Phase 2 trial of concurrent 5-fluorouracil, hydroxyurea, cetuximab, and hyperfractionated intensity-modulated radiation therapy for locally advanced head and neck cancer




The objective of this phase 2 study was to evaluate the tolerability and efficacy of incorporating cetuximab and simultaneous integrated-boost (SIB), intensity-modulated radiation therapy (IMRT) into a well described 5-fluorouracil (5-FU) and hydroxyurea (HU)-based chemoradiation regimen.


Patients with stage IVA and IVB or high-risk stage III squamous cell carcinomas of the head and neck were enrolled on a phase 2 trial. Prior organ-conserving surgery or induction chemotherapy was allowed off protocol. SIB-IMRT was prescribed to low-risk volumes (43.2 gray [Gy] to 48 Gy) and intermediate-risk volumes (54-63 Gy). A separate IMRT cone-down plan was targeted to macroscopic disease (72 Gy). The median radiation dose was 72 Gy (range, 60-72 Gy) administered in 1.5 Gy fractions twice daily during Weeks 1, 3, 5, 7 and 9. Concurrent systemic therapy consisted of 5-FU (600 mg/m2), HU (500 mg twice daily), and cetuximab (250 mg/m2).


From January 2007 through April 2008, 33 patients were enrolled. At a median follow-up of 24 months, the 2-year rates of locoregional control, distant control, disease-free survival, and overall survival were 83%, 79%, 69%, and 86%, respectively. Grade 3 toxicity consisted of mucositis in 33% of patients, radiation dermatitis in 15%of patients, anemia in 18% of patients, leukopenia in 18% of patients, neutropenia in 12% of patients, and thrombocytopenia in 3% of patients. Most patients (64%) were able to tolerate treatment without a feeding tube, and there were no acute or late grade ≥4 adverse events.


The current results indicated that concurrent 5-FU, HU, and cetuximab plus SIB-IMRT is a promising and reasonably well tolerated approach to incorporating molecularly targeted therapy into curative therapy for patients with locally advanced head and neck cancer. Cancer 2011. © 2010 American Cancer Society.

For patients with locally advanced stage (stages III, IVA, and IVB) head and neck squamous cell carcinoma (HNSCC), the traditional approach of radical surgery and radiation therapy has resulted in disappointing cure rates of 30% to 50%.1 In addition, this approach is often is associated with significant cosmetic and functional impairment, resulting in decreased quality of life.2 Current ongoing, active research is exploring combined-modality therapy to improve survival, organ preservation, and function in patients with locally advanced HNSCC.3, 4 It has been demonstrated that chemoradiation is superior to radiation alone for patients with stage IVA and IV HNSCC and selected patients with stage III HNSCC in the organ-preservation, inoperable and adjuvant settings.5-8 Concurrent cisplatin and daily radiotherapy is the most widely investigated treatment regimen, but the results remain suboptimal for some patient subgroups.5, 7-9 For instance, in the most recent US Intergroup study of unresectable stage III and IV disease, patients who received 70 grays (Gy) of radiation in 2-Gy fractions with concurrent cisplatin had a 3-year overall survival rate of 37% and a 3-year disease-specific survival rate of 51%.5 Patients who had stage III and IV HNSCC who received concurrent cetuximab and radiotherapy had a 3-year overall survival rate of 55% and a locoregional control rate of 47%.10

Potential avenues of treatment intensification include incorporating organ-conserving surgery, hyperfractionated radiation, and the use of concurrent multiagent chemotherapy, biologically targeted therapy, or induction chemotherapy.11-14 Although it is a departure from conventional chemoradiation paradigms, 1 treatment regimen that incorporates these strategies is the 5-fluorouracil (5-FU)/hydroxyurea-based chemoradiation approach, which was developed by investigators at the University of Chicago.14, 15 Results recently were reported on the long-term follow-up of multiple phase 2 trials of combined induction carboplatin and paclitaxel for 6 weeks followed by concurrent 5-FU, hydroxyurea, and paclitaxel plus twice-daily radiotherapy for patients with stage IVA and IVB or poor-prognosis stage III head and neck cancer.14-16 Those results were impressive, with a 5-year locoregional control rate of 91%, a 5-year distant control rate of 87%, and a 5-year overall survival rate of 62%.16 Although these results are promising, the regimen has not been adopted widely because of concerns about toxicity and complexity. With the objective of reducing toxicity, a recent phase 2 trial of induction carboplatin and paclitaxel for 6 weeks followed by concurrent 5-FU, hydroxyurea, and gefitinib plus twice-daily radiotherapy resulted in a 3-year disease-free survival rate of 64% and an overall 3-year survival rate of 73% with decreased hematologic and neurologic toxicity.17 Although cetuximab had not been combined previously with 5-FU/hydroxyurea-based chemoradiation, extensive preclinical and clinical data suggest that cetuximab is a well tolerated and effective radiosensitizer for HNSCC.10, 18 Our group hypothesized that the addition of cetuximab would enhance the therapeutic ratio of a highly active chemoradiation regimen.


