A phase 1 study of everolimus plus docetaxel plus cisplatin as induction chemotherapy for patients with locally and/or regionally advanced head and neck cancer




Activation of the phosphatidylinositol-3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) pathway is common in head and neck cancers, and it has been demonstrated that inhibition of mTOR complex 1 sensitizes cell lines to platinum and taxane chemotherapy. The authors conducted a phase 1 study to evaluate the addition of oral everolimus to cisplatin and docetaxel as induction chemotherapy for head and neck cancer.


In this single-institution phase 1 study, 3 doses of daily everolimus were explored: 5 mg daily, 7.5 mg daily (administered as 5 mg daily alternating with 10 mg daily), and 10 mg daily of each 21-day cycle. Cisplatin and docetaxel doses were fixed (both were 75 mg/m2 on day 1 of 21-day cycle) at each dose level with pegfilgrastim support. A standard 3 + 3 dose-escalation plan was used. After induction, patients were removed from protocol.


Eighteen patients were enrolled (15 men, 3 women), and their median Karnofsky performance status was 90. The most common toxicities were hyperglycemia, low hemoglobin, fatigue, and thrombocytopenia. Dose-limiting toxicities (DLTs) were neutropenic fever (1 event at dose level 2, 2 events at dose level 3), and all patients recovered fully from these DLTs. The maximum tolerated dose was exceeded at dose level 3. The progression-free survival rate at 1 year was 87.5% (95% confidence interval, 56.8%-96.7%); and, at 2 years, it was 76.6% (95% confidence interval, 41.2%-92.3%). Activating PI3K catalytic subunit α (PIK3CA) gene mutations were identified in 2 human papillomavirus-associated oropharyngeal cancers.


The phase 2 recommended dose was 7.5 mg daily for everolimus plus cisplatin and docetaxel (both at 75 mg/m2 on day 1 of a 21-day cycle) given with pegfilgrastim support. Cancer 2013. © 2013 American Cancer Society.


For patients with locally and/or regionally advanced head and neck squamous cell cancer (HNSCC), primary treatment with induction docetaxel (Taxotere), cisplatin, and 5-fluorouracil (TPF) chemotherapy followed by definitive radiation-based therapy has emerged in recent years as an effective treatment strategy.1, 2 Induction chemotherapy offers theoretical advantages, including early treatment of micrometastatic disease and the potential for improved long-term speech and swallowing function, compared with primary chemoradiotherapy.3, 4 However, TPF can be associated with significant toxicities. The combination of 3 cytotoxic agents is associated with substantial myelotoxicity. In the TAX324 study, 83% of patients experienced grade 3 or 4 neutropenia, and 12% of patients experienced febrile neutropenia during induction TPF.1 5-Fluorouracil can be logistically cumbersome to administer, because it requires a continuous infusion, and it can be associated with significant mucositis in some patients. In view of these concerns regarding the toxicity and logistical challenges associated with TPF, an appropriate question for clinical research is whether 5-fluorouracil can be replaced with a molecularly targeted agent that may be better tolerated and would be less cumbersome to administer.

Everolimus (Afinitor, RAD001) is an orally available rapamycin analog that allosterically inhibits the mammalian target of rapamycin (mTOR), a serine/threonine (Ser/Thr) kinase. mTOR is central regulator of cell growth and survival in the phosphatidylinositol 3-kinase (PI3K)/Akt pathway and is present in 2 intracellular complexes, mTOR complex 1 (mTORC1) and mTORC2.5, 6 Downstream targets of mTORC1 are the protein translation regulator 4E-BP1 and ribosomal S6 protein kinase (p70S6K). mTORC2 regulates the phosphorylation of Akt on Ser473. Everolimus and other rapamycin analogues potently inhibit mTORC1 and also may inhibit mTORC2 in some contexts.7

