced squamous cell carcinoma of the head and neck

Patients with locally advanced squamous cell carcinoma of the head and neck (SCCHN) are often treated with induction chemotherapy or chemoradiotherapy, but to date without major impact on survival. The combination of cisplatin–5-fluorouracil (5-FU) (PF) has been used as standard induction therapy; however, poor patient survival has stimulated investigation into new agents with potential activity in SCCHN. Docetaxel has significant single-agent activity in SCCHN and has been investigated in combination with PF regimens as induction therapy. The results of six phase II studies of docetaxel–PF regimens (TPF) as induction in locally advanced SCCHN patients are reviewed and reported. Consistently, high 2-year survival rates and overall response rates were demonstrated across the phase II trials in the range 42–82 and 71–100%, respectively. The toxicity profile seen with TPF-based regimens was acceptable. The primary toxicity was neutropenia, which together with gastrointestinal complaints accounted for the majority of adverse events. Given the encouraging phase II experience with TPF-based regimens, two large-scale phase III studies comparing TPF-based regimens with standard PF regimens are underway. The results have significant potential for validating the findings of the phase II studies, demonstrating improved survival and overall response of patients treated with docetaxel-based induction chemotherapy.

Squamous cell cancer of the head and neck (SCCHN) is a significant international problem. Cancer specific to the mouth and pharynx is the sixth most common cancer worldwide, and the third most common cancer among men in developing countries (Johnson, 2001). Men are at higher risk of developing these cancers and are affected 2 -3 times as often as women in industrialised countries. There are approximately 400 000 new cases of SCCHN projected annually (Parkin and Muir, 1992). Most SCCHN cases are locally advanced at presentation, with up to 75% of patients having stage III -IV, M0 disease (Dreyfuss et al, 1996;Parker, 1996). Stage of disease at diagnosis is regarded as the single most important prognostic factor (Forastiere, 2000). Stage I -II SCCHN is often curable with either surgery or radiotherapy, but it is generally accepted that treatment of locally advanced SCCHN should involve a combined modality approach (Forastiere, 2000).
Surgery and/or radiotherapy remain cornerstones of therapy in patients with locally advanced SCCHN. However, induction chemotherapy and chemoradiotherapy are now firmly integrated in locoregional care. Controlled studies have established cisplatin and continuous-infusion 5-fluorouracil (5-FU) as the standard induction regimen for unresectable locally advanced SCCHN patients (Paccagnella et al, 1994;Lefebvre et al, 1996;Domenge et al, 2000;Pignon et al, 2000). Up to 40% of chemonaïve locally advanced SCCHN patients treated with cisplatin -5-FU (PF) have been reported to achieve a complete response, with an overall response rate (ORR) in the range of 85% (Schantz et al, 2001). The main toxicities associated with PF-induction therapy are haematological, digestive and mucositis with the majority of events being grade I or II (Paccagnella et al, 1994;Domenge et al, 2000). Despite inducing locoregional control, high response rates and a significant improvement in survival, PF induction therapy is associated with a relatively poor absolute rate of patient survival, which is similar to that seen with intensive chemoradiotherapy regimens (Paccagnella et al, 1994;Lefebvre et al, 1996;Domenge et al, 2000;Pignon et al, 2000). The dose-intensive cisplatin -5-FU -leucovorin (PFL) combination was subsequently developed in an attempt to improve complete response rates to induction chemotherapy and thus survival. The addition of leucovorin resulted in higher complete response rates at the expense of increased toxicity (Vokes et al, 1990;Schneider et al, 1995;Clark et al, 1997).
