Multimodal intensification therapy for previously untreated advanced resectable squamous cell carcinoma of the oral cavity, oropharynx, or hypopharynx

Authors

  • David E. Schuller M.D.,

    Corresponding author
    1. Department of Otolaryngology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    • Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Room 519, 300 West 10th Avenue, Columbus OH 43210
    Search for more papers by this author
  • John C. Grecula M.D.,

    1. Division of Radiation Oncology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Amit Agrawal M.D.,

    1. Department of Otolaryngology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Chris A. Rhoades M.D.,

    1. Division of Medical Oncology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Debra A. Orr M.S.S.,

    1. Department of Otolaryngology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Donn C. Young Ph.D.,

    1. Biostatistics Unit, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • James P. Malone M.D.,

    1. Department of Otolaryngology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Meredith Merz B.S.

    1. Department of Otolaryngology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University, Columbus, Ohio
    Search for more papers by this author

  • Fax: (614)293-3132

Abstract

BACKGROUND

An intensified treatment regimen for previously untreated Stage III and IV resectable oral cavity, oropharyngeal, or hypopharyngeal squamous cell carcinoma was analyzed to assess disease control, patient compliance, and toxicity.

METHODS

Forty three patients with previously untreated, advanced, resectable squamous cell carcinoma of the oral cavity, oropharynx, or hypopharynx were enrolled in a prospective Phase II institutional clinical trial at a tertiary care comprehensive cancer center. This regimen was a continuum of multimodal treatment in a contracted time interval. It included preoperative slightly accelerated hyperfractionated radiotherapy with concurrent cisplatin, followed immediately with surgery and intraoperative radiotherapy, and completed with early postoperative weekly paclitaxel (beginning on Day 6 after surgery), two additional cisplatin cycles, and concurrent once daily radiotherapy beginning on Day 28 after surgery.

RESULTS

The current trial was designed to reduce the toxicity of the systemic therapy while maintaining or improving locoregional/distant disease control and patient compliance. There were 43 patients enrolled, and the range of time at risk was 2.6 to 24.7 months (median, 14.6 months). Of the 43 registered patients, 43 were evaluable. The locoregional (100%) and systemic (93%) disease control rates were excellent, with low rates of patient noncompliance (21%) and reduced levels of toxicity.

CONCLUSIONS

An intensive treatment regimen that improves disease control and treatment compliance is clearly feasible for this patient population. Future plans include modifications to continue to reduce toxicity and expansion to a multi-center Phase II trial to determine if the single institutional results can be duplicated. Cancer 2002;94:3169–78. © 2002 American Cancer Society.

DOI 10.1002/cncr.10571

Although head and neck carcinoma is relatively infrequent, making up only about 6-7% of all human malignancies,1 it produces a disproportionate health and economic impact on society. The explanation for this fact rests primarily with the reality that the survival rates for the most common disease sites in the head and neck have not improved during the last 60 years. The standard therapy regimens of surgery and postoperative radiation therapy for previously untreated advanced stage resectable cancers arising in the oral cavity, oropharynx, and hypopharynx produce four year survivals of only 38%.2 Multi-modal regimens that have been tested to improve survival over the last 25 years have been encumbered by patient noncompliance, which has compromised the ability to improve survival rates.2

The recognition of the importance of a systemic component to the disease in the 1970s stimulated expanded interest in seeking effective cytotoxic therapy that could be used in a neoadjuvant or adjuvant setting.3–8 The most recently completed Phase III trial in the United States for previously untreated advanced stage resectable disease of the oral cavity, oropharynx, and hypopharynx was sponsored by the National Cancer Institute (NCI) Head and Neck Intergroup.2 Although not showing survival improvement, the study did show a statistically significant decrease in frequency of distant metastases in the experimental arm using chemotherapy. However, the study showed that the combination of locoregional and systemic therapy still produced unsatisfactorily high failure rates for the primary tumors (15.3%), the neck nodes (9.5%), and distant metastases (14.9%).