Study Design and Eligibility

The study opened in January 2007 and reached accrual goals in April 2008. This phase 2 trial (MSSM 06-1155; National Clinical Trials no. 00462735) was approved by the institutional review board. Eligible patients had either HNSCC or poorly differentiated carcinoma. Patients had either stage IVA and IVB disease or high-risk stage III disease according to the sixth edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual, which was defined as a base of tongue or hypopharyngeal primary tumor, or major pathologic risk factors (microscopic positive margins or extracapsular extension). All cases were reviewed at a multidisciplinary conference, which was attended by representatives from the Departments of Head and Neck Surgery, Radiation Oncology, Medical Oncology, Palliative Care, Social Work, and Nutrition. Patients who had received previous head and neck radiation were excluded. Patients had an Eastern Cooperative Oncology performance status ≤2 and adequate bone marrow, kidney, and liver function based on laboratory evaluation. Organ-preserving surgery, which was defined as selective neck dissection or preservation of the tongue, larynx, and orbit, or induction chemotherapy was allowed before registration on the protocol.

This phase 2 study was designed as a Fleming 1-stage design, with a hypothesized 2-year disease-free survival of 46% for cetuximab and radiation alone and 70% for concurrent cetuximab, 5-FU, and hydroxyurea plus hyperfractionated radiotherapy. The targeted accrual was 33 patients (to yield 31 assessable patients), which had an 80% power to reject the null hypothesis with α = .05 using a 2-sided test. Early stopping rules were used in the event of unexpected grade 3, 4, or 5 toxicities. Of 73 patients who required concurrent radiation and radiosensitizing drug therapy for stage III, IVA, or IVB head and neck cancer, 40 patients were excluded from the trial (Fig. 1). Reasons for exclusion were patient refusal (n = 18), previous head and neck radiation (n = 12), stage III disease that did not meet eligibility criteria (n = 4), and ineligibility because of a comorbid illness (n = 6).

Figure 1.

This is a Consolidated Standards of Reporting Trials (or CONSORT) diagram for phase 2 trials. RT indicates radiotherapy.

Initial staging procedures included a history and physical examination, nasolaryngoscopy with biopsy with tumor measurements, dental evaluation, and head-neck and chest imaging studies with or without a positron emission tomography (PET) study. When available, PET data were used for radiation planning using 40% of the maximal standard uptake value for edge delineation.19 Prophylactic feeding tubes were strongly recommended for patients with oral cavity involvement, massive tumors, advanced age, or limited physiologic reserves. All patients signed a written informed consent form. Patients were followed prospectively by the multidisciplinary team usually with PET/computed tomography (CT)-based follow-up studies.

Week On/Week Off Chemoradiotherapy

All patients received concurrent chemoradiotherapy, which consisted of continuous-infusion 5-FU at a dose of 600 mg/m2 daily for 120 hours, hydroxyurea 500 mg orally every 12 hours with the morning dose administered 2 hours before radiation, and cetuximab 250 mg/m2 on Day 1 after the first dose of radiation was administered to all patients (Fig. 2). No anticancer therapy was administered on Days 6 through 14. Patients who had macroscopic disease received 5 cycles of chemoradiation, whereas postsurgical patients with microscopic disease received 4 cycles of chemoradiation. Chemoradiotherapy was withheld only for an absolute neutrophil count <500, fever, infection, or patient refusal, but not for mucositis or dermatitis.

Figure 2.

This is the treatment schema for concurrent 5-fluorouracil, hydroxyurea, cetuximab, and hyperfractionated intensity-modulated radiotherapy (IMRT). Gy indicates grays.

Radiotherapy was administered at 1.5 Gy per fraction twice daily with treatments separated by at least 6 hours on Days 1 through 5 on an alternating week schedule. Radiation was delivered with intensity-modulated radiation therapy (IMRT) planning for all patients. Patients underwent CT simulation, usually with fusion of PET or MRI images to assist with target delineation. Preinduction chemotherapy tumor volumes were targeted.