The need for further improvements in SCCHN care has stimulated intensive investigation into new agents with potential in induction chemotherapy and chemoradiotherapy regimens. New agents such as epidermal growth factor receptor (EGFR) inhibitor ZD1839 (Iressat), antibody C225 (Cetuximabt) and inhibitor , in addition to farnesyl transferase inhibitors, have shown some promise in early studies. Although they are relatively ineffective as single agents, their combination with other agents (radiation and/or cytotoxic drugs) is under evaluation and may be of interest. UFT (a mixed compound of 5-FU and uracil) has shown promising single-agent activity, but combination therapy has yet to be studied. Alpha interferon has been tested as a potentiator of PFL or PF chemotherapy, but a phase III trial has failed to demonstrate any additional advantages from this approach (Vokes et al, 1995;Schrijvers et al, 1998). Perhaps the most promising agents have been the taxanes. In particular, the taxane docetaxel has demonstrated significant single-agent activity in recurrent SCCHN (Catimel et al, 1994;Dreyfuss et al, 1996;Couteau et al, 1999). Docetaxel is similar to paclitaxel with respect to its general mechanism of action (tubulin stabilization and cell cycle arrest), and the two share a number of pharmacological characteristics (Colevas and Posner, 2001). However, docetaxel and paclitaxel have somewhat different pharmacodynamics and toxicities that may be important in combination therapy programmes. Docetaxel, for example, has a greater affinity for tubulin, a longer intracellular half-life, and promotes microtubule stabilization at lower drug concentrations (Lavelle et al, 1995). The major toxicity associated with docetaxel is highly predictable myelotoxicity (Posner, 2001a). More importantly, neuropathy, a dominant side effect of cisplatin, is minimal with docetaxel-containing regimens, but can be significant in paclitaxel-containing regimens (Schrijvers and Vermorken, 2000;Colevas and Posner, 2001;Posner, 2001a). The differing mechanisms of action, and relative nonoverlapping toxicities compared with PF, have prompted interest in the addition of docetaxel to PF regimens.

POTENTIAL OF DOCETAXEL IN SCCHN
In early preclinical studies, docetaxel demonstrated pronounced in vitro and in vivo cytotoxicity against a variety of human cancer cell lines, particularly head and neck cancer (Braakhuis et al, 1994). In fact, docetaxel was more effective than cisplatin in inhibiting the growth of two SCCHN xenograft models (HNX-14C and HNX-22B) (Braakhuis et al, 1994). Further preclinical evidence for docetaxel efficacy has been derived from studies employing a murine model, representative of human head and neck cancer (Sommer et al, 2001). From the drugs investigated (cisplatin, carboplatin, docetaxel, methotrexate, 5-FU), as single agents or in combination, an ifosfamide -docetaxel combination produced the best tumourfree survival.
The clinical efficacy and safety of docetaxel in SCCHN was initially established in patients with metastatic or recurrent/ incurable SCCHN. In three phase II studies (Table 1), single-agent docetaxel (100 mg m À2 ) administered on day 1 every 3 weeks produced response rates ranging from 21 to 42% (Catimel et al, 1994;Dreyfuss et al, 1996;Couteau et al, 1999). As expected, grade III -IV neutropenia was the principal toxicity in these studies. These results compare favourably with other single agents in the management of SCCHN (Schoffski et al, 1998;Posner, 2001a).
Phase I/II studies combining docetaxel with cisplatin and/or 5-FU were undertaken in recurrent or metastatic SCCHN patients to look for improved efficacy and safety. In two trials, docetaxel -5-FU resulted in ORRs of 24 and 27% Tubiana-Mathieu et al, 2000). Docetaxel -cisplatin appeared to be a more effective combination, with overall response rates ranging from 33 to 76% (Kienzer et al, 1998;Schoffski et al, 1999;Specht et al, 2000;Glisson et al, 2002). In the trial conducted by the European Organisation for the Research and Treatment of Cancer (EORTC), docetaxel (100 mg m À2 ) plus cisplatin (75 mg m À2 ) produced a response rate of 54% (Schoffski et al, 1999). Of the 44 patients enrolled in this trial, 22 were chemotherapy-naïve and the response rate in this group was 86%. As with single-agent docetaxel, the main toxicity in each of the combination studies was myelosuppression (Kienzer et al, 1998;Schoffski et al, 1999;Colevas et al, 2000;Specht et al, 2000;Tubiana-Mathieu et al, 2000;Glisson et al, 2002). Mucositis was more commonly seen in the combination studies (particularly docetaxel -5-FU) than with single-agent docetaxel.

