Professional Documents
Culture Documents
Harrington 2015
Harrington 2015
4370
The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2015.61.4370
than 16,000 patients with SCCHN in randomized trials that showed 50 Gy to the low-risk planning target volume 1; Appendix, online only) and
that concomitant chemotherapy added significant benefits in 5-year lapatinib (1,500 mg taken orally once per day) or placebo. This was followed by
event-free survival and overall survival (OS).2 Addition of chemother- administration of lapatinib monotherapy (1,500 mg orally once per day) or
placebo for 1 year in the maintenance phase or until disease relapse. Carbopla-
apy to postoperative radiotherapy has also been tested in two large
tin was allowed if cisplatin could not be given after the first or second cycle
phase III randomized studies in high-risk patients with locally ad- because of nephrotoxicity. Dose reductions or delays in randomly assigned
vanced SCCHN. Both studies reported improved disease-free survival treatment were permitted in case of grade 3, 4, or unacceptable adverse events
(DFS) or progression-free survival, although only one study reported (AEs) or AEs consistently associated with lapatinib. An independent program
improved OS.1,3 Long-term results of one of these trials, Radiation of quality assurance and analysis of radiotherapy plans (Radiotherapy Quality
Therapy Oncology Group 95-01, showed relatively high locoregional Assurance) was implemented (EqualEstro, Paris, France), with the aim of
failure rates and poor DFS (⬍ 30% in both arms), suggesting the need ensuring standardized delivery of radiotherapy across treatment arms.
At the time of completion or early discontinuation of study treatment,
for more effective adjuvant therapy in patients with operable SCCHN patients were observed for recurrence every 4 months for 2 years and then
at high risk of relapse.4 every 6 months until withdrawal from the study or death. After disease recur-
Although previous reports suggested that epidermal growth fac- rence, patients were monitored remotely every 6 months for survival outcome.
tor receptor (EGFR, c-ErbB1) is overexpressed in up to 90% of All patients gave written informed consent; the study was performed in accor-
SCCHN tumors, recent data from the Cancer Genome Atlas Network dance with good clinical practice guidelines and the Declaration of Helsinki
suggest that this is a significant overestimate.5 Even so, EGFR-targeted and approved by local ethics committees.
therapies are potentially beneficial for treating SCCHN,6 and this Patient Population
approach has been validated in a phase III trial that established a new Eligible patients had a histologically confirmed diagnosis of SCCHN at
standard of care for locally advanced SCCHN using the chimeric the oral cavity, oropharynx, hypopharynx, or larynx; pathologic stage II, III, or
anti-EGFR antibody cetuximab combined with radiotherapy.7 Cetux- IVA cancer and ECE of nodal disease and/or positive resection margin of ⱕ 5
imab, as monotherapy or in combination with cisplatin, has also been mm but with no evidence of gross residual disease; primary surgery with a
associated with clinical efficacy in patients with platinum-refractory curative intent completed within 4 to 6 weeks (no later than 8 weeks) before
random assignment; and complete recovery from the surgical procedure.
SCCHN.8,9
Details of exclusion criteria are included in the Appendix.
Heterodimerization and functional cross-talk between members
of the EGFR/ErbB family suggest that combining chemoradiotherapy Study End Points and Assessments
with multitargeted small-molecule tyrosine kinase inhibitors may The primary end point was DFS (time from random assignment to the
be a successful approach for the treatment of SCCHN. To date, earliest date of disease recurrence or death as a result of any cause) based on
EGFR-targeted agents such as humanized monoclonal antibodies10 or radiologic and/or clinical assessment of electronic case report form data by an
independent review committee. DFS was also assessed by the investigator
tyrosine kinase inhibitors,11,12 tested concomitantly with (chemo)ra-
(based on scans, endoscopies, or positive biopsy results).
diotherapy, have failed to demonstrate significant benefits in random- Secondary efficacy end points included OS (time from random assign-
ized trials, emphasizing the need for alternative EGFR-targeting ment until death as a result of any cause); disease-specific survival; time to
agents. One such agent is lapatinib, a small-molecule inhibitor of locoregional control; time to distant relapse; incidence of second primary
EGFR and human epidermal growth factor receptor 2 (HER2, ErbB2). tumor; and clinical outcome with biomarkers including DFS by human pap-
The combination of lapatinib and chemoradiotherapy was well toler- illomavirus (HPV) status15 and EGFR status. Medical resource utilization
ated and showed single-agent activity in a small number of patients (collected via the electronic case report form) and health-related quality of life
(HRQoL) were also evaluated. HRQoL was assessed using the Functional
with locally advanced SCCHN in phase I and II trials.13,14 Therefore, a Assessment of Cancer Therapy–Head and Neck and EQ-5D questionnaires
randomized, placebo-controlled, phase III study was designed to eval- (further details are provided in the Appendix). Other post hoc analyses
uate the efficacy and safety of lapatinib administered concomitantly included DFS in patients with or without ECE and with different
with chemoradiotherapy, followed by maintenance monotherapy for surgical margins.
1 year, in patients with resected SCCHN at high risk of relapse. Exposure to chemoradiotherapy was reported descriptively. Compliance
was assessed by tablet count. Safety end points included toxicities (including
late radiation morbidity), measured by recording the incidence and grading of
PATIENTS AND METHODS AEs and serious AEs (SAEs).