Recent clinical studies have provided encouraging results showing enhanced ability to control locoregional disease with concurrent chemoradiotherapy.9, 10, 11 Not only have these studies shown improvement in the locoregional control, there are some where survival improvement for certain head and neck carcinoma sub-populations has occurred.12, 13 Merlano et al.12 showed a survival improvement using alternating radiotherapy and chemotherapy for previously untreated patients with unresectable head and neck carcinoma. Five year overall survival with a regimen that included alternating cisplatin and flurouracil with radiotherapy was 24% compared to the standard treatment arm of radiotherapy alone at 10% (P = 0.01). A more recent report by Adelstein et al.13 showed a projected three year overall survival rate of 37% for patients with unresectable cancer utilizing concurrent chemoradiotherapy (with cisplatin). The Phase III Head and Neck NCI Intergroup trial for nasopharyngeal carcinoma using concurrent chemoradiotherapy followed by sequential 5FU and cisplatin resulted in almost a doubling of the control arm's five year overall survival rate of 67%, compared to the control arm, which used radiation therapy alone (37%).10, 14 The feasibility and safety of intraoperative radiation therapy for this patient population has also been previously documented.15, 16, 17

The results of the prior investigations motivated the scientists at our institution to initiate a series of pilot studies evaluating a therapeutic regimen designed to improve the survival and compliance for previously untreated, resectable, advanced stage (Stages III and IV) squamous carcinoma arising in the oral cavity, oropharynx, and hypopharynx. These series of studies report results of modifications in the regimen to achieve the stated goals.18, 19, 20, 21

The purpose of the current report is to present the results of the most recently completed study, whose goals were to sustain excellent locoregional/distant disease control and patient protocol compliance while reducing toxicity.

PATIENTS AND METHODS

The patients originated from the Department of Otolaryngology at our institution. The investigational protocol was reviewed and approved by the university's Institutional Review Board and the Comprehensive Cancer Center's Scientific Review Committee. Eligible patients had previously untreated, resectable squamous cell carcinomas of the oral cavity, oropharynx, or hypopharynx. Patients with resectable disease included those whose preoperative assessment determined the probability of surgically removing all clinically detectable disease with negative histologic margins of resection. The patients must have had Stage III or Stage IV disease according to the 1997 American Joint Committee on Cancer and no distant metastases. Stage II hypopharyngeal carcinoma patients were also eligible. A Karnofsky performance index of 60 or higher, adequate bone marrow function (platelet count > 100 × 109/L and absolute neutrophil count > 2.0 × 109/L), creatinine clearance greater than 1.0 mL/s (> 60 mL/minute), and adequate hepatic function (bilirubin level < 31 umol/L [< 1.8 mg/dL]); serum transaminases four times the upper limit were required. Written informed consent was obtained from all patients before the initiation of therapy. Patients with prior malignant neoplasms were excluded unless they were disease free for five years, had adequately treated basal or squamous cell skin carcinomas, or had in situ cervical carcinoma (because of the excellent prognosis associated with these limited cancers). All patients were examined by a multimodality team consisting of a head and neck surgeon (D.E.S., A.A.), a dentist, a medical oncologist (C.A.R.), and a radiation oncologist (J.C.G. or others). Patients with prior bradyarrhythmias, atrioventricular conduction defects, or marginal cardiac function were eligible but underwent cardiac monitoring during treatment. Forty three patients were registered into the trial from May 24, 1999 through December 9, 2000.

The treatment schema (Table 1) was as follows. The perioperative treatment included preoperative chemoradiation and intraoperative radiotherapy. On Days 1-4, the patients were given a slightly accelerated hyperfractionated boost of external beam radiotherapy consisting of 9.1 Gy with 6 MV X-rays delivered to the primary tumor and clinically involved nodes, excluding the spinal cord. The external beam radiotherapy was divided into seven twice-daily treatments of 1.3 Gy with an interfraction interval of at least six hours. Concurrent cisplatin chemotherapy, 30 mg/m2 per day, was delivered intravenously on Days 1-3. Patients were hydrated intravenously with 1 L of 0.45% sodium chloride with 10 mEq of potassium chloride, 3 g of magnesium sulfate, and 40 g of mannitol for two hours before cisplatin therapy. Surgical resection and intraoperative radiotherapy to the site of closest surgical margin were performed on Day 4. For patients with negative surgical margins (as determined by intraoperative frozen section pathologic analysis), an intraoperative dose of 7.5 Gy was delivered with 6 MeV electrons (prescribed to the 90% isodose).