Macroscopic tumor volume received from 72 to 73.5 Gy, microscopic positive margins received 66 Gy, high-risk microscopic disease (resected tumor bed or first echelon of uninvolved lymph node stations) received 54 to 63 Gy, and low-risk microscopic disease (low-risk lymph node stations) received 43.2 to 48 Gy. The microscopic volumes were covered with a simultaneous integrated boost plan. A separate cone down was performed for patients with macroscopic disease. Patients received the lower dose levels if they achieved a partial or complete response to induction chemotherapy. The median dose was 72 Gy (range, 60-72 Gy). The median duration of treatment was 59 days (range, 46-87 days).

Induction Chemotherapy

Induction chemotherapy, which was administered off protocol, was recommended for patients who had N2 or N3 lymph node status with intact primary tumors or status post-tonsillectomy with the objective of decreasing distant metastasis. Measurable disease was not considered a requirement for induction chemotherapy. Twenty patients (61%) received combined docetaxel (75 mg/m2), cisplatin (75 mg/m2), and 5-FU (750 mg/m2 × 5 days) every 3 weeks for 2 cycles.


Seventeen patients (52%) underwent organ-conserving surgery before chemoradiation. This included 8 patients who underwent with resection of the primary site and selective neck dissection, 3 patients who underwent resection of the primary site alone, and 6 patients who underwent neck dissection alone. Adverse pathology included 9 patients with microscopic positive margins and 8 patients with perineural invasion among 11 patients who underwent primary site resection and 9 patients with multiple, pathologically positive lymph nodes and 12 patients with extracapsular extension among 14 patients who underwent pretreatment neck dissection. Four patients underwent selective neck dissection after achieving a complete response of N2 or N3 lymph node disease to chemoradiation.

Quality of Life

Acute and late toxicities were scored using version 3.0 of the National Cancer Institute Common Terminology Criteria for Adverse Events. The validated University of Washington Quality of Life (UW-QOLR) questionnaire was administered to patients before treatment, immediately after the completion of chemoradiation, and at follow-up visits.

Treatment Evaluation and Statistical Considerations

Recurrence-free survival was measured from the initiation of nonsurgical treatment to either the last follow-up, disease progression, or death using intent-to-treat methodology. Failures were classified as local, regional, or distant. Survival curves were calculated using the Kaplan-Meier method with Stata statistical software (version 9.1; Stata Corp., College Station, Tex). Differences in survival curves were calculated with log-rank tests, and differences in mean quality-of-life scores were calculated with 2-sided t tests.


In total, 33 patients were treated, and surviving patients were followed for a median of 24 months (range, 17-32 months). The patient characteristics are summarized in Table 1. The median age was 59 years (range, 18-77 years). Thirty-two patients (97%) had stage IV disease, and 1 patient (3%) had stage III disease. Two patients had recurrent disease before they started chemoradiation. Nineteen patients (58%) had T3 or T4 primary tumors, and 24 patients (73%) had N2 or N3 lymph node disease (Table 2). A percutaneous gastrostomy tube was placed in 15% of patients before radiation was started.3

Table 1. Patient Characteristics
CharacteristicNo of Patients (%)
Median age [range], y59 [18-77]
 Men24 (73)
 Women9 (27)
Performance score 
 012 (36)
 119 (58)
 22 (6)
Charlson comorbidity score 
 0-127 (82)
 ≥26 (18)
 White23 (70)
 Black5 (15)
 Hispanic5 (15)
 Asian0 (0)
Alcohol consumption 
 None10 (30)
 Occasional14 (42)
 Moderate2 (6)
 Heavy7 (16)
 None6 (18)
 Cigar, pipe, or betel nut only3 (9)
 ≤20 pack-y9 (27)
 20.1-40 pack-y7 (21)
 >40 pack-y8 (24)
Primary site 
 Sinonasal2 (6)
 Nasopharynx2 (6)
 Oropharynx14 (42)
 Oral cavity6 (18)
 Salivary gland1 (3)
 Larynx2 (6)
 Hyopharynx4 (12)
 Unknown primary2 (6)
Table 2. Distribution of Tumor and Lymph Node Classifications
 Tumor Classification 
Lymph Node ClassificationTxT1T2T3T4Total
N2a 11 2
Table 3. Acute Toxicity for Week On/Week Off Chemoradiotherapy
 Toxicity Grade, %
Toxicity (n=33)1234
  1. NA indicates not available; ANC, absolute neutrophil count; WBC, white blood cell count; Hgb, hemoglobin; Plt, platelets.


