Annals of Oncology 0: 1–7, 2017

doi:10.1093/annonc/mdx299
Published online 4 July 2017

ORIGINAL ARTICLE

Induction TPF followed by concomitant treatment

versus concomitant treatment alone in locally
advanced head and neck cancer. A phase II–III trial

M. G. Ghi1*, A. Paccagnella1, D. Ferrari2, P. Foa2, D. Alterio3, C. Codeca2, F. Nolè4, E. Verri4, R. Orecchia3,

F. Morelli5, S. Parisi6, C. Mastromauro1, C. A. Mione7, C. Rossetto8, M. Polsinelli9, H. Koussis10,
L. Loreggian11, A. Bonetti12, F. Campostrini13, G. Azzarello14, C. D’Ambrosio15, F. Bertoni16, C. Casanova17,
E. Emiliani18, M. Guaraldi19, F. Bunkheila20, P. Bidoli21, R. M. Niespolo22, A. Gava23, E. Massa24,
A. Frattegiani25, F. Valduga26, G. Pieri27, T. Cipani28, D. Da Corte29, F. Chiappa30 & E. Rulli30, for the GSTTC
(Gruppo di Studio Tumori della Testa e del Collo) Italian Study Group
1
Medical Oncology Department, Ospedale dell’Angelo, Venezia; 2Medical Oncology Unit, Ospedale San Paolo, Milano; 3Advanced Radiotherapy Center; 4Unit of
Urogenital and Head and Neck Oncology, Istituto Europeo di Oncologia, Milano; 5U.O.C. Medical Oncology; 6U.O.C. Radiation Oncology, IRCCS Casa Sollievo della
Sofferenza, San Giovanni, Rotondo; 7Radiotherapy Department, Ospedale SS Giovanni e Paolo, Venezia; 8Medical Oncology Department; 9S.O.C. Radiation
Oncology, Azienda Ospedaliero-Universitaria S.Maria della Misericordia, Udine; 10Medical Oncology Department 2, Istituto Oncologico Veneto- IRCCS, Padova;
11
Radiotherapy Department, Istituto Oncologico Veneto – IRCCS, Padova; 12Medical Oncology Department, Ospedale Mater Salutis, Legnago; 13Radiotherapy
Department, Ospedale Mater Salutis, Legnago; 14Oncology Unit, Department of Internal Medical Sciences, Mirano; 15Medical Oncology Department;
16
Radiotherapy Department, Azienda Ospedaliero Universitaria, Modena; 17Medical Oncology Department; 18Radiotherapy Department, Azienda USL, Ravenna;
19
Medical Oncology Department, Policlinico Sant’Orsola-Malpighi, Bologna; 20Radiotherapy Department, Policlinico Sant’Orsola-Malpighi, Bologna; 21Medical
Oncology Department, Ospedale San Gerardo, Monza; 22Radiotherapy Department, Ospedale San Gerardo, Monza, Ospedale San Gerardo, Monza; 23Radiotherapy
Department, Ospedale Ca’ Foncello, Treviso; 24Department of Medical Science, Universita degli Studi di Cagliari, Cagliari; 25Radiation Oncology Department,
Ospedale S. Maria della Misericordia, Perugia; 26Medical Oncology Department, Ospedale S. Chiara, Trento; 27Medical Oncology Department, AO Triestina, Trieste;
28
Niguarda Cancer Center, Ospedale Niguarda Ca Granda, Milano; 29Oncology Department, Ospedale S. Martino, Belluno; 30Laboratory of Clinical Research, Istituto
di Ricerche Farmacologiche “Mario Negri,” Milano, Italy

*Correspondence to: Dr Maria Grazia Ghi, Medical Oncology Department, Ospedale dell’Angelo, Via Paccagnella 11, Venezia-Mestre 30174, Italy. Tel: þ39-041-9656320;
Fax: þ39-041-9656326; E-mail: [email protected]
Note: This study was previously presented in part at the 49th and 50th Annual Meeting of the American Society of Clinical Oncology, 2013 (preliminary toxicity data) and
2014 (efficacy results) and at the ICHNO 2015 meeting.