Table 1. Intensification Treatment Schema
TreatmentPrestudyDays 1–3Day 4Day 10Day 31Day 32
  1. IORT: intraoperative radiotherapy; HDRB: high dose rate brachytherapy; EBRT: external beam radiotherapy.

SurgeryTriple endoscopy and biopsy, gastrostomy tube placement Surgical resection   
IORT  Boost to area of close (7.5 Gy) or positive margins (10.0 Gy) with 6 MeV electrons (90% isodose) or HDRB at 0.5 cm depth   
EBRT Off cord boost 9.1 Gy over seven twice daily treatments (1.3 Gy fractions with 6 MV photons)Completion of off cord boost 9.1 Gy over seven twice daily treatments (1.3 Gy fractions with 6 MV photons)  Begin 40 Gy/20 fractions, 6 MV photons to primary and nodal area, and 45 Gy/20 fractions, 6 MV photons to lower neck and bilateral supraclavicular areas. If (+) node > 3 cm, bilateral posterior neck electron boosts 10 Gy/5 fractions
Cisplatin 30 mg/m2 daily for three days  30 mg/m2 daily for three days every three weeks, two cycles total 
Paclitaxel   45 mg/m2 over three hours weekly, nine cycles total  

On Day 10, the patients began receiving weekly three hour infusions of paclitaxel at 45 mg/m2. All patients were premedicated with dexamethasone, 20 mg orally at 12 and 6 hours before the beginning of paclitaxel infusion or 20 mg intravenously 30 minutes before paclitaxel infusion. The patients received 300 mg of cimetidine hydrochloride intravenously and 50 mg of diphenhydramine hydrochloride intravenously 30 minutes before paclitaxel therapy. Subsequent courses of the same dose of paclitaxel were given as an outpatient weekly treatment on Days 17, 24, 31, 38, 45, 52, 59, and 66 for a total of nine courses of paclitaxel therapy.

On Day 31, after intravenous hydration, patients received a second course of cisplatin, 30 mg/m2 daily for three days. On Day 32, patients underwent external beam radiotherapy with 6 MV X-rays (an additional 40 Gy were delivered in 20 once daily treatments to the primary tumor site and regional draining lymph nodes, and 45 Gy were delivered in 20 once daily treatments to the lower neck and bilateral supraclavicular areas). Parallel opposed upper neck fields were prescribed to midline, and the lower neck/supraclavicular field was prescribed to the 100% isodose. If a histologically positive node larger than 3 cm was present, bilateral posterior neck electron boosts of 10 Gy (at 100% isodose) were delivered in five treatments. The electron energy was chosen to limit the spinal cord dose to less than 45 Gy total. On Day 52, the third course of cisplatin, 30 mg/m2 daily for three days, was given with hydration as previously described.

Recombinant human granulocyte-colony stimulating factor (G-CSF) was allowed at the discretion of the medical oncologist. Prophylactic antibiotic and antifungal coverage consisting of oral ciprofloxacin, 500 mg twice daily, and oral fluconazole, 100 mg daily, was given to patients who developed absolute neutrophil counts < 0.5 × 109/L and continued until the absolute neutrophil count was above 1.5 × 109/L. Radiotherapy was delayed for an absolute neutrophil count of < 0.5 × 109/L and continued when it was greater than 0.5 × 109/L. If the absolute neutrophil count was < 1.5 × 109/L, then chemotherapy was delayed until it rose above 1.5 × 109/L. If the delay in granulocyte recovery was seven days or longer or if the patient had neutropenic fever, subsequent doses of paclitaxel were reduced to 30 mg/m2 (33% dose reduction).

Table 2 gives the demographics of the patient population. This population included 34 men and 9 women. All were white except for three blacks and one Asian. Ages ranged from 30 to 78 years, with a median of 58 years. Twenty patients had oropharyngeal, 15 oral cavity, and 8 hypopharyngeal primary site cancers. Twenty eight percent of patients (12 out of 43) had Stage III clinical disease at presentation. The remaining 72% (31 out of 43) had M0 Stage IV disease.