DOCETAXEL INDUCTION THERAPY IN SCCHN: PHASE II RESULTS
The promising response rates and tolerability profiles exhibited in recurrent or metastatic SCCHN patients treated with docetaxelcisplatin led to the investigation of docetaxel regimens as induction therapy for locally advanced disease. Two phase II induction studies with docetaxel 75 mg m À2 plus cisplatin 75 or 100 mg m À2 i.v. every 3 weeks followed by definitive radiotherapy in patients with locally advanced SCCHN have been reported (Mel et al, 2000;Caponigro et al, 2001). In an intention-to-treat analysis, objective responses of 62% and 46% were seen in patients during the induction phase, with corresponding complete response rates of 17% (Mel et al, 2000) and 11% (Caponigro et al, 2001), respectively. Grade III -IV neutropenia was experienced by 74% of patients in one study, although febrile neutropenia affected only 17% of patients (Mel et al, 2000). In the second study, 61% of patients experienced grade III -IV neutropenia, which was complicated by fatal sepsis in 2 (4%) patients (Caponigro et al, 2001). The authors suggest that these toxic deaths might have been related to the higher dose of cisplatin (100 mg m À2 ) used in their study (Caponigro et al, 2001). The incidences of grade III -IV nonhaematological toxicity, including gastrointestinal complications, were relatively low in both studies. These results, in a heterogeneous group of patients with moderate-to-poor performance status, were similar to the data seen with PF. With data from recurrent patients, these studies demonstrate that the docetaxel -cisplatin combination has interactive efficacy similar to that seen with PF.
Rather than substituting cisplatin or 5-FU with a newer SCCHNactive agent such as docetaxel, in the curative setting, most research is now focused on the addition of docetaxel to PF-based regimens. Studies involving docetaxel -PF-based regimens demonstrate a high likelihood of superior clinical benefit vs current treatment options in terms of response rate and survival.
In an attempt to reduce hospitalisation, the TPFL-4 regimen was developed in which G-CSF was started earlier and treatment was compressed into the first 4 days of the 4-week cycle (Colevas et al, 1999). In all, 30 stage III -IV SCCHN patients received two loading doses of oral leucovorin 100 mg followed by docetaxel (60 mg m À2 ) d1 and then infusions of cisplatin (31.25 mg m À2 day À1 ), 5-FU (700 mg m À2 day À1 ) and leucovorin (500 mg m À2 day À1 ) over 4 days. G-CSF was given d4 -10 and ciprofloxacin on d5 -15. The ORR was 93%, with 63% CRs and 30% PRs. Primary tumour site clinical and pathological responses were 93 and 68%, respectively. The originally reported 2-year overall survival and disease-free survival rates were 87 and 57%, respectively (Colevas et al, 1999). A more recent analysis (median follow-up 30 months) reported the 2-year overall survival and disease-free survival rates as 83 and 53%, respectively . Haematological toxicity was generally mild, with grade III -IV neutropenia, thrombocytopenia, and anaemia reported during 8, 6 and 2% of cycles, respectively (Colevas et al, 1999). Mucositis was the most prominent nonhaematological grade III -IV event (48% of cycles), followed by nausea/vomiting (15% of cycles). These values represent major reductions in haematological and infectious toxicities compared with patients in TPFL-5. As a result, hospitalization was reduced to approximately 14%, the main cause of which was dehydration.

Docetaxel plus PF regimens
The TPFL regimens are intensive and associated with considerable toxicity. Patients routinely require G-CSF support and a significant proportion are hospitalised either for treatment or for treatmentrelated toxicity. In many patients age and comorbidities rule out such dose-intensive chemotherapy regimens. Intermediate doseintensive regimens involving docetaxel in combination with PF may be equally efficacious, providing alternatives in these patient groups. Three trials in which this intermediate-dose approach was investigated are discussed here and summarised in Table 3.