Background: Platinum-based chemoradiation (CCRT) is the standard treatment for Locally Advanced Head and Neck
Squamous-Cell Carcinoma (LAHNSCC). Cetuximab/RT (CET/RT) is an alternative treatment option to CCRT. The efficacy of
induction chemotherapy (IC) followed by chemoradiation compared to chemoradiation alone has not been demonstrated in
randomized clinical trials. The goals of this phase II-III trial were to assess: (i) the overall survival (OS) of IC versus no-induction
(no-IC) and (ii) the Grade 3–4 in-field mucosal toxicity of CCRT versus CET/RT. The present paper focuses on the analysis of
efficacy.
Materials and methods: Patients with LAHNSCC were randomized to receive concomitant treatment alone [CCRT (Arm A1) or
CET/RT (Arm A2)], or three cycles of induction docetaxel/cisplatin/5 fluorouracil (TPF) followed by CCRT (Arm B1) or followed by
CET/RT (Arm B2). The superiority hypothesis of OS comparison of IC versus no-IC (Arms B1 þ B2 versus A1 þ A2) required 204
deaths to detect an absolute 3-year OS difference of 12% (HR 0.675, with 80% power at two-sided 5% significance level).
Results: 414 out of 421 patients were finally analyzed: 206 in the IC and 208 in the no-IC arm. Six patients were excluded
because of major violation and one because of metastatic disease at diagnosis. With a median follow-up of 44.8 months, OS was
significantly higher in the IC arm (HR 0.74; 95% CI 0.56–0.97; P ¼ 0.031). Complete Responses (P ¼ 0.0028), Progression Free
Survival (P ¼ 0.013) and the Loco-regional Control (P ¼ 0.036) were also significantly higher in the IC arm. Compliance to
concomitant treatments was not affected by induction TPF.

C The Author 2017. Published by Oxford University Press on behalf of the European Society for Medical Oncology.
V
All rights reserved. For Permissions, please email: [email protected].
Original article Annals of Oncology
Conclusions: IC followed by concomitant treatment improved the outcome of patients with LAHNSCC without compromising
compliance to the concomitant treatments. The degree of the benefit of IC could be different according to the type of the
subsequent concomitant strategy.
Clinical Trial Number: NCT01086826, www.clinicaltrials.gov
Key words: head and neck cancer, concomitant chemoradiotherapy, induction chemotherapy

Introduction Eligible patients aged 18 years had to have a histological/cytological

previously untreated stage III–IV LAHNSCC of the oral cavity, orophar-
Squamous-cell carcinoma of the head and neck represents 5% of ynx, hypopharynx, ECOG Performance Status of 0–1, life expect-
newly diagnosed cancers in adults and more than 550 000 new ancy 6 months, adequate bone marrow and organ function; no
cases are predicted annually worldwide [1]. Most patients present peripheral neuropathy, altered hearing G2, or weight loss >20% in the
loco-regionally advanced disease (LAHNSCC) not suitable for previous 3 months.
radical resection or low surgical curability. Phase III studies and Patients were deemed unsuitable for radical surgery after evaluation of
a multidisciplinary team. Inoperability criteria were technical reasons
individual patient data Meta-Analysis of Chemotherapy in Head
(tumor fixation/invasion to either base of the skull, cervical vertebrae,
and Neck Cancer (MACH-NC) [2, 3] have shown that concomi- nasopharynx, or fixed lymph nodes) or low surgical curability (T3–4,
tant platinum-based chemoradiotherapy (CCRT) is the optimal N2–3 excluding T1N2). Patients unresectable due to medical conditions
treatment for LAHNSCC. were not eligible.
In 2006, a phase III trial comparing radiotherapy (RT) alone RT was delivered according to the Italian national guidelines on
versus cetuximab/RT (CET/RT) showed that CET/RT also pro- ‘Quality guarantee in Radiotherapy’ (Istituto Superiore di Sanita, rap-
vides superior overall survival (OS) and loco-regional control porti ISTISAN 02/20, in Italian).
(LRC) [4, 5]. The study was conducted in accordance with the Declaration of
Helsinki and the protocol was approved by Institutional Ethic
Induction chemotherapy (IC) has a proven role in organ pres-