Table 2. Patient Population
CharacteristicNo. of patients
Gender 
 Male34 (79%)
 Female9 (21%)
Race 
 White39 (91%)
 Black3 (7%)
 Asian1 (2%)
Age 
 Range30–78 years
 Median57.5 years
Tumor site 
 Oral cavity15 (35%)
 Oropharynx20 (46%)
 Hypopharynx8 (19%)
Overall stage 
 Stage III12 (28%)
 Stage IV31 (72%)

In the current group of patients with advanced stage disease, surgical ablation was accompanied by primary closure in 17 patients (39%). The majority of patients (56%, 24 out of 43) required some type of myocutaneous/myofascial flap or free flap reconstruction (Table 3). In addition, 10 patients (23%) also required the use of mandibular plate fixation/reconstruction whenever mandibular osteotomy or mandibulectomy was used either to gain access or to achieve adequate resection of their primary site disease.

Table 3. Types of Reconstructive Techniques Used
Reconstruction typeNo. of patients (%)
Primary closure17 (39%)
Split-thickness skin graft8 (19%)
Reconstruction plate10 (23%)
Myocutaneous flap/free flap24 (56%)

RESULTS

Compliance

Patient directed noncompliance (defined as the patient refusing continuing treatment per protocol) was 21% (9 out of 43). Two patients decided to have their postoperative adjuvant treatment closer to home (both received only postoperative radiotherapy). One patient refused postoperative chemotherapy and missed six postoperative radiation treatments. Four patients refused the remainder of their chemotherapy midcourse but completed radiotherapy as per protocol. Only 2 of 43 patients refused further chemoradiotherapy directly related to toxicity (one Grade 3 mucositis and one Grade 3 nausea/vomiting).

Protocol compliance was defined as receiving all courses of treatment per protocol. Total protocol completion was 53% (23 of 43 patients). The 20 patients (47%) not completing the protocol included 10 (23%) from toxicity, 9 (21%) from patient directed noncompliance, and 1 (2%) from mortality attributed to comorbid conditions.

Toxicity

Toxicity (Table 4) was graded primarily using the RTOG Criteria for Acute and Late Effects. The Southwest Oncology Group/NCI Common Toxicity Criteria were used for categories not covered by RTOG.

Table 4. Toxicities of the Regimen for 43 Patients
Type of toxicityAcuteLate
345345
  • CVA: cerebrovascular accident.

  • Data are given as number of patients. Acute toxicity occurred from day 1 to day 90 and late, after day 90. The numbers 3, 4, and 5 indicate the grade.

  • a

    Grade according to Radiation Therapy Oncology Group Criteria.

  • b

    Grade according to Southwest Oncology Group/National Cancer Institute Common Toxicity Criteria.

Operative      
 Pharyngeal fistula15    
 Flap hematoma1     
 Flap donor site dehiscence1     
 Flap survival failure 1    
Nonoperative      
 Hematologica9     
 Infections requiring hospitalizationb3  1  
 Mucositisa19     
 Gastrointestinala141 1  
 Cardiovascularb 21 1 
 CVAb 4    
 Neuropathyb1     
 Xerostomiaa      
 Hearing lossb 1    

Acute morbidity was defined as adverse results that occurred within 90 days after initiating perioperative treatment and is summarized in Table 4. During this period, one patient died. This patient was a 65 year old man with Stage IV carcinoma of the tongue base and suffered cardiopulmonary arrest of unknown cause on Day 12 (postoperative Day 8). Findings at the time of autopsy suggested cardiac ischemia (prior history of myocardial infarction, poor cardiac ejection fraction of 10-15 %; the patient received one course of cisplatin and one course of paclitaxel).

Four patients (9%) experienced notable (Grade 3) neutropenia or leukopenia, and five patients (12%) experienced Grade 3 anemia. Of the 43 patients from the current series, three (7%) required hospitalization for pneumonia. Nineteen (44%) of 43 patients developed RTOG Grade 3 mucositis. Grade 2 xerostomia developed in 5 (12%) of 43 patients. One patient developed Grade 4 neuropathy (footdrop) during postoperative chemoradiation; the last two cycles of chemotherapy were withheld in that patient. One patient developed sensorineural hearing loss at the end of therapy.