Clinical study of inhibitors of the PI3K/Akt/mTOR pathway in HNSCC is supported by the findings of the Head and Neck Cancer Tissue Array Initiative, which observed alterations of this pathway in most of the >400 pathology tumor cores that were analyzed.8 In another study that evaluated somatic point mutations and copy number alterations in 26 oral cavity squamous cell cancers, activating alterations in the epidermal growth factor receptor (EGFR)/PI3K pathway were observed in 74% of tumors.9 Additional justification for the study of mTOR inhibitors in the head and neck cancer stems from an extensive preclinical literature, which demonstrates additive or synergistic effects when mTOR inhibitors are combined with platinum and/or taxane agents.10-14

In a phase 1 study of daily everolimus plus low-dose weekly cisplatin for patients with advanced solid tumors, we observed that the regimen was well tolerated.15 One patient with sinus carcinoma experienced a partial response, and 1 patient with advanced oropharyngeal cancer experienced prolonged stabilization of disease. Similarly, everolimus was well tolerated when combined with docetaxel in phase 1, and the phase 2 recommended dose from that study was everolimus 10 mg/daily plus docetaxel 70 mg/m2.16

The current phase 1 study evaluates the addition of everolimus to induction chemotherapy in patients with stage III/IVB HNSCC. At the time this study was designed in 2008, we did not wish to combine everolimus plus 5-fluorouracil because of safety concerns. A phase 1 study of an mTOR inhibitor (CCI-779) plus 5-fluorouracil/leucovorin was closed because of dose-limiting mucositis.17 We postulated that everolimus may provide an effective and safe alternative to 5-fluorouracil during induction chemotherapy in head and neck cancer. Everolimus appears to have a more favorable side-effect profile than infusional 5-fluorouracil and would be less cumbersome to administer.

Therefore, we designed a phase 1 study to define the phase 2 recommended dose of everolimus when added to high-dose cisplatin plus docetaxel as induction chemotherapy for patients with HNSCC, with pegfilgrastim support. We selected the cisplatin dose of 75 mg/m2 that was used in TAX323 rather than the cisplatin dose of 100 mg/m2 that was used in TAX324.1, 2 The cisplatin dose selected (75 mg/m2) reflects concerns about the potential of mTOR inhibitors to intensify the myelosuppressive effects of cytotoxic chemotherapy. In this context, we believed that the lower dose of cisplatin was appropriate from a safety standpoint. The selection of the lower dose of cisplatin also was influenced by preclinical data indicating that mTOR inhibitors may sensitize cancer cells to lower doses of platinum-based chemotherapy, thereby creating an opportunity to develop clinical regimens with submaximal dosing of cisplatin that may be better tolerated than regimens incorporating maximal dose levels of cisplatin.10-14


Patient Eligibility Criteria

This was a single-institution, phase 1 study that was approved by our hospital's institutional review board. All patients provided written informed consent. Eligible patients had previously untreated stage III through IVB HNSCC or nasopharyngeal cancer (World Health Organization type I, II, or III), were aged ≥18 years, and had a Karnofsky performance status ≥70%. Baseline laboratory requirements were an absolute neutrophil count ≥1.5 × 109/L, platelets ≥100 × 109/L, hemoglobin >10 g/dL, serum creatinine ≤1.3 mg/dL, serum bilirubin ≤1.0 mg/dL, activated partial thromboplastin time <1.5 times the upper limits of normal (ULN), and an International Normalized Ratio <1.5. Aspartate or alanine aminotransferase elevations up to 1.5 times the ULN were allowed if the alkaline phosphatase level was ≤2.5 times the ULN.

Patients were excluded for hearing loss requiring hearing aid or intervention; multifocal peripheral sensory alterations or paresthesias interfering with activities of daily living; oxygen saturation ≤88% at rest on room air by pulse oximetry; liver disease, such as cirrhosis or severe hepatic impairment (Childs-Pugh class C); immunization with live vaccine within 1 week of study entry; hyperglycemia grade ≥2 (Common Terminology Criteria for Adverse Events [CTCAE], version 3.0); therapeutic anticoagulation with coumadin; requirement for chronic treatment with steroids (prednisone 5 mg daily); impairment of gastrointestinal function or gastrointestinal disease that could significantly alter the absorption of everolimus; known infection with human immunodeficiency virus; or other active infection or serious underlying medical condition that would impair the patient's ability to receive protocol treatment.