Committees . All patients provided written informed consent.
ervation and in reducing distant failure [2, 3, 6], however, its abil-
ity to prolong OS has not been demonstrated. The MACH-NC
showed that IC cisplatin/5-fluorouracil (PF) followed by local
Treatment plan
treatment was associated with a small but significant improve- Patients were randomized to receive concomitant treatment alone
ment in OS [hazard ratio (HR) 0.90; 95% CI 0.82–0.99] and in [CCRT (arm A1) or CET/RT (arm A2)] or three cycles of IC followed by
CCRT (arm B1) or followed by CET/RT (arm B2). The CET/RT arms
distant failures (HR 0.63; 95% CI 0.45–0.89) [2, 3]. In a recent
were numerically unbalanced in accordance with the revised trial design.
MACH-NC update, induction PF plus a taxane [6] increased pro-
Standard-fractionated RT (2 Gy/day, 5 days/week) of 70 Gy was de-
gression free survival (PFS) (HR 0.78; 95% CI, 0.69–0.87) and OS livered to the primary tumor and 60 Gy to the neck depending on
versus PF (HR 0.72; 95% CI 0.63–0.83). whether a neck dissection was indicated/planned. 3D conformal radio-
Our trial was designed to assess the OS of concomitant treat- therapy or IMRT were allowed.
ment with or without IC and the in-field mucosal toxicity of The CCRT arm consisted of two cycles of cisplatin 20 mg/m2 from
CCRT versus CET/RT. days 1 to 4 plus 5-fluorouracil 800 mg/m2/day, 96-h continuous infusion,
This report is focused on the comparison of IC versus no-IC. administered during weeks 1 and 6 of the radiation treatment.
In the CET/RT arm, cetuximab was initiated 1 week before RT at
400 mg/m2 loading dose and followed by 250 mg/m2 weekly, for 7 weeks.
In the IC arms, three cycles of docetaxel 75 mg/m2 day 1, cisplatin
Patients and methods 80 mg/m2 day 1 and 5-fluorouracil 800 mg/m2/day 96-h continuous infu-
sion (TPF) were administered every 3 weeks with antibiotic prophylaxis
Study design after each cycle. Within 3–5 weeks after IC, patients received CCRT or
CET/RT provided there was no disease progression.
This is a multicenter, open-label, randomized phase II-III trial. In the Prophylactic neck dissection was planned for stage N2–N3 patients achiev-
first part [phase II randomized part (XRP6976-F/2501)] patients were ing pathological CR at the primary site and radiological CR at the neck.
randomized to receive CCRT or IC followed by the same CCRT. Since Salvage surgery was considered for residual disease after treatment.
the target difference of 15% in centrally reviewed radiological complete
responses (CR) in favor of IC had been achieved [7], the trial continued
as a phase III (NCT01086826), as pre-planned. Since CET/RT combin-
End points
ation had become an alternative treatment option to CCRT [4, 5] OS of no-IC versus IC (from the date of randomization until death from
within the study design timeframe, a 22 randomization was intro- any cause) was the primary efficacy end point. Survival was also expressed
duced for the phase III part of the trial according to the reciprocal con- as the proportion of patients alive after 3 years.
trol design [8]. Secondary efficacy endpoints included: response rate (RR) 6–8 weeks
Single-step double randomization, was centrally managed, computer- after concomitant treatment and 3–4 weeks after IC, according to the
generated, with a permuted-block scheme and a 1:1 ratio, stratified by RECIST criteria; LRC defined as the absence of loco-regional progression
T (T1–2 versus T3–4), N (N0–1 versus N2–3), and primary site (oral cav- or death due to loco-regional disease (deaths due to primary cancer with-
ity/oropharynx versus hypopharynx). Once assigned to the IC or no-IC out documented recurrence and/or deaths from unknown causes were
arm, the randomization to CCRT or CET/RT was stratified by IC attributed to loco-regional disease) [9]; PFS was calculated from ran-
assignment. domization to progression, second primary or death from any cause.
An Independent Data Monitoring Committee (IDMC) reviewed ac- Toxicity was graded using the Common Toxicity Criteria for Adverse
crual, safety, and aggregate data on efficacy. Events (version 2.0).