Four patients developed a cerebrovascular accident (CVA). The first patient's CVA followed prestudy staging panendoscopy prior to instituting therapy. There were three CVAs during therapy. One patient developed a CVA on postoperative Day 2. Another was readmitted on postoperative Day 18 with a lacunar-type infarct. The third suffered severe cerebral ischemic insult on Day 66 following an acute episode of massive gastrointestinal bleeding from his gastrostomy tube site that required emergent surgical exploration via open laparotomy. Protocol treatment was discontinued in these three patients.

Six patients (14%) developed pharyngo/orocutaneous fistulae in the earlypostoperative period. Five of the six fistulae required eventual surgical correction, and one healed with local wound care. Two of these patients completed treatment per protocol. The other four patients did not complete treatment per protocol criteria due to toxicity from chemotherapy or patient noncompliance but proceeded to receive adjuvant radiotherapy. One patient developed respiratory distress during her postoperative adjuvant treatment secondary to presumed aspiration/pneumonia and was admitted to an outside facility, after which she then elected to undergo further adjuvant treatment at that facility. Another patient developed adeep venous thrombosis of the left upper extremity at the site of a subclavian central venous catheter on postoperative Day 6. The patient subsequently developed ischemia and gangrene of the left upper extremity and right great toe due to heparin induced thrombocytopenia. For these reasons, postoperative chemotherapy was withheld. Amputations of the left hand and right great toe were performed at two months and six months respectively after his initial surgery. Two patients experienced complications relating to their pectoralis flap donor sites: one patient developed a chest wall hematoma (Day 17) requiring evacuation, and the other patient developed wound dehiscence at his donor site. One patient experienced failure of a radial forearm free flap on postoperative Day 17 that required debridement and reconstruction with a levator scapulae flap.

Late toxicity was defined as adverse events occurring more than 90 days after initiation of treatment. One patient developed a prevertebral abscess necessitating incision and drainage four months after completion of therapy that resulted in transient quadriplegia. This patient subsequently recovered and underwent additional workup that revealed no evidence of recurrent cancer. Six patients experienced problems relating to their mandibular hardware: five patients developed delayed plate exposure externally, and one patient experienced mandibular infection and loosening of her mandibular plate/screws. All five patients required eventual removal of their mandibular hardware.

Locoregional Control/Rate of Distant Metastases

The overall locoregional control rate was 100%. The rate of distant metastases was 6.9% (3 out of 43). In the 23 patients who completed all therapy per protocol, the rate of distant metastases was 8.7% (2 out of 23). One patient (2.3%) developed a second primary lesion (small cell lung carcinoma) three months after beginning treatment for his head and neck carcinoma.

Survival

The 43 patients have been followed for a median time at risk of 14.6 months (range, 2.6–24.7 months). Table 5 lists the outcomes by primary tumor site and stage of disease.

Table 5. Clinical Staging and Patient Outcome
SiteTumor stageNodal statusOutcome
N0N1N2N3
  • NED: alive with no evidence of disease; DOD: dead of disease; DM: distant metastases; DOC: dead of complications; DwoD: dead without evidence of disease; AWD: alive with disease.

  • Staging done using 1997 American Joint Committee on Cancer classification.

  • Numbers indicate patient reference number.

  • a

    Protocol noncompliance.

Oral cavityT1     
 T2     
 T311, 12, 19, 32a4, 34a15a 4, 11, 12, 15, 19, 32, 34 NED
 T427a, 33, 408, 20a14, 24, 25 14, 20, 24, 27, 33, 40 NED
      8 DOD, DM
      25 DOC
OropharynxT1 26  26 NED
 T2  13a, 30, 31a2213, 22, 30, 31 NED
 T323a, 387, 21, 28a, 366, 35a, 37296, 21, 28, 29, 31, 35, 36, 37, 38 NED
      23 DwoD
 T442a10a, 43a9a3, 5a3, 10, 42, 43 NED
      9 AWD, DM
      5 DwoD
HypopharynxT1     
 T2     
 T3  17a, 18a, 39a4, 7a4, 16, 18, 39, 41 NED
     16a, 417 DOD, DM
      17 2nd primary (lung)
 T41  22 NED
      1 DwoD

Thirty five patients (81%) were alive and without evidence of disease at the time of writing. Two patients (5%) were alive with evidence of cancer. Of these two, one patient with a second primary cancer (lung) diagnosed three months after initiation of therapy for Stage IVA pyriform sinus carcinoma developed progression of his lung disease. He did not receive any postoperative taxol/cisplatin secondary to a CVA. The second patient developed brain metastases 13 months after surgery. One patient with Stage IVA hypopharyngeal carcinoma developed a pulmonary metastasis at 13 months and underwent a wedge resection. He subsequently died without evidence of disease from a ruptured abdominal aortic aneurysm six months after the lung resection.