Tumor EGFR protein expression levels were evaluated by IHC. HPV
status was determined by evaluating the intratumoral expression levels of
Study Design and Treatment CDKN2A (p16) by IHC and the presence of DNA from high-risk HPV geno-
This was a randomized, double-blind, placebo-controlled, multicenter, types (HPV16 and HPV18) by chromogenic in situ hybridization (Appendix
phase III study (ClinicalTrials.gov identifier: NCT00424255; GlaxoSmithKline Table A1, online only). Samples that were p16 positive/HPV negative were
protocol number: EGF102988), conducted between December 2006 and re-examined by consensus high-risk HPV polymerase chain reaction for the
November 2013. Patients with resected high-risk SCCHN (defined as ECE of detection of HPV DNA. Samples that were p16 positive/HPV positive were
nodal disease or positive resection margin [ⱕ 5 mm]) were randomly assigned defined as harboring biologically relevant oncogenic HPV infection. Further
in a 1:1 ratio to lapatinib or placebo arms using a GlaxoSmithKline internal details are provided in the Appendix.
random assignment system, stratifying by nodal stage (N0 or N1 v N2), Clinical safety and tolerability were assessed in the safety population
primary tumor site, geographical region, and EGFR expression (immunohis- (comprising all patients who took at least one dose of study medication). AEs
tochemistry [IHC] EGFR3⫹ v non-EGFR3⫹). were graded according to the National Cancer Institute Common Terminol-
At the start of the treatment phase, randomly assigned treatment (lapa- ogy Criteria for Adverse Events (version 3) and coded using the Medical
tinib or placebo) was administered once daily for 3 to 7 days before commenc- Dictionary for Regulatory Activities (version 15.1).
ing the 6- to 7-week combined therapy phase, where patients received
postoperative adjuvant concurrent chemoradiotherapy (intravenous cisplatin Statistical Analysis
100 mg/m2 per day on days 1, 22, and 43 plus 2 Gy of radiation per day, 5 days The planned sample size was 680 patients (340 patients per arm). To
per week, for a total of 66 Gy to the high-risk planning target volume 2 and detect a hazard ratio (HR) of 0.7206 with 80% power, approximately 298
independently assessed events (disease recurrences or deaths as a result of any treatment arms had radiotherapy plans that were approved by the
cause) would be required for the final analysis. The expected accrual rate was independent quality assurance group.
15 patients per month for 45 months. Median follow-up time (with 95% CI)
was calculated using the reverse censoring method.16,17 Further details on the
statistical properties of the trial design are provided in the Appendix. Efficacy
All efficacy analyses were conducted on the intent-to-treat (ITT) popu- According to the study design, approximately 298 independently
lation, which comprised all patients randomly assigned to study treatment, assessed DFS events (disease recurrence or death) would be required
regardless of whether they received study medication. DFS and OS were for the final analysis. However, because of apparent plateauing of
summarized using Kaplan-Meier curves (all P values are two-sided unless investigator-reported events (and having reached a median follow-up
otherwise stated). Further details are provided in the Appendix. time of 35.3 months), it was decided to report the study using data
collected up to a clinical cutoff of March 29, 2013. At this time, there
RESULTS were 232 independently assessed DFS events.
A 1.0
B 1.0
CRT + lapatinib CRT + lapatinib
CRT + placebo CRT + placebo
Disease-Free Survival
Disease-Free Survival
0.8 0.8
(probability)
(probability)
0.6 0.6
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70
Time Since Random Assignment (months) Time Since Random Assignment (months)
No. at risk No. at risk
Lapatinib 346 215 177 130 88 42 11 1 Lapatinib 346 237 188 135 90 44 11 1
Placebo 342 209 172 127 89 45 8 Placebo 342 224 183 131 91 47 11
Fig 2. Kaplan-Meier estimates of disease-free survival (DFS) in intent-to-treat population: (A) independently assessed and (B) investigator assessed. CRT,
chemoradiotherapy; NR, not reached.
placebo and lapatinib arms, respectively. Lymphopenia (5% of pa- occurrence of late radiation morbidity events was similar between
tients in both arms) and mucosal inflammation (nine patients [3%] the placebo and lapatinib arms (20% and 16% of patients,
and 17 patients [5%] in the placebo and lapatinib arms, respectively), respectively), and grade ⱖ 3 events occurred in 4% of patients in
were the most commonly experienced SAEs. both arms.
Sixteen patients (5%) and 25 patients (7%) experienced fatal AEs
in the placebo and lapatinib arm, respectively. Sepsis and pneumonia
were the most common known reasons for fatal SAEs (ⱕ 1% of DISCUSSION
patients in each arm). Two patients (⬍ 1%) in the placebo arm and
three patients (⬍ 1%) in the lapatinib arm had fatal AEs that were Despite a strong scientific rationale and supportive data from earlier
considered related to study medication by the investigator. The phase I and II studies, this phase III trial of lapatinib in combination
0 1 2 3 4 5
1.0 CRT + lapatinib Table 3. Summary of On-Therapy and Follow-Up Overall Incidence of
CRT + placebo Adverse Events Occurring in ⱖ 15% of Patients in Either Treatment Arm
(overall and by grade [safety population])
0.8
No. of Patients (%)
Overall Survival
(probability)
single agent in patients with recurrent and/or meta- neck: Rationale for future randomised trials in hu-
REFERENCES static squamous cell carcinoma of the head and man papilloma virus-negative disease. Eur J Cancer
neck who failed to respond to platinum-based ther- 49:1609-1618, 2013
1. Cooper JS, Pajak TF, Forastiere AA, et al: apy. J Clin Oncol 25:2171-2177, 2007 15. Ritchie JM, Smith EM, Summersgill KF, et al:
Postoperative concurrent radiotherapy and chemo- 9. Baselga J, Trigo JM, Bourhis J, et al: Phase II Human papillomavirus infection as a prognostic fac-
therapy for high-risk squamous-cell carcinoma of the multicenter study of the antiepidermal growth factor tor in carcinomas of the oral cavity and oropharynx.
head and neck. N Engl J Med 350:1937-1944, 2004 receptor monoclonal antibody cetuximab in combi- Int J Cancer 104:336-344, 2003
2. Pignon JP, le Maître A, Maillard E, et al: nation with platinum-based chemotherapy in pa- 16. Shuster JJ: Median follow-up in clinical trials.