2 | Ghi et al.
Annals of Oncology Original article
Statistical methods Results
The trial was initially designed to have a power of 0.85 to detect an absolute dif-
ference of 15% in 3-year OS (from 45% to 60%), corresponding to a HR of Patient characteristics
0.64, a ¼ 0.05, two-sided. At the planned interim analysis (June 2010), the
IDMC suggested increasing the sample size according to the good steady-level 421 patients were randomized in 48 centers. 101 entered the
of recruitment and to the preliminary safety data. A more conservative HR of phase II part of the trial between January 2003 and January 2006
0.675 (12% difference in 3-year OS, from 52.5% to 64.5%), was therefore while the other 321 were randomized between March 2008 and
adopted (power 0.80, a ¼ 0.05, two-sided). Consequently, 420 patients (in-
April 2012. Final analysis was conducted on 414 patients. Seven
stead of 350) had to be randomized to observe at least 204 events. The protocol
was amended accordingly (July 2010). patients were not included: one due to metastatic disease and six
Kaplan–Meier curves for survival with the Mantel-Cox version of the as outcome information were not submitted.
log-rank test were compared. Cox proportional hazards model was used 206 patients were randomized to the no-IC arm and 208 to the
to estimate treatment effect adjusted for prognostic factors [age, T stage, IC arm (Figure 1).
N stage, ECOG-PS, site, phase of recruitment (II versus III), type of con- Baseline patient characteristics were well balanced (Table 1).
comitant treatment (CCRT versus CET/RT)]. In the analysis of LRC,
deaths not caused by loco-regional disease and distant relapses were re-
garded as competing events [10]. Safety
The differences in relative effect size according to the type of concomi-
tant treatment were described by forest plots of HR and relative confi- 204 patients were evaluable for toxicity during IC (supplemen-
dence interval at 95% level (95% CI), and tested for interaction. tary Table S1, available at Annals of Oncology online) and 384
Analyses were conducted according to the Intention-To-Treat (ITT) during concomitant treatment (Table 2).
except for compliance, safety and RR, performed considering patients in During induction TPF, neutropenia (27.5%) was the most rele-
the arm of treatment they actually received. vant G3–4 toxicity; febrile neutropenia was 11%. During