Six patients have died from this study group. One death occurred during treatment as previously described. In addition, three patients died without disease (one from ruptured abdominal aortic aneurysm, one from respiratory complications/pneumonia, one from unknown causes). Two patients died with distant metastatic disease. Only one of these two patients received all of the treatment per protocol.

Survival was defined as the period from the first day of preoperative radiation therapy to the last date of followup or death. Patients alive or who died of nondisease related causes were considered censored at the last date of followup or death. Data were analyzed using Kaplan-Meier techniques. To provide background information in context of the previous intensification pilots carried out at this institution, Kaplan-Meier survival data from previously published data derived from the first two intensification pilots carried out at this institution (IR1 and 2) as well as the current pilot (IR3) are also shown in Table 6 in addition to locoregional and distant disease control rates. Although available for the first intensification pilot, five year survival data will not be available for the second pilot for approximately two more years.

Table 6. Comparative Locoregional/Distant Control Rates and Kaplan-Meier Disease Specific Survival Data on All Registered Patients
Intensification pilot studiesLocoregional recurrencesaDistant metastasesaSurvival (1 year)Survival (2 year)Survival (5 year)
  • IR: preoperative cisplatin/radiation (Days 1–3), surgery/intraoperative and radiation (Day 4), postoperative cisplatin/radiation (Days 29–56).

  • IR 2: identical to IR 1 plus taxol (135 mg/m2) on Days 24, 45, and 66.

  • a

    Percentage; median months at risk.

  • b

    Not yet reached.

IR 18%; 4019%; 4083%76%60%
IR 212%; 257%; 2594%86% 
IR 30%; 14.67%; 14.688%  

DISCUSSION

The current article reports the recent results of continuing investigations which began in 1993 into the feasibility, toxicity, and responses to a multi-modal therapeutic regimen designed to improve survival and patient compliance for patients with previously untreated resectable squamous carcinoma arising in the oropharynx, oral cavity, and hypopharynx. The ultimate goal of showing a survival advantage for a new treatment regimen requires an orderly series of prior investigations that are intended to develop confidence that a new proposed regimen has a substantial chance of showing a survival advantage when compared to standard therapy in a Phase III trial. Until the time when such a Phase III trial is undertaken, completed, and analyzed, there can be no claims made about comparative efficacies among different treatment regimens.

The current regimen was developed with the intent to intensify therapy at the primary tumor site, regional neck nodes, and distant sites because of the previously cited unacceptably high failure rates in all three areas with past trials. The schema is intended for the perioperative therapies to be a continuum with the sole purpose of enhancing locoregional control. The preoperative chemoradiation (Days 1-3) is not imagined to create a response alone, but rather as a component of the entire perioperative regimen that also includes surgery and intraoperative radiotherapy. The slightly accelerated hyperfractionated schema was chosen not to explore hyperfractionation per se, but rather to deliver a clinically significant dose for microscopic disease when used in combination with radiosensitizing cisplatin and a modest dose of intraoperative radiotherapy over a short time period. It is well known that the fraction size determines the rate of late effects/complications of intraoperative radiotherapy and radiotherapy in general.22, 23 Intraoperative radiation therapy, however, can theoretically enhance the primary tumor site disease control. Therefore, the first studies to be completed incorporated intraoperative radiotherapy, and the results showed that the use of this technology was feasible for this patient population and resulted in no increased perioperative complications.14, 24–29 These investigations provided a level of confidence that intraoperative radiation therapy could be a part of the regimen that would enhance primary site disease control.