Meta-analysis of chemotherapy in head and neck tients with platinum-refractory metastatic and/or J Clin Oncol 9:191-192, 1991
cancer (MACH-NC): An update on 93 randomised recurrent squamous cell carcinoma of the head and 17. Schemper M, Smith TL: A note on quantifying
trials and 17,346 patients. Radiother Oncol 92:4-14, neck. J Clin Oncol 23:5568-5577, 2005 follow-up in studies of failure time. Control Clin
2009 10. Mesía R, Henke M, Fortin A, et al: Chemora- Trials 17:343-346, 1996
3. Bernier J, Domenge C, Ozsahin M, et al: diotherapy with or without panitumumab in patients 18. de Souza JA, Davis DW, Zhang Y, et al: A
Postoperative irradiation with or without concomi- with unresected, locally advanced squamous-cell
phase II study of lapatinib in recurrent/metastatic
tant chemotherapy for locally advanced head and carcinoma of the head and neck (CONCERT-1): A
squamous cell carcinoma of the head and neck. Clin
neck cancer. N Engl J Med 350:1945-1952, 2004 randomised, controlled, open-label phase 2 trial.
Cancer Res 18:2336-2343, 2012
4. Cooper JS, Zhang Q, Pajak TF, et al: Long- Lancet Oncol 16:208-220, 2015
19. Ang KK, Zhang Q, Rosenthal DI, et al: Ran-
term follow-up of the RTOG 9501/intergroup phase 11. Gregoire V, Hamoir M, Chen C, et al: Gefitinib
domized phase III trial of concurrent accelerated
III trial: Postoperative concurrent radiation therapy plus cisplatin and radiotherapy in previously un-
radiation plus cisplatin with or without cetuximab for
and chemotherapy in high-risk squamous cell carci- treated head and neck squamous cell carcinoma:
stage III to IV head and neck carcinoma: RTOG
noma of the head and neck. Int J Radiat Oncol Biol A phase II, randomized, double-blind, placebo-
0522. J Clin Oncol 32:2940-2950, 2014
Phys 84:1198-1205, 2012 controlled study. Radiother Oncol 100:62-69, 2011
5. Cancer Genome Atlas Network: Comprehen- 12. Martins RG, Parvathaneni U, Bauman JE, et 20. Cameron D, Casey M, Oliva C, et al: Lapatinib
sive genomic characterization of head and neck al: Cisplatin and radiotherapy with or without erlo- plus capecitabine in women with HER-2-positive
squamous cell carcinomas. Nature 517:576-582, tinib in locally advanced squamous cell carcinoma of advanced breast cancer: Final survival analysis of a
2015 the head and neck: A randomized phase II trial. J Clin phase III randomized trial. Oncologist 15:924-934,
6. Kalyankrishna S, Grandis JR: Epidermal Oncol 31:1415-1421, 2013. 2010
growth factor receptor biology in head and neck 13. Harrington KJ, El-Hariry IA, Holford CS, et al: 21. Ravaud A, Hawkins R, Gardner JP, et al:
cancer. J Clin Oncol 24:2666-2672, 2006 Phase I study of lapatinib in combination with Lapatinib versus hormone therapy in patients with
7. Bonner JA, Harari PM, Giralt J, et al: Radio- chemoradiation in patients with locally advanced advanced renal cell carcinoma: A randomized phase
therapy plus cetuximab for squamous-cell carci- squamous cell carcinoma of the head and neck. J III clinical trial. J Clin Oncol 26:2285-2291, 2008
noma of the head and neck. N Engl J Med 354: Clin Oncol 27:1100-1107, 2009 22. Wülfing C, Machiels JP, Richel DJ, et al: A
567-578, 2006 14. Harrington K, Berrier A, Robinson M, et al: single-arm, multicenter, open-label phase 2 study of
8. Vermorken JB, Trigo J, Hitt R, et al: Open- Randomised phase II study of oral lapatinib com- lapatinib as the second-line treatment of patients
label, uncontrolled, multicenter phase II study to bined with chemoradiotherapy in patients with ad- with locally advanced or metastatic transitional cell
evaluate the efficacy and toxicity of cetuximab as a vanced squamous cell carcinoma of the head and carcinoma. Cancer 115:2881-2890, 2009
Affiliations
Kevin Harrington, Institute of Cancer Research and Royal Marsden Hospital, London; Hisham Mehanna, Institute of Head and Neck
Studies and Education, University of Birmingham, Birmingham; Natalie Franklin and John Farrell, GlaxoSmithKline, Uxbridge, United
Kingdom; Stephane Temam and Jean Bourhis, Institut Gustave- Roussy, Villejuif; Ida D’Onofrio, Hôpital Forcilles, Paris, France; Anil
D’Cruz, Tata Memorial Hospital, Mumbai; Minish Jain, Ruby Hall Clinic–Cancer Centre, Pune, India; Georgy Manikhas, St Petersburg City
Oncology Dispensary, St Petersburg, Russia; Zsuzsanna Horvath, Szent Imre University Teaching Hospital Budapest, Budapest, Hungary; Yan
Sun, Beijing Cancer Hospital, Beijing, China; Stefan Dietzsch, Hospital and Policlinic for Radiation Therapy and Radio-oncology, Leipzig,
Germany; Pavol Dubinsky, East Slovakia Cancer Institute, Kosice, Slovakia; Petra Holeckova, Institute of Radiation Oncology, Hospital Na
Bulovce, and 1st Medical Faculty of Charles University, Prague, Czech Republic; Iman El-Hariry, Synta Pharmaceutical, Lexington, MA; Paul
Wissel and Mayur Amonkar, Novartis Pharmaceuticals, East Hanover, NJ; Catherine Ellis and Sergio Santillana, GlaxoSmithKline,
Philadelphia, PA; and Nigel Biswas-Baldwin, Philippe Legenne, Thelma Netherway, Jing Wang-Silvanto, and Nazma Ahmed, Novartis
Pharma AG, Basel, Switzerland.