Treatment Plan

Patients received an induction chemotherapy regimen consisting of daily everolimus (per the dose-escalation scheme) plus concurrent docetaxel and cisplatin (both 75 mg/m2 intravenously on day 1, every 21 days). The dose escalation was for everolimus only, and the doses of cisplatin and docetaxel were fixed. The study regimen is provided in Table 1. Pegfilgrastim (6 mg subcutaneously) was administered on day 2 of each cycle. Three cycles of induction chemotherapy were planned for each patient. After induction chemotherapy, patients were removed from the study, and definitive management of the primary tumor/involved neck lymph nodes was at the discretion of the treating team, according to standard practice.

Table 1. Dose-Escalation Scheme
Dose LevelEverolimus, Oral or by PEG Tube on Days 1-21 of Each 3-Week CycleDocetaxel, Intravenously on Day 1, Every 21 days, mg/m2Cisplatin, Intravenously On Day 1, Every 21 Days, mg/m2
  • Abbreviations: PEG tube, percutaneous endoscopic gastrostomy tube; QOD, every other day.

  • a

    Dose level 2 (everolimus 7.5 mg/day) is administered as 5 mg/day and 10 mg/day on alternating days, because 7.5-mg tablets are not available.

15 mg/d7575
2a5 mg/QOD (on odd-numbered days) alternating with 10 mg QOD (even numbered days)7575
310 mg/d7575

Dose Escalation and Definition of Dose-Limiting Toxicity

A standard 3 + 3 dose escalation plan was followed. Patient 1 was enrolled at dose level 1. If 1 of the initial 3 patients in a dose level experienced dose-limiting toxicity (DLT), then 3 additional patients would be enrolled at that dose level. If 0 of 3 patients or if 1 of 6 patients experienced DLT at a given dose level, then 3 additional patients could be enrolled at the next higher dose level. Completion of 1 cycle without DLT by all patients at any dose level was required before any patients could be enrolled at the next dose level. If 2 or more patients in any cohort of up to 6 patients experienced DLT during cycle 1, then the maximum tolerated dose (MTD) was exceeded. The MTD was defined as the highest dose level at which ≤1 of 6 patients experienced DLT during cycle 1.

DLT definitions were modified from prior studies of induction chemotherapy in head and neck cancer.18, 19 DLT was defined as grade 4 neutropenia for more than 7 days, febrile neutropenia (absolute neutrophil count <1.0 × 109/L, fever ≥38.5°C), grade 4 thrombocytopenia, or any grade 3 toxicities other than those specified as non-DLT in the protocol. The protocol specified that only those events that were judged as possibly, probably, or definitely related to everolimus were deemed DLTs. Adverse events that we believed were related to cisplatin, docetaxel, or head and neck cancer, at the investigator's discretion, were not necessarily considered DLTs.

Safety Evaluations

Before each cycle of therapy, patients underwent clinical evaluation and comprehensive laboratory tests (complete blood cell count, basic metabolic panel, liver function tests, and lipid panel). Patients were evaluated weekly in the clinic during cycle 1, including routine laboratory studies (complete blood cell count, basic metabolic panel). During cycles 2 and 3, patients were evaluated in the clinic on day 1 and underwent at least 1 additional clinic evaluation per cycle with routine laboratory studies. Patients were followed for toxicity for 30 days after the last dose of everolimus. Adverse events were graded according to National Cancer Institute's CTCAE, version 3.0. Patients who experienced DLT during cycle 1 were removed from the study, and subsequent management off protocol was according to standards of care for head and neck cancer. Patients who maintained active follow-up underwent a safety assessment approximately 30 days after the off-study date.