In a similar study (TAX 017HN), 48 patients with locally advanced SCCHN received docetaxel 75 mg m À2 and cisplatin 75 mg m À2 (level I) or 100 mg m À2 (level II) d1, followed by continuous infusions of 5-FU 750 mg m À2 day À1 over 5 days . Owing to infectious complications seen in the first 18 patients, ciprofloxacin was added from d5 to 15. Prior to this, six patients developed infections leading to hospitalisation. Grade III -IV neutropenia was reported in 88 and 63% of patients in the level I and II groups, respectively. An ORR of 71% was observed. At a recent follow-up (median 24 months), the 2-year overall survival rate was 42% (Posner et al, 2001c). Janinis et al (2001) administered docetaxel 80 mg m À2 d1, cisplatin 40 mg m À2 d2 and d3 and continuous 5-FU 1000 mg m À2 day À1 d1 -3 every 4 weeks in chemonaïve patients with locally advanced SCCHN. All patients received G-CSF d4 -9. A maximum of four chemotherapy cycles were allowed and radiation therapy was planned after completion of chemotherapy. The ORR postchemotherapy was 90% with a CR of 20%. After radiotherapy, the ORR was 95% with a CR that increased to 73%. After a median follow-up of 36 months, median disease-free and overall survival had not been reached. The 2-year survival rate was 60%. Grade III -IV toxicity was limited to leucopenia (25%), febrile neutropenia (10%), grade IV infection (10%) and grade IV diarrhoea (5%). The most common acute nonhaematological toxicities included alopecia, mucositis and peripheral sensory neuropathy. Alopecia was reversible, and mucositis was mild and did not require hospitalisation. Peripheral neuropathy was seen in 25% of patients, but was of mild degree with late occurrence.

ONGOING PHASE III INDUCTION STUDIES INCORPORATING DOCETAXEL
It is apparent that docetaxel, when used in conjunction with standard induction therapy (cisplatin, 5-FU with or without leucovorin), has significant activity in locally advanced SCCHN. Consistently high 2-year survival rates and ORRs have been  (9); vomiting (7); dehydration (7); thrombocytopenia (5); neurological-hearing (5); liver enzyme abnormalities (5) Janinis et al ( observed across phase II studies in the range 42 -82 and 71 -100%, respectively (Posner et al, 2001c). Each of the phase II studies discussed has used a slightly different regimen with varying concentrations of chemotherapy in 3-or 4-week cycles. The study results have suggested somewhat different response rates, overall survival rates and toxicity profiles. This may be attributed to factors such as differences in patient selection or subtle differences in schedule-related multidrug interactions. Given the encouraging phase II experience with docetaxel induction regimens, and the need to further investigate the various regimens in a randomised, controlled setting, phase III studies of two TPF-based regimens vs PF are warranted and are under way . A phase III trial (TAX 324) is being conducted mainly in the USA, with some centres in Europe and South America. This trial is evaluating three cycles of TPF (docetaxel 75 mg m À2 d1, cisplatin 100 mg m À2 d1 and 5-FU 1000 mg m À2 d1 -4 every 3 weeks) vs three cycles of PF (cisplatin 100 mg m À2 d1 and 5-FU 1000 mg m À2 d1 -5 every 3 weeks). This trial comprises a sequential therapy design incorporating postinduction chemoradiotherapy. Patients in both arms are treated with definitive chemoradiation, with weekly carboplatin for a maximum of seven doses following induction chemotherapy. A second study conducted by the EORTC (TAX 323) is evaluating four cycles of TPF (docetaxel 75 mg m À2 d1, cisplatin 75 mg m À2 d1 and 5-FU 750 mg m À2 d1 -5 every 3 weeks) vs four cycles of PF (cisplatin 100 mg m À2 d1 and 5-FU 1000 mg m À2 d1 -5 every 3 weeks). A 7week course of radiotherapy is to commence within 4 -7 weeks of completing the last chemotherapy cycle. These ongoing, large-scale, randomised trials are expected to confirm the positive benefits of adding docetaxel to PF-based regimens, as already demonstrated in the phase II setting.