This perioperative multimodal regimen was combined with early weekly paclitaxel for nine courses and two additional courses of cisplatin to produce both a cytotoxic impact on microscopic distant disease while also providing radiosensitization of the postoperative radiotherapy. Concurrent chemoradiotherapy has been shown to improve local control in the primary treatment of advanced head and neck carcinomas, with some studies showing a survival benefit.26, 30 Recently, the European Organization for Research and Treatment of Cancer has also shown a significant improvement in disease free survival, overall survival, local control, and time to progression with the addition of postoperative cisplatin with concurrent once daily radiotherapy in comparison to postoperative once daily radiotherapy alone in the treatment of locally advanced squamous cell carcinomas of the oral cavity, oropharynx, larynx, or hypopharynx.31 The results of the initial pilot at our institution showed extraordinary locoregional control (95%), excellent compliance (92%), and acceptable toxicities.18 That schema was designed to provide a continuum of therapeutic modalities in a contracted time interval to improve compliance, in contrast to prior multi-modal trials which included separate phases of single modality treatment2, 6, 8 over an extended period of time. Prior trials that included time intervals between treatment modalities seemed to provide an increased opportunity for patient directed refusal of subsequent treatment.

The second pilot was designed to increase the effectiveness of systemic cytotoxicity for this disease. Paclitaxel was identified as the agent to be evaluated based on promising results utilizing it for recurrent/metastatic cancer in the Ohio State University pilots.19 In addition to work by Schuller et al.18 and Grecula et al.,19 paclitaxel was evaluated by using an induction schema for this exact patient population with Stage III or IV previously untreated resectable squamous carcinoma involving the oral cavity, oropharynx, or hypopharynx. This group of 42 patients yielded 4 patients (10%) with complete responses and 17 (40%) with partial responses, for an overall response rate of 50%.20

The results of this paclitaxel induction study showing activity resulted in the intensification regimen being modified by adding paclitaxel in an adjuvant setting. Paclitaxel was administered on Days 24, 45, and 66 at a dosage of 135 mg/m2 for 24 hours with routine granulocyte colony stimulating factor support.19 This pilot study, with a median time at risk of 25 months, showed continuation of an extremely low primary tumor (12%) and regional neck node (9%) treatment failure rate as well as of a decrease in distant metastases (8%, 2 out of 25) when treatment was completed as per the protocol. However, this regimen produced an increased toxicity of the systemic regimen with 2 of the 43 patients experiencing Grade 5 hematologic toxicity, 16 (37%) Grade 4, and 10 (23%) Grade 3. These results prompted the modification of the regimen that is the basis for the current report.

The patients in the original pilot study were subsequently evaluated to assess locoregional treatment response following an extended time at risk. With a median time at risk of 40 months, the overall compliance for the group of patients in the first pilot was 73%, with primary tumor site control being 97% and regional nodal control 95%. The rate of distant metastases in this first group of patients was 19%.21 All of these results from previous pilot studies provided encouragement for the authors to develop an additional modification to this regimen, which was producing prolonged disease control rates.

The results of the current third pilot study show that this group of patients was skewed toward advanced stage disease, with the majority having N2 or N3 neck nodes and 72% of the patient population having Stage IV cancers. The preponderance of patients with large volume disease in the current study presented a formidable challenge for this new therapeutic regimen.

Protocol compliance for the current patient population represents the cumulative effect of comorbid conditions and patient directed noncompliance. Once again, the patient directed noncompliance was substantially lower (21%) than what has been experienced in previously reported multi-modal treatment programs for this identical patient population.6, 18 The patient directed noncompliance rate for the current study is similar to the other two previously reported Ohio State University pilot studies. The patient directed noncompliance for the 123 patients registered to all three pilot studies over the last seven years is 12%. Once again, this compliance rate is better than what has been reported in previous studies of other regimens.6, 32

The frequency of toxicities associated with this treatment regimen reflect the challenges of this patient population with multiple comorbid conditions. Pre-existing cardiac and vascular disease certainly affected the types of toxicities experienced during the current pilot study. However, the frequency of chemotherapy related toxicities and the utilization of G-CSF was markedly decreased from that reported for the second intensification regimen. Although there were four CVAs in this group of 43 patients, one occurred prior to instituting any therapy and one occurred in association with shock associated with a massive gastrointestinal bleed at the site of the patient's gastrostomy tube. The other two patients' CVAs were felt to be related to comorbid vascular conditions rather than adverse effect of therapy on carotid artery bloodflow.