■ ■ ■
Postoperative Adjuvant Lapatinib and Concurrent Chemoradiotherapy Followed by Maintenance Lapatinib Monotherapy in High-Risk Patients
With Resected Squamous Cell Carcinoma of the Head and Neck: A Phase III, Randomized, Double-Blind, Placebo-Controlled Study
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are
self-held unless noted. I ⫽ Immediate Family Member, Inst ⫽ My Institution. Relationships may not relate to the subject matter of this manuscript. For more
information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.
Kevin Harrington No relationship to disclose
Honoraria: Merck Sharp & Dohme, Amgen, Oncos Therapeutics,
Cellgene Iman El-Hariry
Consulting or Advisory Role: Merck Sharp & Dohme, Amgen (Inst), No relationship to disclose
Viralytics Inc (Inst), Lytix Natalie Franklin
Speakers’ Bureau: Merck Sharp & Dohme, Amgen Employment: GlaxoSmithKline, Roche
Research Funding: Oncolytics Biotech (Inst), Genelux (Inst), Viralytics Stock or Other Ownership: GlaxoSmithKline
(Inst), AstraZeneca (Inst)
Nigel Biswas-Baldwin
Stephane Temam Employment: GlaxoSmithKline
No relationship to disclose Stock or Other Ownership: GlaxoSmithKline
Hisham Mehanna Philippe Legenne
Employment: Warwickshire Head Neck Clinic Employment: GlaxoSmithKline
Leadership: Warwickshire Head neck Clinic, Warwickshire Head Neck
Stock or Other Ownership: GlaxoSmithKline
Clinic (I)
Stock or Other Ownership: Warwickshire Head Neck Clinic Paul Wissel
Honoraria: AstraZeneca Employment: GlaxoSmithKline
Speakers’ Bureau: Merck Sharp & Dohme, Sanofi Pasteur, Merck Stock or Other Ownership: GlaxoSmithKline
Research Funding: GlaxoSmithKline (Inst), Merck Sharp & Dohme
(Inst), Sanofi Pasteur (Inst), Silence Therapeutics (Inst), Thelma Netherway
GlaxoSmithKline Biologicals (Inst) Employment: GlaxoSmithKline, Novartis
Travel, Accommodations, Expenses: Sanofi Pasteur, Merck Sharp & Stock or Other Ownership: GlaxoSmithKline, Novartis
Dohme, Merck Travel, Accommodations, Expenses: GlaxoSmithKline, Novartis
Acknowledgment
We thank all patients and their families, investigators, and study staff; the independent data monitoring committee; and the BioClinica
(independent review committee), EqualEstro (Radiotherapy Quality Assurance), and PAREXEL (independent statistics) teams. The authors also
thank Jen Carver, Jhangir Irani, Sejal Patel, Zuheb Ali, and all members of the GlaxoSmithKline central and monitoring team. Lapatinib is an asset
of Novartis AG as of March 2, 2015. Editorial support was provided by Karen Yee, PhD, of Fishawack Indicia, funded by GlaxoSmithKline and
Novartis Pharmaceuticals Corporation.
Appendix
Methods
Radiation treatment. A detailed description of radiation target volumes and dose prescription is provided in the Radiation Therapy
Protocol Guidelines (Data Supplement). Patients in both arms of this study received a conventionally fractionated course of radiotherapy.
Irradiation was planned using either a standard two-dimensional technique or a three-dimensional conformal technique. The use of
intensity-modulated radiation therapy was not allowed. The selection of technique was left to the discretion of each center. Only linear
accelerator– based treatments (photon energy ⱖ 4 MV or electrons) were allowed.
Exclusion criteria. Patients were excluded if they had tumors of the nasopharynx, paranasal sinuses, or nasal cavity or tumors with
histology other than squamous cell carcinomas; evidence of distant metastases, gross postoperative residual disease, or second primary
tumor; any prior or concurrent anticancer treatment or concurrent treatment with an investigational agent; or history of another
malignancy within the last 5 years (excluding resected basal or squamous skin cancers).