Efficacy Assessments

Pretreatment assessment included cross-sectional imaging of the neck (computed tomography of the neck with contrast or magnetic resonance imaging with gadolinium) within 6 weeks before therapy. Restaging imaging studies occurred within 2 to 6 weeks after the final treatment with cisplatin/docetaxel (typically during cycle 3 for patients who were not removed from the study earlier). Radiologic response criteria were based on previously published Response Evaluation Criteria in Solid Tumors (RECIST) guidelines.20 However, confirmation of response was not obtained, because such confirmatory radiologic assessments would be confounded by the impact of subsequent definitive locoregional therapy off protocol. Overall survival and progression-free survival were estimated according the method of Kaplan and Meier.

Human Papillomavirus Assay

For most patients with primary tumors of the base of tongue or tonsil, the presence or absence of high-risk human papillomavirus (HPV) strains was determined using the Ventana HPV III family 16 probe (Ventana Medical Systems, Inc., Tucson, Ariz).21 For patients with other primary sites, HPV testing is not routinely obtained in our clinical practice.

Mass Spectrometry Genotyping

Mass spectrometry genotyping using the iPLEX system (Sequenom, San Diego, Calif) was performed according to established methods.22, 23 Genomic DNA was extracted from formalin-fixed, paraffin-embedded specimens. Polymerase chain reaction amplification of 100-base pair to 150-base pair fragments with primers bracketing the mutation of interest was conducted. Extension primers designed immediately adjacent to the mutation site allowed extension by 1 nucleotide, depending on the template sequence, and allelic-specific differences in mass between extended products were measured by time of flight after laser deionization. Wild-type and mutant alleles were distinguished with the use of Sequenom software. We used the multiplexed system developed by Dr. Solit and colleagues for genotyping PIK3CA, AKT1, KRAS, NRAS and BRAF.24


Patient Characteristics

Between August 25, 2009 and April 22, 2011, 18 patients were enrolled. Baseline characteristics of the entire enrolled study population are summarized in Table 2. Among 18 enrolled patients, most were men (n = 15), and median the age was 56 years (range, 44-70 years). The median Karnofsky performance status was 90 (range, 80-100). The most common primary tumor site was the oropharynx (n = 14). Most patients had stage IVA disease (n = 14).

Table 2. Characteristics of the Study Population, n = 18
CharacteristicNo. of Patients (%)
  • a

    Thirteen patients had primary tumor in base of tongue and/or tonsil; 1 patient had soft palate primary. The analysis includes 1 patient with tonsil squamous cell carcinoma who was enrolled but never treated.

 Men15 (83)
 Women3 (17)
Age: Median [range], y56, [44-70]
Karnofsky performance status; Median [range]90 [80-100]
Primary site 
 Nasal cavity1
 Oral cavity1
Stage, clinical 
Tobacco history, pack-years 
 Never smoker of cigarettes7
 >0 to ≤101
 >10 to ≤359

Among 14 patients with oropharyngeal primary tumors, 7 were positive for the p16 immunohistochemistry assay. The p16 assay was not done in the other oropharyngeal tumors. Six of the patients who had oropharyngeal squamous cell carcinoma were never-smokers.

Dose Escalation and Description of Dose-Limiting Toxicities

There were no DLTs among 3 patients who were treated at dose level 1, and enrollment for dose level 2 was initiated. Eight patients were enrolled at dose level 2. Two patients who were enrolled at dose level 2 were deemed inevaluable for toxicity assessment and were replaced. The safety evaluation was based on 6 evaluable patients at dose level 2, and there was 1 DLT. He was a man aged 56 years with stage IVA oropharyngeal squamous cell carcinoma who developed neutropenic fever during cycle 1. He recovered fully from this event and completed treatment for his cancer off protocol. Also at dose level 2, 1 patient was removed from the study during cycle 2 because of a grade 2 microperforation of the small bowel, which was not deemed to be a DLT.