The advent of neoadjuvant and adjuvant chemotherapy and/or radiotherapy regimens initially created concerns among surgeons about the possibility of increased postoperative healing complications. However, those concerns were somewhat dispelled by the results of Phase III trials showing no increased healing complications.6 In the current study, 71% of operations (42 out of 59) involved a variety of reconstructive techniques involving tissue transfer and/or insertion of alloplastic materials. In spite of the increased healing demands associated with these clinical situations, the frequency of pharyngocutaneous fistula (14%) and wound infections (7%) compares favorably to the frequency of the healing complications reported in the literature for patients not receiving neoadjuvant or adjuvant therapy.33

As stated earlier, it is certainly not the intent to claim survival advantage with this treatment regimen compared to other therapeutic approaches, but it is important to continue to assess disease control. The locoregional control rate in the current pilot study was 100% of the 43 patients registered to this trial. The rate of distant metastases was 9% of the 43 patients registered to this trial. These results document that the extraordinary locoregional control rates initially reported for the first intensification pilot study have now been sustained through two subsequent pilot studies.

The results of the first intensification pilot study showed excellent locoregional control. However, the cisplatin dosage used in the current regimen was not sufficient to have a positive impact on the frequency of distant metastases. Even with the presence of distant metastases in the current study, the Kaplan-Meier disease free survivals are still quite encouraging (Table 6).

The second intensification pilot study was intended to modify the original treatment regimen by maintaining the locoregional disease approach and adding a stronger systemic component with the administration of taxol. This taxol regimen was actually determined by a pilot study at our institution involving 45 patients,20 where taxol was used in an induction schema pilot followed by surgery and postoperative radiation therapy. This neoadjuvant schema provided a means of assessing taxol's activity in this group of advanced staged patients with previously untreated advanced stage disease. To our knowledge, no prior reports were present in the literature. The responses noted in this setting showed that taxol as a single agent produced a complete response of 10%, a partial response of 40%, and an overall response rate of 50%.20

The second intensification regimen was modified from the first with the addition of taxol. However, unacceptably frequent hematologic toxicities occurred in that study. Locoregional control for the second pilot study continued to be extraordinarily high. In fact, the frequency of distant metastases in protocol compliant patients decreased from 5 of 30 patients (17%) in the first intensification protocol to 2 of 25 patients (8%). The Kaplan-Meier disease specific survival for this regimen showed 94% at one year and 86% at two years. Five year survival rates for the second intensification pilot will not be available for another two years.

The current most recently completed trial evaluated the intensification regimen modified by decreasing the individual taxol dosages and increasing the frequency of administration so that the total dosage was identical to the second trial where taxol may have decreased distant metastases. The taxol is started relatively early in the regimen as a means of decreasing the time interval for the patients' total treatment and of having an early impact on microscopic distant disease. The current study has shown continued high patient compliance as well as locoregional control and infrequent distant metastases in those patients who complete the treatment per the protocol.

The multi-modal treatment program requires interdisciplinary cooperation and specialized radiotherapeutic technology. However, the promising results of this regimen for a disease where therapy has not been improved in 60 years is motivation to continue assessing feasibility.

Subsequent pilots under development are focused on further decreasing therapeutic toxicities with this current regimen. Following determination of feasibility and an assessment of toxicities, the plan is to expand the evaluation of this regimen by involving other institutions to determine if the single institutional Ohio State University results can be achieved in a multi-institutional setting. The endpoints of locoregional control, distant metastatic rate, and compliance will be used for the multi-institutional Phase II study. The global strategic plan is to continue this progressive step-wise approach of validating unique efficacy until or if this regimen to be proposed for a Phase III trial.

Acknowledgements

The authors acknowledge support from the Ohio State University Comprehensive Cancer Center - Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Head and Neck Oncology Program. The authors also thank Jessie L.-S. Au, Ph.D., Pharm.D., College of Pharmacy, Ohio State University; Marsha L. Hauger, R.N., Division of Medical Oncology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University; Subir Nag, M.D. and Reinhard A. Gahbauer, M.D., Division of Radiation Oncology, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University; Jeffrey R. Haller, M.D., Division of Otolaryngology-Head and Neck Surgery, University of Utah; Carol M. Bier-Laning, M.D., Department of Otolaryngology, Loyola University; and Pramod Sharma, M.D., Division of Otolaryngology-Head and Neck Surgery, University of Michigan.

Ancillary