Concomitant Chemoradiotherapy—Feasibility and Tolerability

Patients received 99.7% of the prescribed radiation dose and 98% of the prescribed chemotherapy dose intensity. One patient discontinued radiation before completion because of grade 3 mucositis, and another patient withdrew consent after 2 cycles of concurrent chemoradiation but completed concurrent cisplatin and accelerated radiation off protocol. No grade 4 toxicity was observed. A feeding tube or total parental nutrition was required by 36% of patients during or after treatment. The median weight loss was 12% (range, 0%-17%). Acute, moderate xerostomia was observed in 48% of patients; grade 3 mucositis was observed in 33% of patients; and grade 3 dermatitis was observed in 15% of patients. Only 18% of patients developed grade 3 leukopenia, 12% developed grade 3 neutropenia, 18% developed grade 3 anemia, and 3% developed grade 3 thrombocytopenia. No patients developed neutropenic fever or infection during concurrent chemoradiation. Grade 2 or 3 acneiform rash was observed in 15% of patients. Thrush, mouth sores related to 5-FU, fatigue, thickened phlegm, and altered taste were common acute adverse events that were managed with supportive measures. Greater than 99% of the prescribed radiation dose was delivered. Two patients had unplanned treatment delays of ≥7 days because of noncompliance.

Survival and Patterns of Failure

The 2-year locoregional control, distant control, disease-free survival, and overall survival rates were 83%, 79%, 69%, and 86%, respectively (Fig. 3). The 2-year local control rate was 94%, the 2-year regional control rate was 86%, and the 2-year rate of locoregional control with surgical salvage was 87%. To date, 10 patients (30%) have experienced recurrence of their head and neck cancer. Patterns of failure were local only in 1 patient, regional only in 2 patients, distant only in 5 patients, and both locoregional and distant in 2 patients. In addition, there was 1 second primary lung adenocarcinoma and 1 second primary squamous cell carcinoma of the thoracic esophagus. There was no significant difference in 2-year disease-free survival between patients with oropharyngeal primaries versus nonoropharyngeal primaries (71% vs 67%, respectively; P = .7).

Figure 3.

These charts illustrate (Top Left) locoregional control (LRC), (Top Right) distant control (DC), (Bottom Left) disease-free survival (DFS), and (Bottom Right) overall survival (OS) in patients with locally advanced head and neck cancer.

Long-Term Toxicity

To date, grade ≥ 3 late toxicities included 1 case each of grade 3 frontal bone necrosis and grade 3 esophageal stricture. Additionally, one patient with tumor that extended to the middle ear developed cartilage necrosis and unilateral hearing loss that required reconstruction. The rates of long-term (>6 months) percutaneous endoscopic gastrostomy (PEG) tube dependence resulting from treatment was 3%, and another 2 patients reported solid food dysphagia. One additional patient had a PEG tube in place for 1 year that was removed after successful esophageal dilatation. It is noteworthy that 1 patient developed grade 2 skin hypopigmentation and grade 2 skin telangiectasia at the site of grade 3 acute dermatitis (see Fig. 4, top left). Grade 2 xerostomia was noted in 33% of patients. No patients developed any visual problems or required tracheotomy because of toxicity.

Figure 4.

(Top) A late grade 2 telangiectasia is observed at the site of grade 3 acute skin toxicity 26 months after chemoradiation. (Bottom) This chart illustrates the mean patient-reported University of Washington Quality-of-Life (UW-QOLR) scores.

Quality of Life

Twenty patients completed the UW-QOLR questionnaire before and after therapy (Fig. 4, top right). Compared with baseline scores, quality-of-life scores were decreased after treatment (P < .001) and at 1 month after chemoradiation (P = .09). However, the scores returned to baseline as early as 4 months after chemoradiation.