Regarding the 2 patients at dose level 2 who were deemed inevaluable for safety, 1 was a woman aged 70 years with stage IVB oropharyngeal squamous cell carcinoma who had baseline grade 1 hyponatremia. She was admitted to hospital on day 5 of cycle 1 with grade 3 hyponatremia. The patient and her family indicated that she had only taken 2 doses of everolimus instead of daily dosing as prescribed, but a pill diary was not provided. It was believed that her toxicities were related to cisplatin and that everolimus probably was not related. She was removed from study because of excessive cisplatin-related toxicities and lack of adherence to the everolimus regimen. She was discharged in stable condition after a 10-day hospitalization, during which hyponatremia worsened to grade 4 and then resolved to baseline. The other patient who was deemed inevaluable at dose level 2 was a man aged 50 years with stage IVA oropharyngeal squamous cell carcinoma. He tolerated treatment well; however, during cycle 3, the investigator learned that the patient had been taking everolimus irregularly. The patient presented pill diaries indicating appropriate study drug administration; but, after extensive discussion, his pill diaries were deemed unreliable. Thus, the patient was deemed inevaluable for the safety endpoint. After 5 of 6 evaluable patients completed cycle 1 at dose level 2 without DLT, the decision was made to begin enrollment at dose level 3.

Seven patients were enrolled at dose level 3, but 1 was never treated on the study. Among 6 evaluable patients at dose level 3, there were 2 DLTs. One patient was a man aged 60 years with stage IVA oropharyngeal squamous cell carcinoma, and the other was a man aged 59 years with stage IVA oral squamous cell carcinoma. Both events were febrile neutropenia, and the patients recovered fully. Another patient at dose level 3, a man aged 55 years with stage IVA oropharyngeal squamous cell carcinoma, was removed from study after cycle 1 because of poor tolerance in the context of an episode of acute prostatitis and transient creatinine elevation that the investigator did not believe constituted DLT.

We deemed that the MTD was exceeded at dose level 3 because of the 2 episodes of febrile neutropenia. Dose level 2 was designated the MTD, because there was only 1 DLT among 6 evaluable patients at dose level 2.

Adverse Events Associated with Induction Chemotherapy

Fifteen patients were evaluable for toxicity among 18 enrolled patients. Not included in the safety analysis are 2 patients who were removed from study because of noncompliance with oral everolimus and 1 patient who was never treated on study. Table 3 lists all adverse events, regardless of attribution, that occurred in >30% of the study population regardless of grade or were ≥3 grade in at least 2 patients. The most common adverse events of any grade were hyperglycemia, low hemoglobin, thrombocytopenia, and fatigue. The relation between everolimus and hyperglycemia was difficult to evaluate because of the contributory effect of steroids given as the standard of care with docetaxel and cisplatin chemotherapy. The most common grade 3 or 4 adverse events were leukopenia and lymphopenia. There were no treatment-related deaths. There were no instances of grade 3 mucositis on clinical examination, but 3 patients (20%) experienced grade 2 mucositis on clinical examination. One patient with grade 2 mucositis on clinical examination had grade 3 functional mucositis according to CTCAE version 3.0.

Table 3. Summary of Adverse Events, Regardless of Attribution
 No. of Adverse Events (%)
Adverse EventAny GradeGrade 3Grade 4
  1. Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase.

Hyperglycemia15 (100)5 (33)1 (7)
Low hemoglobin14 (93)1 (7)0 (0)
Thrombocytopenia13 (87)1 (7)1 (7)
Fatigue13 (87)2 (13)0 (0)
ALT elevation12 (80)0 (0)0 (0)
Hypercholesterolemia12 (80)0 (0)0 (0)
Hypertriglyceridemia12 (80)0 (0)0 (0)
AST elevation11 (73)0 (0)0 (0)
Leukopenia10 (67)6 (40)1 (7)
Lymphopenia10 (67)6 (40)4 (27)
Hypomagnesemia10 (67)0 (0)0 (0)
Hypoalbuminemia9 (60)0 (0)0v
Diarrhea9 (60)0 (0)0 (0)
Nausea9 (60)0 (0)0 (0)
Hyponatremia9 (60)1 (7)0 (0)
Constipation8 (53)0 (0)0 (0)
Creatinine elevation8 (53)0 (0)0 (0)
Rash, acneiform8 (53)0 (0)0 (0)
Alkaline phosphatase elevation7 (47)0 (0)0v
Neutropenia7 (47)3 (20)3 (20)
Hypokalemia7 (47)3 (20)0 (0)
Dysgeusia6 (40)0 (0)0 (0)
Febrile neutropenia3 (20)2 (13)1 (7)