Randomization
(n = 421)

Concomitant IC->concomitant
n = 211 n = 210

n = 4 missing data° n = 2 missing data°

n = 1 ineligible

CCRT CET/RT TPF->CCRT TPF->CET/RT

(Arm A1) (Arm A2) (Arm B1) (Arm B2)
n = 128 n = 78 n = 129 n = 79

n = 206 evaluable for OS n = 208 evaluable for OS

n = 12 not evaluable for

n = 15 not evaluable for response to IC:
response: - never started 5
- never started 3 - consent withdrawn 1
- consent withdraw 2 - missing data 6
- lost to FU 2
- switch 2
- missing data 5 n = 27 not evaluable for
- unknown 1 response to concomitant:
- never started 7
- consent withdraw 5
- lost to FU 2
- switch 7
- missing data 4
- surgery after IC 1
- unknown 1

n = 201 evaluable for safety n = 204 evaluable for safety to IC

n = 183 evaluable for safety to concomitant
°uncompliant Center

Figure 1. Consort diagram.

doi:10.1093/annonc/mdx299 | 3
Original article Annals of Oncology
Table 1. Characteristics of the patients Table 2. G3–4 hematological and non-hematological toxicity during con-
comitant treatment
Concomitant IC->concomitant
(arms A11A2) (arms B11B2) Concomitant IC->concomitant P value
n 5206 (%) n 5208 (%) (arms A11A2) (arms B11B2)
n5201 (%) n 5183 (%)
Median age (years) 60 61
Range 27–81 37–78 Leukopenia 3 (1.5) 5 (3) 0.396
Gender M/F 168/38 (82/18) 170/38 (82/18) Neutropenia 2 (1) 8 (4) 0.038
ECOG PS Febrile neutropenia 2 (1) 0 0.200
0 171 (83) 162 (78) Anemia 1 (0.5) 4 (2) 0.145
1 34 (16.5) 46 (22) Thrombocitopenia 0 2 (1) 0.137
Missing 1 (0.5) 0 In-field mucositis 83 (41) 63 (34.5) 0.166
Tumor site In-field dermatitis 30 (15) 26 (14) 0.842
Oropharynx 114 (55) 119 (57) Skin rash 12 (6) 3 (1.5) 0.028
Oral cavity 44 (21.5) 38 (23.5) Fever 6 (3) 2 (1) 0.195
Hypopharynx 48 (23.5) 49 (18.5) Liver 1 (0.5) 0 0.344
Multiple site 0 2 (1) Renal G2–4 1a (0.5) 2a (1) 0.508
AJCC clinical stage Neurological 1 (0.5) 0 0.269
III 71 (34) 60 (29) Cardiac 0 0 –
IV 134 (65.5) 148 (71) Nausea/vomiting 0 0 –
Unknown 1 (0.5) 0 Diarrhea 0 0 –
T stage
a
Tx 2 (1) 1 (0.5) All grade 2.
T1 12 (6) 12 (6)
T2 36 (17.5) 45 (22)
T3 75 (36.5) 55 (26)
T4 81 (39) 95 (45.5)
N stage
Compliance
Nx 3 (1.5) 3 (1.5) 194 patients completed the three planned IC cycles. About nine
N0 22 (11) 22 (10.5) patients had to interrupt and five patients never started IC.
N1 32 (15.5) 31 (15) Compliance with concomitant treatment, independently from
N2 135 (65.5) 134 (65) the type of concomitant strategy, was similar in the IC and no-IC
N3 14 (6.5) 18 (9) arms (supplementary Table S2, available at Annals of Oncology
Reason for inoperability online). No significant differences were observed in completion
Technical unresectability 84 (41) 82 (40)
of the concomitant treatment (PF/cetuximab), median RT dose
Low surg.curability/organ pres. 95 (46) 103 (49.5)
and duration and number of patients who required RT interrup-
Missing 27 (13) 23 (10.5)
tion more than three consecutive days.
CCRT, concomitant chemoradiotherapy; AJCC, American Join Committee
on Cancer. Response rate
196 patients (94%) were evaluable for response to IC and 372
(90%) were evaluable after concomitant treatment.
After IC, overall RR (ORR) was 76%. ORRs after concomitant
treatment were similar, 80% versus 81%, while CRs were signifi-
cantly higher in the IC arm: 42.5% versus 28%, P ¼ 0.0028 (sup-
concomitant treatment, neutropenia G3–4 was significantly plementary Table S3, available at Annals of Oncology online).
higher in the IC arm (4% versus 1%, P ¼ 0.038). No significant More patients in the IC arm (8.5% versus 4%) received prophy-
differences were observed in G3–4 in-field mucositis and derma- lactic neck dissection (N2–N3 initial stage). Salvage surgery was
titis (34.5% versus 41% and 14% versus 15%, respectively) and in performed in 21.5% of patients in the IC arm versus 26.5% in the
other G3–4 toxicities, apart from skin-rash resulting from cetuxi- no-IC arm (P ¼ 0.229) (supplementary Figure S1, available at
mab administration in the no-IC arm (6% versus 1.5%; Annals of Oncology online).
P ¼ 0.028). The maximum grade of renal toxicity observed was
G2 in three patients.
Twelve early deaths (death within 30 days of last treatment ad-
Efficacy
ministration) occurred: five in the IC arm and seven in the no-IC After a median follow-up of 44.8 months 204 patients had died:
arm. The causes of early deaths were progressive disease (1), tox- 97 in the IC arm and 107 in the no-IC arm (supplementary Table
icity [5: 3 hematological toxicities (2 in the IC), 1 gastric perfor- S4, available at Annals of Oncology online). Median OS and the 3-
ation (IC arm) and 1 sepsis (IC arm)], unrelated to study year OS was higher in the IC arm: 54.7 months [interquartile
treatment (3) and unknown (3). range (IQR) 16.2–99.2] versus 31.7 months (IQR 12.4–71.2) and