Health-related quality of life (HRQoL) assessment. Functional Assessment of Cancer Therapy–Head and Neck (FACT-H&N,
comprising Functional Assessment of Cancer Therapy–General [FACT-G] and a head and neck subscale) and EQ-5D questionnaires
were completed at day 1 (before dose), at the end of chemotherapy, every 8 weeks during the first year of follow-up, and at discontinuation
of randomly assigned therapy. During the next 2 years of follow-up, patients were asked to complete the questionnaire every 4 months.
A minimal clinically important difference (ie, a clinically meaningful change) is defined as ⫹6 (improvement) and ⫺12 (worsening)
for the FACT-H&N total score, ⫹4 (improvement) and ⫺8 (worsening) for the FACT-G score, and ⫹3 (improvement) and ⫺3
(worsening) for the head and neck subscale score and Trial Outcome Index (TOI; Ringash J, et al: Qual Life Res 13:725-733, 2004); 0.08
for the EQ-5D utility score; and 7 for the EQ-5D thermometer score (Pickard AS, et al: Health Qual Life Outcomes 5:70, 2007).
Tissue samples. Tissue sections (4 m) and/or tissue blocks of the formalin-fixed and paraffin-embedded (FFPE) diagnostic biopsies
or resection specimens were collected. For the analysis, where only tissue sections were available, these were used. Where tissue blocks were
available, tissue microarrays (TMAs) were constructed. Briefly, hematoxylin and eosin–stained sections were prepared from the blocks to
confirm the presence of tumor, and the slides were annotated by a pathologist to guide TMA core selection. Up to four 0.6-mm diameter
tumor cores per block were transferred from the donor block to the recipient TMA block. A hematoxylin and eosin–stained section of the
prepared TMA was examined to assess adequate tumor sampling. Thick tissue curls (10 m) were taken from available blocks for DNA
extraction and downstream polymerase chain reaction (PCR) analysis.
HPV analysis protocol. All samples were analyzed for p16 expression by immunohistochemistry (IHC). Samples were assessed for
high-risk (HR) HPV DNA using a protocol based on that described and validated by Smeets et al (Int J Cancer 121:2465-2472, 2007).
Samples were assessed by HPV in situ hybridization (ISH). For samples that were found to be p16 positive/HPV negative by ISH using
TMAs, whole tissue sections were tested to reduce misclassification as a consequence of possible sampling limitations inherent to TMA
analysis. If the sample remained p16 positive/HPV negative by ISH on the whole tissue section, the sample was then examined by
consensus HR-HPV PCR to control for the suboptimal sensitivity of HPV ISH compared with target amplification techniques (Smeets SJ,
et al: Int J Cancer 121:2465-2472, 2007). This latter protocol has been previously validated and proven to have the same accuracy as using
p16 and HPV-PCR (Thavaraj S, et al: J Clin Pathol 64:308-312, 2011). For all samples in this study, samples that showed evidence of HPV
DNA (either by ISH or consensus PCR) and showed high p16 expression (HPV positive/p16 positive) were defined as harboring
biologically relevant oncogenic HPV infection (Smeets SJ, et al: Int J Cancer 121:2465-2472, 2007).
HPV and p16 testing methods. p16 IHC was carried out using a proprietary kit (CINtec Histology; MTM Laboratories AG,
Heidelberg, Germany) on a Benchmark Autostainer (Ventana Medical Systems, Tucson, AZ). Tissue from an oropharyngeal squamous
cell carcinoma with high p16 expression was used as a positive control. The primary antibody was omitted from negative controls. A binary
(positive and negative) scoring system was adopted, and p16 IHC was scored as positive if there was strong and diffuse nuclear and
cytoplasmic staining present in more than 70% of the malignant cells (Singhi AD, et al: Cancer 116:2166-2173, 2010). All other staining
patterns were scored as negative.
HR-HPV ISH was carried out using proprietary reagents (Inform HPV III Family 16 Probe [B]; Ventana) on a Benchmark
Autostainer (Ventana). The Inform HPV III Family 16 Probe (B) detects HR genotypes 16, 18, 31, 33, 35, 39, 51, 52, 56, 58, and 66. The
following three control samples were used: FFPE CaSki cells (HPV-16 positive; 600 copies per cell), HeLa cells (HPV-18 positive; 10 to 50
copies per cell), and C-33A (HPV-negative cell line; Ventana). A binary scoring system was used. The HR-HPV ISH test was scored as
positive if there was a blue reaction product that colocalized with the nuclei of malignant cells. Diffuse staining of tumor and stromal
tissues, considered to represent nonspecific chromogen precipitate, was scored as negative. Pale staining limited to the nucleoli of cells and
staining of occasional leukocytes and stromal cells were also disregarded, in line with the manufacturer’s instructions (Ventana).
For consensus PCR for oncogenic HPV, DNA was extracted from tissue curls using the Qiagen (Hilden, Germany) FFPE kit (without
xylene washes). The yield of DNA was quantified using nanodrop spectrophotometry. DNA adequacy was assessed by amplification of the
human -globin gene. HPV typing was performed using the GP5⫹/6⫹ primer set (Jacobs MV, et al: Clin Microbiol 35:791-795, 1997).
Samples were tested by PCR enzyme immunoassay using a cocktail of probes for 14 HR-HPV types and separately with a cocktail of probes
for six low-risk types. Samples testing positive for ab HR-HPV type were subject to second PCR and assayed with each of the 14 HR probes
independently of each other. A sample was defined as HR-HPV DNA positive if it tested positive in the HR-HPV cocktail reaction and an
HR-HPV genotype was identified in the second PCR reaction.