Subsequent Definitive Therapy

Patients were removed from study after induction chemotherapy on this protocol. Subsequent management was at the discretion of the treating physician and was not specified by the protocol. The reported adverse events are only those associated with induction chemotherapy, and not those associated with subsequent definitive therapy. Among 18 enrolled patients, 16 received subsequent definitive radiation therapy at this center. The concurrent chemotherapy regimens for these patients were weekly cisplatin (n = 5), cetuximab plus experimental agent on another clinical trial (n = 4), weekly carboplatin plus paclitaxel (n = 2), weekly cisplatin plus experimental agent on another clinical trial (n = 2), weekly cisplatin plus paclitaxel, weekly cetuximab, and bolus cisplatin every 3 weeks (n = 1 each). One patient was enrolled but not treated on the study and received therapy with his local oncologist. The woman aged 70 years with oropharyngeal cancer at dose level 2 who was removed from study during cycle 1 because of excessive cisplatin-related toxicities and noncompliance with everolimus declined subsequent anticancer therapy.

Tissue Correlates

Table 4 indicates the tumor site, clinical staging, HPV status, and mutational profiling results. Among 12 treated patients who had primary tumors arising in the base of tongue or tonsil, 7 had tumors that contained high-risk HPV DNA, 4 were not tested, and 1 was HPV-negative. Two HPV-associated squamous cell cancers of the base of tongue were identified that harbored activating mutations in the helical domain of PIK3CA. One HPV-negative squamous cell carcinoma of the oral cavity tested positive for a KRAS mutation. No other mutations were detected among the tumor samples from patients in the study with this assay panel.

Table 4. Additional Patient and Tumor Characteristics, With Response Assessmentsa
Dose LevelAge, ySexCigarette History, Pack-YearsHPV ISH StatusMutational Profiling ResultPrimary Tumor, Site, and StageBest Response or Reason for I nevaluable Status
  • Abbreviations: AE, adverse event; BOT, base of tongue; DLT, dose-limiting toxicity; HPV, human papillomavirus; ISH, immunohistochemistry; KRAS, A146T, alanine to threonine substitution at codon 146; PIK3CA, E545Q, glutamic acid to glutamine substitution at codon 545.

  • a

    This does not include 1 patient who was enrolled but never treated (tonsil primary, never smoker).

  • b

    WT indicates that no mutations were detected in the assay panel for the study.

156Woman30Not doneWTbHypopharynx, T3N2bM0−21%
154ManNeverPositiveWTBOT, T2N2cM0−47%
144Man11Not doneWTNasal Cavity, T3N2cM0−22%
270Woman50Not doneWTBOT, T2N3M0Removed from study in cycle 1 because of noncompliance with oral regimen and cisplatin-related toxicities
260Man11Not doneWTSoft palate, T2N2bM0Off study because of AE in cycle 2
248Woman10Not doneWTBOT/tonsil, T2N3M0−11%
256Man35PositiveWTBOT, T1N2bM0Off study because of DLT in cycle 1
252ManNeverNot doneWTBOT, T3N2cM0−64%
246ManNeverNegativeWTBOT, T1N2bM0−32%
250ManNeverNot doneWTTonsil, T1N2bM0Noncompliance with oral regimen
256ManNeverNegativeWTNasopharynx, T1N2M0−27%
368Man23PositiveWTTonsil, T2N2cM0−53%
360M30PositiveWTBOT, T3N3M0Off study because of DLT in cycle 1
357Man30PositiveWTTonsil, T4aN2cM0−39%
359Man18PositivePIK3CA, E545QBOT, T2N2bM0−58%
355ManNeverPositivePIK3CA, E545QBOT, T2N2aM0Off study because of AE in cycle 1 (not DLT)
359Man30NegativeKRAS, A146TOral cavity, T4aN2bM0Off study because of DLT in cycle 1