4 | Ghi et al.
Annals of Oncology Original article
A 1.0 B 1.0
A: No IC A: No IC
0.9 0.9 B: IC
B: IC

0.8 0.8

Progression free survival

0.7 0.7
Overall survival

0.6 0.6

0.5 0.5

0.4 0.4
Number of events Number of events
A: 107 (51.9 %) A: 126 (61.2 %)
0.3 B: 97 (46.6 %) 0.3 B: 115 (55.3 %)

0.2 0.2
Log-rank: Chi2=4.69 df=1 p =0.030 Log-rank: Chi2=6.20 df=1 p =0.013
0.1 0.1
HR=0.74 (95% CI 0.56 - 0.97) p = 0.031 not adjusted HR=0.72 (95% CI 0.56 - 0.93) p = 0.013 not adjusted
HR=0.73 (95% CI 0.55 - 0.97) p = 0.029 adjusted HR=0.72 (95% CI 0.55 - 0.93) p = 0.013 adjusted
0.0 0.0
0 6 12 18 24 30 36 42 48 54 60 0 6 12 18 24 30 36 42 48 54 60
Time to event (months) Time to event (months)
Patients at risk Patients at risk
Time 0 6 12 18 24 30 36 42 48 54 60 Time 0 6 12 18 24 30 36 42 48 54 60
A 206 171 141 120 98 76 61 52 28 22 22 A 206 156 112 95 78 60 57 42 36 20 20
B 208 188 158 136 115 97 83 66 51 31 27 B 208 178 140 118 100 82 67 61 61 29 29

C 1.0 D 1.0

0.9 A: no IC 0.9 A: no IC
HR=0.74 (95% CI 0.55 - 0.98) p =0.0365 HR=0.76 (95% CI 0.46-1.25) p =2747
B: IC B: IC
0.8 0.8
Cumulative incidence (%)

Cumulative incidence (%)

0.7 0.7

0.6 0.6
Number of events Number of events
0.5 0.5
A: 99 (48.1 %) A: 33 (16.0 %)

0.4 B: 85 (40.9 %) 0.4 B: 27 (13.0 %)

0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0
0 6 12 18 24 30 36 42 48 54 60 0 6 12 18 24 30 36 42 48 54 60
Time to event (months) Time to event (months)
Patients at risk Patients at risk
Time 0 6 12 18 24 30 36 42 48 54 60 Time 0 6 12 18 24 30 36 42 48 54 60
A 206 156 112 95 78 60 57 42 36 20 20 A 206 156 112 95 78 60 57 42 36 20 20
B 208 178 140 118 100 82 67 61 61 29 29 B 208 178 140 118 100 82 67 61 61 29 29

Figure 2. Kaplan–Meier curves for OS (A) and PFS (B) of IC versus no-IC and cumulative incidence (competing risk analysis) for loco-regional*
(C) and distant events (D). *Loco-regional progression, death from cancer without documented progression or death from unknown causes
were considered loco-regional failure.