Scoring methods. The criteria used to score the p16 expression and HR-HPV ISH were defined before commencing the study. Two
pathologists analyzed the test results independently. Before commencing the study, a training set of 45 tonsil squamous cell carcinomas
was used to calibrate the two scorers, obtaining greater than 95% concordance (data not shown). The p16 IHC and HR-HPV ISH tests
were assessed separately alongside corresponding hematoxylin and eosin–stained sections, and the scores were collated by the study
coordinator. Any discordant scores (5%) between the pathologists were resolved at a meeting between the pathologists to establish a
consensus. In addition, all samples scored as p16 negative/HPV positive and p16 positive/HPV negative were reviewed by the two
pathologists together to ensure correct categorization.
Hypotheses. The study was designed to provide evidence to either support the null hypothesis (H0: ⱖ 1) or reject it in favor of the
alternative hypothesis (HA: ⬍ 1), where is the hazard ratio (HR) for disease-free survival (DFS; lapatinib/placebo). Assuming the DFS
curves are consistent with proportional hazards, then the null hypothesis represents equality of the median DFS in the two treatment arms,
or a lower median DFS in the lapatinib arm, and the alternative hypothesis represents a higher median DFS in the lapatinib arm.
Sample size assumptions. The following assumptions were made in the estimation of the required sample size: exponential distribu-
tions of DFS; an HR of 0.7206, which originates from estimated DFS rates at 2 years after random assignment of 65% in the lapatinib arm
and 55% in the placebo arm (an absolute increase of 10%), resulting in an approximate median DFS of 39 and 28 months in the lapatinib
arm and placebo arm, respectively; a 1:1 random assignment scheme, with a total of approximately 680 patients, 340 patients per arm (on
the basis of observed cure rates in the studies reported by Bernier et al3 and Cooper et al,4 it was anticipated that a portion of patients
entered onto the study may never reach an end point of recurrence of disease; therefore, more patients were recruited to allow for this); a
2.5% risk of erroneously claiming superiority of lapatinib in the presence of no true underlying difference (one-sided type I error); an 80%
chance of successfully detecting am HR of 0.7206, which equates to detecting a 39% increase in median DFS in patients who receive
lapatinib (39 months; HR, 0.0179) relative to placebo (28 months; HR, 0.0249); and an average constant accrual rate of 15 patients per
month for 45 months.
Efficacy end points. Greenwood’s formula was used to calculate the SE of the Kaplan-Meier curve estimates. The Pike estimator of the
HR and associated 95% CIs were also presented. DFS was compared between treatment arms using a nonstratified log-rank test. Cox
regression models were used to explore the effect of prognostic factors, including stratification factors and HPV status. The adjusted HR,
95% CI, and two-sided Wald 2 P values were presented.
HRQoL. For HRQoL analyses, data collected from FACT-H&N and EQ-5D questionnaires were summarized descriptively and
listed to show FACT-H&N responses for subscale, total, FACT-G, and TOI scores, and EQ-5D domain scores, thermometer values, and
derived utility scores. Changes from baseline were also summarized for each treatment arm with an analysis of covariance performed to
compare differences between study arms in FACT-G scores, TOI scores, EQ-5D utility scores, and thermometer scores.
Quality-of-life responder analyses were also performed to indicate scores that were clinically meaningful. Patients with missing
baseline assessments for FACT-H&N were excluded from analysis.
Results
Secondary end point: HRQoL. For HRQoL, at the end of chemoradiotherapy, clinically meaningful decreases from baseline were
observed in both arms in all HRQoL measures, as follows: FACT-H&N total (lapatinib, ⫺17.9; placebo, ⫺13.5), FACT-G (lapatinib,
⫺11.1; placebo, ⫺7.8), TOI (lapatinib, ⫺16.0; placebo, ⫺12.2), head and neck subscale (lapatinib, ⫺7.0; placebo, ⫺5.6), EQ-5D utility
(lapatinib, ⫺0.14; placebo, ⫺0.12), and EQ-5D thermometer (lapatinib, ⫺8.4; placebo, ⫺7.7) scores. All of these declines (except both
scores of EQ-5D) were found to be significantly worse (P ⬍ .05) for the patients on lapatinib than those on placebo. Results for FACT-G
and EQ-5D utility are illustrated in Appendix Figures A1 and A2 (showing changes in scores from baseline across all the treatment phases
in the study).
During the maintenance period, no clinically meaningful decreases from baseline were observed in both arms of all HRQoL measures
except TOI scores at maintenance week 8 (lapatinib, ⫺8.5; placebo, ⫺5.0). However, patients receiving lapatinib had significantly less
favorable changes (P ⬍ .05) than patients given placebo in FACT-H&N total score, FACT-G score (Fig A2), and TOI at multiple time
points throughout the maintenance period.
On treatment withdrawal, both groups displayed clinically meaningful declines in FACT-H&N total (lapatinib, ⫺14.9; placebo,
⫺16.2), FACT-G (lapatinib, ⫺10.1; placebo, ⫺12.2; Fig A1), TOI (lapatinib, ⫺11.9; placebo, ⫺12.5), head and neck subscale (lapatinib,
⫺4.9; placebo, ⫺4.1), EQ-5D utility (lapatinib, ⫺0.11; placebo, ⫺0.23), and EQ-5D thermometer scores (lapatinib, ⫺3.3; placebo,
⫺11.4). However, the differences between treatment arms was not statistically significant (P ⬎ .05) except for EQ-5D, with both scores
significantly in favor of lapatinib compared with placebo (P ⬍ .05; Appendix Fig 3A).