Patients who received at least 2 cycles of study treatment were deemed evaluable for radiologic response evaluation. Table 4 describes the radiologic response data for the 10 patients who were evaluable for response and gives the reasons for inevaluable status for the patients who were treated but were not evaluable for response. Among evaluable patients, all experienced decreases in measurable disease, and 6 met criteria for an unconfirmed response.

Survival analysis for the study population is descriptive only. The survival data cutoff date was June 30, 2012. Among 17 patients who received any treatment on study (median follow-up, 17.8 months), the 1-year overall survival rate was 100%, and the 2-year overall survival rate was 91% (95% confidence interval, 50.8%-98.7%). Progression-free survival at was 87.5% at 1 year (95% confidence interval, 56.8%-96.7%) and 76.6% at 2 years (95% confidence interval, 41.2%-92.3%).


The current study demonstrated the safety and tolerability of a novel induction chemotherapy regimen in which everolimus was added to standard doses of cisplatin and docetaxel (both 75 mg/m2 on day 1 of a 21-day cycle, with pegfilgrastim support) for patients with locally and/or regionally advanced head and neck cancer. The phase 2 recommended dose of everolimus is 7.5 mg daily, administered as 5 mg daily and 10 mg daily on alternating days. The DLT in this study was neutropenic fever.

Previous phase 1 studies that led to development of the TPF regimen are instructive with regard to interpreting data from the current study.18, 19 In the prior studies, a dose level would be deemed unacceptable if 3 of 6 patients developed the same DLT.18, 19 The current study defined MTD as the maximum dose level at which ≤1 of 6 patients experienced any DLT that the investigators believed was related to everolimus. If the DLT definition in the current protocol had been more similar to the DLT definitions used in the phase 1 studies for TPF, then it is possible that the 10-mg daily dose of everolimus could have been deemed acceptable. However, in accordance with the dose-escalation rules in the protocol, the recommended dose of everolimus in this study was 7.5 mg daily.

At dose levels 2 and 3, there was a wide range of tolerability of this regimen among the enrolled patients. For some patients, a dose reduction in everolimus to 5 mg daily may be necessary if significant adverse events occur. We note that pharmacokinetic data from the phase 1 study of everolimus monotherapy indicated that most patients achieve therapeutic levels of everolimus at a dose ≥5 mg daily.25 Thus, the combination of everolimus 5 mg daily plus docetaxel and cisplatin would likely constitute exposure to therapeutic everolimus levels for most patients who required this dose reduction.

For patients with head and neck cancer who are receiving induction chemotherapy, prophylaxis to reduce the risk of neutropenic fever is the standard of care. The TAX323 and TAX324 studies featured prophylactic ciprofloxacin on days 5 through 15 of each cycle.1, 2 The current protocol did not include prophylactic ciprofloxacin because of theoretical concerns about the potential emergence of resistant organisms with prophylactic antibiotic use and concerns about potential drug-drug interactions because of the cytochrome P450, family 3, subfamily A, polypeptide 4 (CYP3A4)-inhibitory effects of ciprofloxacin.26, 27 Further study would be necessary to determine whether there is any clinical benefit or harm associated with pegfilgrastim prophylaxis versus antibiotic prophylaxis for patients with head and neck cancer who are receiving the induction chemotherapy regimen of everolimus plus cisplatin plus docetaxel.

When patients received TPF with prophylactic ciprofloxacin in large randomized phase 3 trials, the incidence of febrile neutropenia was 5% in the TAX323 study and 12% in the TAX324 study.1, 2 The incidence of febrile neutropenia in the current phase 2 study was 20% (3 events among 15 evaluable patients). However, 2 of the episodes of febrile neutropenia in our study occurred at dose level 3, which was deemed to exceed the MTD. To estimate the incidence of febrile neutropenia with everolimus plus cisplatin and docetaxel, a larger phase 2 study of the dose level 2 regimen would be informative.