DOCETAXEL-BASED CHEMORADIATION REGIMENS FOR LOCALLY ADVANCED DISEASE
In recent controlled studies, PF-based chemoradiation regimens, used as induction or definitive therapy, have demonstrated improved efficacy compared with conventional or hyperfractionated radiotherapy alone in locally advanced SCCHN (Vokes et al, 1995;Brizel et al, 1998). The activity of docetaxel in SCCHN, coupled with the drug's in vitro radiosensitizing properties (Creane et al, 1999;Mason et al, 1999;Pradier et al, 2001), provide the basis for clinical trials with docetaxel as a component of concurrent or alternating chemoradiation in locally advanced SCCHN (Koukourakis et al, 1999;Hesse et al, 2000;Tishler et al, 2002).
Docetaxel-based chemoradiation regimens have been shown to be feasible and active as induction and adjuvant therapy. The phase I/II study by Tishler et al (2002) considered daily radiation concurrent with weekly docetaxel in 21 patients with stage III -IV SCCHN, who responded poorly to induction chemotherapy, and demonstrated an ORR of 86% (CR 57%). Radiation was delivered at 2 Gy day À1 , to a total dose of 66 -74 Gy. The MTD of weekly docetaxel in this regimen was 25 mg m À2 . Mucositis was the major acute toxicity, occurring at grade III in all patients receiving docetaxel 25 mg m À2 . Swallowing problems represented the main long-term toxicity, although no patient remained feeding-tube dependent. Three-year survival was 460% in this poor prognosis group of patients.
Chemoradiation therapies with docetaxel -cisplatin have achieved an ORR as high as 100% (Varveris et al, 1999;Budach et al, 2000). The chemoradiation protocol considered by Budach et al (2000) was devised to reduce oral toxicity. A total of 15 patients with inoperable, recurrent head and neck cancer received three cycles of docetaxel (50 -60 mg m À2 d1) and cisplatin (15 mg m À2 d2 -5) during weeks 1, 4 and 7 alternating with two courses of radiotherapy (5 Â 2 Gy; total dose 40 Gy) in weeks 2 -3 and 5 -6. The recommended dose of docetaxel was 50 mg m À2 , as 60 mg m À2 led to unacceptable systemic toxicity in the first 12 patients treated. An ORR of 92% was achieved, with acceptable levels of oral toxicity. Varveris et al (1999) demonstrated that the radiosensitizing effect of docetaxel -cisplatin on hyperfractionated radiotherapy enables an ORR of 100% (CR 59%). Docetaxel and cisplatin were both given at 30 mg m À2 twice weekly to 54 patients. This dose was reduced to 15 -20 mg m À2 because of severe acute toxicity in the first 10 patients. Major toxicities included grade III -IV mucositis (86%), grade I -III hypersensitivity reactions (11%) and grade III myelotoxicity (11%).
Further studies are needed to elucidate the relative efficacy and toxicity of these concurrent radiation regimens vs the traditional sequential approach of induction chemotherapy followed by definitive radiotherapy.

CONCLUSIONS
Unmet needs in SCCHN care have stimulated intensive investigation into new agents with potential in induction chemotherapy and chemoradiotherapy regimens. Docetaxel has emerged as one of the most active agents in SCCHN, and is particularly efficacious as a component of induction therapy in locally advanced disease. The results of phase II studies investigating TPF induction regimens demonstrate consistently high 2-year survival rates and ORRs across all trials, in the range of 42 -82 and 71 -100%, respectively (Posner et al, 2001c). An acceptable toxicity profile is seen with TPF-based regimens, with neutropenia and gastrointestinal complaints accounting for the majority of adverse events. The encouraging phase II results are being validated in the ongoing, large-scale, randomised phase III studies comparing TPF with PF regimens.