Overall responder analyses showed that HRQoL worsened in more lapatinib-treated patients than patients taking placebo (P ⬍ .05)
according to FACT-H&N total scores and FACT-G scores, but the differences were not significant in TOI and head and neck subscale
scores (P ⬎ .05; Appendix Table A6).
EGF102988 Study Sites and Investigators
Sites in order of patient recruitment numbers were as follows: India (n ⫽ 119), France (n ⫽ 116), China (n ⫽ 80), Germany (n ⫽ 51),
Hungary (n ⫽ 41), the Russian Federation (n ⫽ 40), Spain (n ⫽ 36), the Czech Republic (n ⫽ 29), Slovakia (n ⫽ 24), Thailand (n ⫽ 22),
United Kingdom (n ⫽ 21), Greece (n ⫽ 21), Croatia (n ⫽ 18), Italy (n ⫽ 17), Austria (n ⫽ 15), the Philippines (n ⫽ 14), Argentina (n ⫽
9), Hong Kong (n ⫽ 8), United States of America (n ⫽ 3), Canada (n ⫽ 3), and Estonia (n ⫽ 1).
Argentina (n ⫽ 3): Blajman, Cesar R., MD; Campos, Claudio A., MD; Fein, Luis E., MD.
Austria (n ⫽ 3): Greil, Richard, MD; Kornek, Gabriela, MD; Scholtz, Arne-Wulf, MD.
Canada (n ⫽ 2): Jha, Naresh, MD, FRCPC, MBBS; Nabid, Abdenour, MD.
China (n ⫽ 7): Gao, Li, MD; Hu, Chaosu, MD; Hu, Guoqing, MD; Pan, Jianji, PhD; Sun, Yan, MD; Wu, Shao Xiong, MD; Yuan,
Zhiyong, MD.
Croatia (n ⫽ 2): Prgomet, Drago, MD; Virag, Mîso Mihajlo, MD.
Czech Republic (n ⫽ 4): Cervena, Renata, MD; Cincibuch, Jan, MD; Feltl, David, MD, PhD; Holeckova, Petra, MD.
Estonia (n ⫽ 1): Kuddu, Maire, MD.
France (n ⫽ 15): Babin, Emmanuel, MD; Banal, Alain, MD; Borel, Christian, MD; Cupissol, Didier, MD, PhD; Degardin, Marian,
MD; Delalande, Cecile, MD; Delord, Jean-Pierre, MD; D’Onofrio, Ida, MD; Duparc, Angelique, MD; Jadaud, Eric, MD; Lacau Saint
Guily, Jean, MD; Prevost, Alain, MD; Racadot, Séverine, MD; Temam, Stephane, MD; Zawadi, Ayman, MD.
Germany (n ⫽ 6): Classen, Johannes, Dr Med; Debus, Juergen P, Dr Med; Kortmann, Rolf-Dieter, Dr Med; Lang, Stephan, MD, PhD;
Schreiber, Andreas, Dr Med, MRCP; Vordermark, Dirk, Dr Med, PhD.
Greece (n ⫽ 4): Antonadou, Theodosia, MD; Maroudias, Nikolaos, MD; Psyrri, Amanda, MD; Skarlos, Dimosthenis, MD, PhD.
Hong Kong (n ⫽ 2): Au, Kwok Hung, MBBS; Chan, Yu Wai, Jimmy, MBBS.
Hungary (n ⫽ 3): Csejtei, András, MD; Horvai, Géza, MD; Remenár, Éva, MD.
India (n ⫽ 5): Almel, Sachin, MD; D’Cruz, Anil K., MBBS, MS; Iyer, Subramania, MS, MCh, FRCS; Jain, Minish M, MD; Kumar, R.
Rejnish, MD.
Italy (n ⫽ 5): Bacigalupo, Almalina, MD; Bianco, Roberto, MD; Caponigro, Francesco, MD; Nolè, Franco, MD; Villa, Eugenio, MD.
Philippines (n ⫽ 1): Pontejos, Alfredo Q., Jr, MD.
Russia (n ⫽ 3): Biakhov, Mikhail, MD, DSc; Manikhas, Georgy M., MD, PhD; Romanov, Ilya S., MD, PhD.
Slovakia (n ⫽ 1): Dubinsky, Pavol, MD.
Spain (n ⫽ 11): Ales, Jose E., MD; Ballester Navarro, Inmaculada, MD; Del Campo, Jose Maria, MD; Espinosa Arranz, Enrique, MD;
García Sáenz, Jose Angel, MD; Iglesias Docampo, Lara; Lavernia, Javier, MD; López, Rafael, MD; Martínez Galán, Joaquina, MD; Morales
Murillo, Serafin, MD; Rubio, Jordi, MD.
Thailand (n ⫽ 2): Chitapanarux, Imjai, MD; Eursritanakorn, Sirima.
United Kingdom (n ⫽ 7): Goodchild, Kathleen, BSc (Hons), MBBS, MRCP, FRCR; Guerrero Urbano, Maria Teresa, PhD, FRCR;
Harrington, Kevin J., BSc, MBBS, MRCP (United Kingdom); Junor, Elizabeth, MRCP, MD, FRCR; Kelly, Charles G., MB ChB, MSc,
FRCP; Robinson, Martin H., MD, MRCP; Simcock, Richard, MBBS, MRCP, FRCR.