There is significant interest in adding epidermal growth factor receptor (EGFR) inhibitors to existing chemoradiation regimens.17, 20-22 In a phase 2 trial, cetuximab was added to an intensive regimen of concurrent cisplatin and concomitant boost radiotherapy. Although promising 3-year locoregional control and overall survival rates of 78% and 71%, respectively, were reported, that trial was closed prematurely because of excessive adverse events.22 Compared with these data and with published results from studies of concurrent cetuximab and radiation, we reported promising rates of locoregional control and survival without excessive toxicity.10, 22 In the current study, we demonstrated that concurrent 5-FU, hydroxyurea, cetuximab, and hyperfractionated IMRT is a promising and reasonably well tolerated approach to incorporating molecularly targeted therapy into the curative treatment of locally advanced HNSCC. This regimen can be preceded by either surgery or induction chemotherapy without affecting tolerability.

Compared with published data from the University of Chicago 9502 regimen of induction carboplatin and paclitaxel for 6 weeks followed by concurrent 5-FU, hydroxyurea, paclitaxel, and hyperfractionated radiation, we observed a numerically higher rate of locoregional and distant failure at 2 years despite relatively short follow-up.16 One possibility is that substitution of cetuximab for paclitaxel reduced treatment intensity, particularly because cetuximab was not continued during the week off. It is noteworthy that our 2-year survival and disease control data are similar to those reported with 5-FU, hydroxyurea, gefitinib, and hyperfractionated radiotherapy.16 In designing future studies, we must consider whether the relatively modest, elective lymph node doses that were considered optimal in the setting of concurrent paclitaxel are adequate when substituting an EGFR inhibitor.16 A second possibility may be the inclusion of a different patient population with postsurgical patients who have disease with adverse pathology.7 A third possibility is our routine use of PET/CT in surveillance, which may detect locoregional and distant failures earlier.23 The excellent 2-year overall survival, combined with a relatively high distant metastasis rate despite extensive use of induction chemotherapy, raises the possibility of earlier detection. Finally, based on promising phase 2 data suggesting that 6 weeks of taxane-based chemotherapy is effective for preventing distant metastases, only 2 cycles of induction docetaxel, 5-FU, and cisplatin were administered.14, 15 However, more recent phase 3 data suggest that 3 or 4 cycles of docetaxel, 5-FU, and cisplatin is the preferred induction regimen and may further reduce the incidence of distant metastases.13

Relative to published results of intensive 5-FU/hydroxyurea-based chemoradiation using cisplatin or paclitaxel, rates of mucosal and hematologic toxicity were low.14, 15 Whether this observation is because of decreasing chemotherapy intensity, exclusive use of IMRT with increased dose conformality associated with a simultaneous integrated boost approach, patient selection, or reduced doses of radiation in surgically resected patients is unclear. There may be an increase in acute and late cutaneous toxicity associated with this treatment regimen, although long-term quality of life after treatment generally is good. Because the half-life of cetuximab is approximately 7 days and the toxicity profile was favorable, currently, we are evaluating the safety and efficacy of intensifying this regimen by continuing cetuximab on Day 8 of each 14-day cycle (Fig. 2).24

In the literature, there is growing concern among some oncologists that acute toxicity caused by treatment intensification is reaching the limits of tolerability.25 An ongoing, randomized phase 2 trial at the University of Chicago is comparing concurrent 5-FU, hydroxyurea, cetuximab, and hyperfractionated radiation versus concurrent cisplatin, cetuximab, and concomitant boost radiotherapy after induction docetaxel, 5-FU, and cisplatin may identify a regimen that achieves an optimal balance between toxicity and efficacy.4 Ultimately, we may individualize treatment intensity based on predictors of response to conventional and intensified therapies.26 Predictors of response to EGFR inhibitors in HNSCC remain elusive, and it is unknown whether human papillomavirus (HPV) status will be useful in selecting appropriate therapy.27 Currently, our group is reviewing the HPV and EGFR status of treated patients to determine whether specific subgroups benefited from protocol therapy. On the basis of the published literature, patients with HPV-positive disease have a more favorable prognosis with standard therapy and are unlikely to benefit from further treatment intensification.28 Conversely, HPV-negative patients continue to experience high rates of treatment failure despite intensive combined-modality therapy and may benefit from novel therapeutic approaches. In conclusion, the current results indicate that concurrent 5-FU, hydroxyurea, cetuximab, and hyperfractionated IMRT is a promising and reasonably well tolerated approach for incorporating molecularly targeted therapy into curative therapy for patients with locally advanced head and neck cancer.


Supported in part by the Mount Sinai School of Medicine and the Ellen Katz Foundation. J.K. received research funding from OSI Pharmaceuticals. S.H.P. and J.K. are on the Speakers Bureau of Sanofi-Aventis.