The DLTs observed in this study were related to neutropenia, consistent with prior observations regarding the additive myelosuppressive effects of mTOR inhibitors when combined with platinum-based and/or taxane-based chemotherapy. A shared feature of those studies is the clinically apparent intensification of myelosuppression with the addition of an mTOR inhibitor.28-33 In a phase 1 study of patients with recurrent or metastatic squamous cell cancer, we observed that full-dose temsirolimus (25 mg per week intravenously) could be safely administered with low-dose weekly carboplatin and paclitaxel on days 1 and 8 of a 21-day cycle, but not with a more intensive schedule of carboplatin and paclitaxel because of myelosuppression.33

In a phase 1 study of weekly everolimus plus docetaxel every 3 weeks for patients with metastatic breast cancer, the trial was terminated in the context of excessive neutropenia.31 In a phase 1 study of patients with recurrent/refractory nonsmall cell lung cancer, the recommended dose was everolimus 5 mg daily (on days 1-19) plus docetaxel 60 mg/m2 every 3 weeks.32 The ability to deliver everolimus 7.5 mg daily with full-dose docetaxel and cisplatin (both 75 mg/m2, every 21 days) in the current study differs from those experiences and may be related to several factors. All patients in the current study were previously untreated, whereas 80% of patients in the breast cancer study had received >2 prior chemotherapy regimens for metastatic disease,31 and most patients in the lung cancer study had received at least 1 prior regimen.32 All patients in our head and neck cancer study received pegfilgrastim support, which was not a component of the breast cancer or lung cancer protocols. Potentially, the intensification of chemotherapy-induced myelosuppression may be less severe with everolimus administered at low doses on a daily basis compared with the weekly administration of everolimus at higher dose in the breast cancer study.

Tissue correlative assays in this study identified 2 squamous cell carcinomas of the base of tongue with activating mutations in the helical domain of PIK3CA, and both of those tumors were associated with HPV. Larger studies are needed to determine whether activating PIK3CA mutations may occur in a clinically significant proportion of HPV-associated tumors and whether such lesions may be associated with the efficacy of everolimus. Activation of the PI3K/Akt/mTOR pathway is a characteristic of both HPV-positive and HPV-negative HNSCC,34, 35 and there appear to be multiple mechanisms that activate this pathway in head and neck cancers. Loss of phosphatase and tensin homolog (PTEN) may be more common in HPV-negative tumors than in HPV-positive tumors.36 Preclinically, it has been observed that the HPV 16 E6 oncoprotein activates mTORC1 signaling and cap-dependent translation.37, 38 Further delineation of the mechanisms of mTOR activation in HPV-positive and HPV-negative tumors would guide the interpretation of data regarding candidate biomarkers in future phase 2 studies involving everolimus.

This phase 1 study defines an induction chemotherapy regimen in which patients receive everolimus (7.5 mg daily) plus docetaxel and cisplatin (both 75 mg/m2 on day 1 of a 21-day cycle) with pegfilgrastim support. Efficacy assessments are descriptive only in this small phase I study in which patients subsequently received definitive therapy for local/regional disease off protocol. A larger phase 2 study of this induction chemotherapy regimen followed by definitive locoregional therapy to assess efficacy and explore candidate biomarkers is warranted. We have recently reported the results from a phase 1 study of everolimus plus low-dose weekly cisplatin and definitive intensity-modulated radiation therapy (IMRT) for patients with head and neck cancer,39 and that chemoradiation regimen could be used after induction chemotherapy with everolimus plus docetaxel and cisplatin in an efficacy-based phase 2 study.


This investigator-initiated study was supported with funding from Novartis Pharmaceuticals.


Dr. Matthew G. Fury and Dr. Alan L. Ho have served on advisory boards for Novartis, the funding entity for this study.