United States (n ⫽ 2): Cordero, Joehassin, MD; Reimers, Hans J. Joachim, MD, PhD.
Table A1. Derivation of Overall HPV Status Based on Separate HPV/p16 Results
p16
Result HPV ISH Result HPV PCR Result Overall Status for Analysis
Positive Positive Not tested Positive
Positive Negative Positive Positive
Positive Negative Negative Negative
Positive Negative Not tested/unknown Unknown
Positive Unknown Not tested Unknown
Negative Negative Not tested Negative
Negative Positive Not tested Negative
Negative Unknown Not tested Negative
Unknown Any Not tested Unknown
Abbreviations: HPV, human papillomavirus; ISH, in situ hybridization; PCR, polymerase chain reaction.
Abbreviations: HR, hazard ratio; NR, not reached; OR, odds ratio.
Table A3. Unscheduled Medical Resource Use: Time Points at Which the Highest Proportion of Patients Reported Use, by Item, Through the Maintenance
Period and the Withdrawal From Treatment Visit
Lapatinib (n ⫽ 346) Placebo (n ⫽ 342)
Category and Type of Resource No. (%) Reported at No. (%) Reported at
Medical office visits
Primary care physician visits 22 (6) MW8 22 (6) MW8
Medical specialist visit 38 (11) MW16 39 (11) MW8
Nurse practitioner, physician’s assistant, or nurse visits 3 (1) End of CRT, MW8 4 (1) End of CRT, MW8, MW16
Telephone calls 3 (1) MW8 4 (1) End of CRT, MW8
Laboratory tests/radiology procedures
Nonstudy laboratory visits 11 (3) End of CRT, MW8 13 (4) MW8
Nonstudy radiology visits 10 (3) End of CRT 11 (3) MW16
Other nonstudy visits 6 (2) MW8 11 (3) WD
Home health care: home health care visits 2 (1) MW8 2 (1) WD
Hospital extended care
Days in general ward 19 (5) End of CRT 12 (4) End of CRT
Days in intensive care 2 (1) WD 2 (1) MW16, WD
Days in extended care facility 1 (⬍ 1) MW56, WD 1 (⬍ 1) WD
Urgent care: emergency room visits 5 (1) End of CRT 7 (2) End of CRT, MW8
% Change
Scale and Type
of Change Lapatinib (n ⫽ 307) Placebo (n ⫽ 306) P
FACT H&N subscale .082
Improvement 19 25
Stable 4 5
Worsening 78 70
FACT TOI .098
Improvement 18 22
Stable ⬍1 2
Worsening 82 76
FACT-G .038
Improvement 30 39
Stable 3 4
Worsening 66 57
FACT-H&N total .007
Improvement 24 33
Stable 4 7
Worsening 72 60
NOTE. Overall responder status determined at the first visit that showed a clinically relevant improvement or worsening. Percentages may not add to 100% as a
result of rounding.
Abbreviations: FACT, Functional Assessment of Cancer Therapy; FACT-G, Functional Assessment of Cancer Therapy–General; FACT-H&N, Functional Assessment
of Cancer Therapy–Head and Neck; TOI, Trial Outcome Index.
1.0 HPV−/placebo
HPV−/lapatinib
Disease-Free Survival
HPV+/placebo
0.8 HPV+/lapatinib
(probability)
0.6
0.4
HPV/p16− HPV/p16+
0.2 Placebo arm Lapatinib arm Placebo arm Lapatinib arm
n 284 276 21 23
Events: disease recurrence 93 (33) 102 (37) 3 (14) 5 (22)
or death, n (%)
0 Median DFS (95% CI) NR (54.6 to NR) 51.5 (42.6 to 61.2) NR (NR to NR) NR (NR to NR)
Hazard ratio (95% CI) 1.16 (0.88 to 1.54) 1.32 (0.33 to 5.33)
Two-sided P .29 .70
0 10 20 30 40 50 60 70
Fig A1. Kaplan-Meier estimates of disease-free survival (DFS; intent-to-treat population) independently assessed by human papillomavirus (HPV) status (HPV and
p16). NR, not reached.
py -
en al
ra mo
lin
tm w
Maintenance Weeks
*
t
ea ra
se
he e
ot ch
tr hd
Ba
di f
it
ra d o
om W
En
fr
Scheduled Assessments
Fig A2. Adjusted mean change from baseline in the Functional Assessment of Cancer Therapy–General score. (*) Denotes a statistically significant (P ⬍ .05)
difference versus placebo. Estimates that fall outside of the shaded region can be interpreted as a clinically meaningful change based on Ringash et al (Qual Life Res
13:725-733, 2004).
0.08
Adjusted Mean Change
From Baseline (95% CI)
−0.08
−0.16
−0.24 Lapatinib
Placebo
−0.32
8 16* 24 32 40 48 56
e
py -
en al
ra mo
lin
tm w
Maintenance Weeks
*
t
ea ra
se
he e
ot ch
tr hd
Ba
di f
it
ra o
om W
d
En
fr
Scheduled Assessments
Fig A3. Adjusted mean change from baseline in the EQ-5D utility index score. (*) Denotes a statistically significant (P ⬍ .05) difference versus placebo. Estimates
that fall outside of the shaded region can be interpreted as a clinically meaningful change based on Pickard et al (Health Qual Life Outcomes 5:70, 2007).