57.5% versus 46.5%, respectively (Figure 2A). Median PFS and non-oropharynx cancer (supplementary Figure S2, available at
the 3-year PFS were also in favor of IC: 30.5 months (IQR 11.2– Annals of Oncology online) and in patients receiving CET/RT
99.2) versus 18.5 months (IQR 7.0–71.2) and 47% versus 38.5%, after IC (HR 0.57; 95% CI 0.35–0.92) (supplementary Figure S3,
respectively (Figure 2B). available at Annals of Oncology online). The interaction test did
A statistically significant improvement in OS (HR 0.74; 95% CI not reach the statistical significance (P ¼ 0.088). A similar trend
0.56–0.97; P ¼ 0.031) and in PFS (HR 0.72; 95% CI, 0.56–0.93; was observed for PFS (supplementary Figure S4, available at
P ¼ 0.013) was detected in favor of IC and confirmed after adjust- Annals of Oncology online).
ment for prognostic factors (supplementary Tables S5 and S6,
available at Annals of Oncology online).
Loco-regional failure (Figure 2C) was lower in the IC arm:
41% versus 48% (HR 0.74; 95% CI 0.55–0.98; P ¼ 0.036) with a Discussion
borderline advantage at multivariate analysis (P ¼ 0.06, supple- Our trial aimed to test the hypothesis that adding induction TPF
mentary Table S7, available at Annals of Oncology online). Distant to concomitant treatment (CCRT or CET/RT) could improve
failure, with or without loco-regional recurrence, was 13% in the OS. The TPF and CCRT regimen derived from our previous feasi-
IC arm versus 16% in the no-IC arm (HR 0.76; 95% CI 0.46–1.25; bility study [11] based on the results of a GORTEC trial [12].
P ¼ 0.274) (Figure 2D). In accordance with study design, time-related events (OS, PFS,
The potential different effect of IC in relation to the primary LRC) are shown only for IC versus no-IC comparison. Since the
tumor site and to the subsequent concomitant strategy was also CET/RT arms have a shorter follow-up time and less patients
evaluated. Apparently, the effect was marked in patients with than CCRT arms, the comparison of the two concomitant

doi:10.1093/annonc/mdx299 | 5
Original article Annals of Oncology
treatments in terms of time-related events is likely to be heavily aim of limiting the adverse effects of concomitant chemotherapy,
biased. For these reasons CCRT versus CET/RT were compared unplanned dose reduction and RT treatment breaks.
only in terms of G3–4 acute toxicity (co-primary end point). Retrospective analyses in LAHNSCC seem to suggest that for cis-
This is the first trial showing a significant OS benefit for IC platin monotherapy 200 mg/m2 total dose could be as effective as
consistent with a significant improvement in CRs and PFS with a 300 mg/m [2, 9, 21]. In our trial the cumulative dose of cisplatin
borderline improvement in LRC. is 160 mg/m2, however 5FU was added with the aim of enhancing
We believe that the positive effect of IC on OS could be related the activity of the concomitant treatment both on loco-regional
to a combined effect on local and distant control. The hypothesis and distant control.
that induction TPF interacts positively with concomitant treat- Currently, CET/RT is considered an alternative treatment op-
ment, increasing both CRs and LRC, seems consistent with the tion to CCRT. The results of phase III trials comparing these two
MACH-NC meta-analysis [6] in which induction TPF was asso- concomitant strategies are not yet available and the retrospective
ciated with a better LRC over PF. comparisons have the limitations of selection bias and heteroge-
Recently, three phase III trials [13–15] failed to demonstrate a neous population [22]. Some data have been provided by a
survival benefit for the addition of IC to CCRT, although in two randomized phase-II larynx preservation trial comparing CCRT or
trials [13, 14] the planned accrual was not achieved. CET/RT in responding to induction TPF patients [23]. No differ-
The OS reported in our trial for the concomitant arm (3-year ences were observed in the organ preservation end points although
OS 46.5%) is similar to that reported in European trials [15–17] LRC seemed to be better in the CCRT arm. To learn more about
and lower than that observed in US trials [9, 13, 14, 20]. Studies the possible different efficacy of these two concomitant strategies
including US and non-US population had an intermediate OS [4, we have to wait for the results of the ongoing phase III trials.
5, 18]. It is our opinion that US patients are selected for fewer co- In conclusion, on the basis of our findings, adding induction
morbidities and these characteristics, together with higher inci- TPF to concomitant treatment significantly improves CRs, PFS
dence of HPV-related cancer in USA, are predictive of better and OS without compromising compliance to the concomitant
outcomes. treatments. The positive effect in LRC also appears to be a key
The OS multivariate Cox analysis confirmed the prognostic component of the IC strategy. The degree of benefit may differ ac-
value for PS, T stage and oropharynx primary sites but not for N cording to the type of the subsequent concomitant strategies.
stage. This observation seems consistent with a report on oro- Since this trial was not designed to investigate questions different
pharynx cancer [19], in which nodal level rather than nodal stage from the efficacy comparison of IC versus no-IC, we are not able
correlated with prognosis. Since our trial was designed in the to establish whether the benefit of IC is limited to specific sub-
early 2000s, HPV analysis was not planned. Exploratory analysis groups of patients or applies to the entire population.
for OS showed a higher effect of IC in non-oropharynx cancer, Although currently IC cannot be considered the standard of
however these results should be interpreted with caution. The care of LAHNSCC, nonetheless it should be regarded as an option
retrospective evaluation of HPV is ongoing and will enable us to for poor prognosis patients. Further efforts should also be made
better assess the results of our trial. to identify the best concomitant treatment to be administered
Subgroup analysis showed a possible superior effect of IC when after IC.
followed by CET/RT rather than CCRT. Although the interaction
test was not statistically significant (P ¼ 0.088), we cannot ex-
clude a possible interaction due to the lack of statistical power for Acknowledgements
this evaluation. The possible interaction was only quantitative, We thank Laura McMahon and Paola Bardelle (Project
not qualitative, since a survival benefit was observed for IC in Manager); Jean Pierre Pignon, MD, PhD and Pierre Blanchard,
each concomitant treatment. These data should be interpreted MD (Meta-Analysis Unit, Biostatistics and Epidemiology
with caution and were only for explorative purpose. Department, Institute Gustave Roussy, Villejuif Cedex, France);
Currently, we have no explanation for the possible different ef- Ezra E. Cohen, MD (University of California, San Diego,
fect of IC when followed by CCRT or CET/RT. We know that Moores Cancer Center); Enzo Bagnardi, Maurizio Marangolo
combining CCRT and a biological drug simultaneously is not a and Giancarlo Martignoni (Independent Data Monitoring
winning strategy [20] and we can only speculate a possible higher Committee); Margaret Fraser.
sensitizing effect of cetuximab on cancer cells surviving the The authors commemorate the colleague Irene Floriani of the
platinum-based IC. Whether induction TPF might have different Mario Negri Institute, who passed away on 12 January 2016 and
efficacy depending on the subsequent concomitant treatment de- who dedicated her life to research. She has been involved in the
livered is an intriguing issue and needs further exploration. statistical design and coordination of this trial.
With regard to safety to concomitant treatment, G3–4 neutro- Access to data: MGG and AP had full access to all the data in
penia was significantly higher, but negligible, in patients receiving the study and take responsibility for the integrity of the data
IC. Conversely, G3–4 non-hematological toxicities were not and the accuracy of the data analysis.
worse and compliance was similar too. Good compliance to IC
and to the subsequent concomitant treatment might be attrib-
uted to the schedule and doses of the induction TPF and of the
CCRT regimen adopted [11]. Funding
As discussed in our previous publication [7], we believe that The randomized phase II part of the study (XRP 6976 F/2501)
the comparator CCRT arm should be considered appropriate. In was sponsored by Sanofi-Aventis, Italy. The sponsor for the
our trial, two cycles of concomitant PF were planned with the phase III part of the study (NCT01086826), is the onlus AVAPO

6 | Ghi et al.
Annals of Oncology Original article
(Associazione Volontari Assistenza ai Pazienti Oncologici), concomitant chemoradiotherapy: a phase II feasibility study. Int J Radiat
Venezia, Italy. The funding sources had no role in the design Oncol Biol Phys 2004; 59: 481–487.
12. Calais J, Alfonsi M, Etienne B et al. Randomized trial of radiation therapy
and conduct of the study, collection, management, analysis/in-
versus concomitant chemotherapy and radiation therapy for advanced-
terpretation of the data, preparation, review/approval of the stage oropharynx carcinoma. J Natl Cancer Inst 1999; 91(24):
manuscript and decision to submit the manuscript for 2081–2086.
publication. 13. Haddad R, O’Neill A, Rabinowits G et al. Induction chemotherapy fol-
lowed by concurrent chemoradiotherapy (sequential chemoradiother-
apy) versus concurrent chemoradiotherapy alone in locally advanced
head and neck cancer (PARADIGM): a randomised phase 3 trial. Lancet
Disclosure Oncol 2013; 14: 257–264.
The authors have declared no conflicts of interest. 14. Cohen EE, Karrison TG, Kocherginsky M et al. Phase III randomized trial
of induction chemotherapy in patients with N2 or N3 locally advanced
Head and Neck Cancer. J Clin Oncol 2014; 32: 2735–2743.
References 15. Hitt R, Grau JJ, L opez-Pousa A et al. A randomized phase III trial com-
paring induction chemotherapy followed by chemoradiotherapy versus
1. International Agency for Research on Cancer. World Health
chemoradiotherapy alone as treatment of unresectable head and neck
Organization. Globocan 2012. http://globocan.iarc.fr.
cancer. Ann Oncol 2014; 25: 216–225.
2. Pignon JP, Bouhris J, Domenge C on behalf of the MACH-NC
16. Bourhis J, Lapeyre M, Tortochaux J et al. Accelerated radiotherapy and
Collaborative group. Chemotherapy added to locoregional treatment for
concomitant high dose chemotherapy in non resectable stage IV locally
head and neck squamous-cell carcinoma: three meta-analyses of updated
advanced HNSCC: results of a GORTEC randomized trial. Radiother
individual data. Lancet 2000; 355: 949–955.
Oncol 2011; 100: 56–61.
3. Pignon JP, Le Maı̂tre A, Maillard E et al. Meta-analysis of chemotherapy
17. Bourhis J, Sire C, Graff P et al. Concomitant chemoradiotherapy
in head and neck cancer (MACH-NC): an update on 93 randomized tri-
versus acceleration of radiotherapy with or without concomitant
als and 17,346 patients. Radiother Oncol 2009; 92: 4–14.
chemotherapy in locally advanced head and neck carcinoma (GORTEC
4. Bonner J, Harari PM, Giralt J et al. Radiotherapy plus cetuximab for
99-02): an open-label phase 3 randomised trial. Lancet Oncol 2012; 13:
squamous-cell carcinoma of the head and neck. N Engl J Med 2006; 354:
145–153.
567–578.
18. Rischin D, Peters LJ et al. Tirapazamine, cisplatin, and radiation versus
5. Bonner JA, Harari PM, Giralt J et al. Radiotherapy plus cetuximab for
cisplatin and radiation for advanced squamous cell carcinoma of the
locoregionally advanced head and neck cancer: 5-year survival data from
head and neck (TROG 02.02, HeadSTART): a phase III trial of the
a phase 3 randomised trial, and relation between cetuximab-induced
Trans-Tasman Radiation Oncology Group. J Clin Oncol 2010; 28(18):
rash and survival. Lancet Oncol 2010; 11: 21–28.
2989–2995.
6. Blanchard P, Bourhis J, Lacas B et al. Taxane-cisplatin-fluorouracil as in-
19. Garden AS, Kies MS, Morrison WH et al. Outcomes and patterns of care
duction chemotherapy in locally advanced head and neck cancers: an in-
of patients with locally advanced oropharyngeal carcinoma treated in the
dividual patient data meta-analysis of the meta-analysis of chemotherapy
early 21st century. Radiat Oncol 2013; 8: 21–30.
in Head and Neck Cancer Group. J Clin Oncol 2013; 31: 2854–2860.
7. Paccagnella A, Ghi MG, Loreggian L et al. Concomitant chemoradiother- 20. Ang KK, Zhang Q, Rosenthal DI et al. Randomized phase III trial of con-
apy versus induction docetaxel, cisplatin and 5 fluorouracil (TPF) followed current accelerated radiation plus cisplatin with or without cetuximab
by concomitant chemoradiotherapy in locally advanced head and neck for stage iii to iv head and neck carcinoma: RTOG 0522. J Clin Oncol
cancer: a phase II randomized study. Ann Oncol 2010; 21: 1515–1522. 2014; 32: 2940–2950.
8. Byar DP. Factorial and reciprocal control designs. Stat Med 1990; 9(1–2): 21. Spreafico A, Hui Huang S, Xu W et al. Differential impact of cisplatin
55–63. in Italian. dose intensity on human papillomavirus-related and HPV-unrelated
9. Nguyen-Tan PF, Zhang Q, Ang KK et al. Randomized phase III trial to locoregionally advanced head and neck squamous cell carcinoma
test accelerated versus standard fractionation in combination with con- (LAHNSCC). J Clin Oncol 2015; 33: abstract 6020.
current cisplatin for head and neck carcinomas in the Radiation Therapy 22. Petrelli F, Coinu A, Riboldi V et al. Concomitant platinum-based chemo-
Oncology Group 0129 Trial: long-term report of efficacy and toxicity. therapy or cetuximab with radiotherapy for locally advanced head and
J Clin Oncol 2014; 32: 3858–3867. neck cancer: a systematic review and meta-analysis of published studies.
10. Gray RJ. A proportional hazards model for the subdistribution of a com- Oral Oncol 2014; 50(11): 1041–1048.
peting risk. J Am Stat Assoc 1999; 94(446): 496–509. 23. Lefebvre JL, Pointreau Y, Rolland F et al. Induction chemotherapy fol-
11. Ghi MG, Paccagnella A, D’Amanzo P et al. Neoadjuvant docetaxel, cis- lowed by either chemoradiotherapy or bioradiotherapy for larynx preser-
platin, 5-fluorouracil before concurrent chemoradiotherapy in locally vation: the TREMPLIN Randomized Phase II Study. J Clin Oncol 2013;
advanced squamous cell carcinoma of the head and neck versus 31: 853–859.

doi:10.1093/annonc/mdx299 | 7