Methods

Location of trial: US and Canada

Number of centres: > 150

Trial ID: RTOG 0522; NCT00265941

Recruitment period: November 2005 to March 2009

Funding source: NCI U10 CA21661, U10 CA37422 and Bristol‐Myers Squibb

Participants

Inclusion criteria: people with stage III‐IV carcinoma of the oropharynx, larynx and hypopharynx; having Zubrod performance of 0‐1 and meeting pre‐defined blood chemistry criteria

Exclusion criteria: prior malignancy; primary tumour in OC, nasopharynx, sinus or salivary glands; prior excision of primary tumour; prior treatment (chemotherapy or RT); severe active co‐morbidity (defined) or pregnancy

Sex (M/F): 88%/12%

Age at baseline (years): Gp A: range 34‐76, median 58; Gp B: range 31‐79, median 57

OP: Gp A: 312/444 (70%), Gp B: 313/447 (70%)

Number randomised: 940

Number evaluated: 891

Interventions

Comparison: cisplatin + RT + cetuximab vs. cisplatin + RT

Gp A (n = 444): loading dose cetuximab 400 mg/m² and 6‐7 weekly doses of cetuximab 250 mg/m² given in conjunction with RT 70‐72 Gy (6 wk) + 2 cycles of cisplatin (q3 weeks)

Gp B (n = 447): RT 70‐72 Gy (6 wk) + 2 cycles of cisplatin (q3 weeks)

Outcomes

Primary: PFS

Secondary: OS, locoregional failure, distant metastasis, acute and late adverse events, quality of life (reported separately)

Duration of follow‐up: median follow‐up 3.8 years

Notes

Sample size calculation: "RTOG 0522 was initially designed with a sample size of 720 patients to detect a 25% reduction in the hazard associated with disease‐free survival with 80% power and a one‐sided test at the P=.025 level.
The primary end point was changed to PFS in 2008 to allow comparisons with the end point in the international meta‐analysis of event‐free survival, which has been shown to be a surrogate for OS.
In addition, because the control group had better‐than expected disease‐free survival/PFS, the sample size was increased to 945 patients to allow detection of a 25% reduction in the risk of PFS failure with 84% statistical power and a one‐sided final test at the P=.0238 significance level, after three interim analyses and a planned final analysis at 434 treatment failures.
When the third planned interim analysis yielded a conditional power of less than 10%, the data monitoring committee recommended early reporting of results with 371 failure events"

Selected reported adverse events: most frequently reported (top 5 + any event category) included in Table 8 8: for full listing of reported adverse effects, see original study reference

Toxicity evaluation assessment tool: National Cancer Institute Common Terminology Criteria for Adverse Events (version 3)

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Stratified, computer‐generated central randomisation: "Patients were...randomly assigned to...arm A or...arm B in a 1:1 ratio using permuted block random assignment"

Allocation concealment (selection bias)

Low risk

Central allocation revealed to investigator after participant registration

Blinding of participants (performance bias)

Unclear risk

Open label, no placebo. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms; however, primary outcomes considered to be objective (e.g. OS)

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Outcomes evaluated according to pre‐specified objective criteria and reviewed by blinded observers

Incomplete outcome data (attrition bias)
All outcomes

Low risk

90% of participants received 2 cisplatin cycles and 74% of participants randomised to cetuximab received loading dose plus ≥ 6 doses. Of 940 participants accrued, 2 lost to follow‐up and 43 were ineligible on review. Similar numbers lost from each group

Selective reporting (reporting bias)

Low risk

Planned outcomes reported

Other bias

Unclear risk

Intervention contamination between groups. Conduct of study affected by interim results (recruited additional participants after interim results indicated control group demonstrated better PFS/DFS outcomes than anticipated; sample size was changed several times mid‐study)

Study ID

Specified adverse effect

Intervention arm

Control arm

Ang 2014

Comparison: mAb

(CRT + cetuximab vs. CRT (no placebo))

Acute period : ≤ 90 days from start of RT (Gp A n = 444; Gp B n = 447)

Any event

All grades: n = 431 (97%);
Grades 3‐4: n = 395 (89%)

All grades: n = 434 (97%);
Grades 3‐4: n = 389 (87%)

Dysphagia

All grades: n = 364 (82%);
Grades 3‐4: n = 235 (53%)

All grades: n = 384 (86%);
Grades 3‐4: n = 255 (57%)

Mucositis

All grades: n = 364 (82%);
Grades 3‐4: n = 191 (43%)

All grades: n = 322 (72%);
Grades 3‐4: n = 147 (33%)

Skin reaction outside portal (pruritus; dermatitis exfoliative NOS; acne NOS; nail disorder NOS)

All grades: n = 364 (82%);
Grades 3‐4: n = 89 (20%)

All grades: n = 63 (14%);
Grades 3‐4: n = 4 (1%)

Skin reaction inside portal (radiation dermatitis NOS; radiation recall syndrome)

All grades: n = 346 (78%);
Grades 3‐4: n = 111 (25%)

All grades: n = 353 (79%);
Grades 3‐4: n = 67 (15%)

Fatigue

All grades: n = 289 (65%);
Grades 3‐4: n = 62 (14%)

All grades: n = 268 (60%);
Grades 3‐4: n = 40 (9%)

Late period : > 90 days from start of RT (Gp A n = 415; Gp B n = 432)

Any event

All grades: n = 403 (97%);

Grades 3‐4: n = 249 (60%)

All grades: n = 419 (97%);

Grades 3‐4: n = 233 (54%)

Dysphagia

All grades: n = 357 (86%);

Grades 3‐4: n = 154 (37%)

All grades: n = 359 (83%);

Grades 3‐4: n = 156 (36%)

Dry mouth

All grades: n = 311 (75%);

Grades 3‐4: n = 21 (5%)

All grades: n = 324 (75%);

Grades 3‐4: n = 17 (4%)

Skin fibrosis

All grades: n = 311 (46%);

Grades 3‐4: n = 8 (2%)

All grades: n = 190 (44%);

Grades 3‐4: n = 4 (1%)

Fatigue

All grades: n = 170 (41%);

Grades 3‐4: n = 12 (3%)

All grades: n = 194 (45%);

Grades 3‐4: n = 13 (3%)

Laryngeal oedema

All grades: n = 166 (40%);

Grades 3‐4: n = 17 (4%)

All grades: n = 181 (42%);

Grades 3‐4: n = 13 (3%)

Bonner 2006

Comparison: mAb

RT + cetuximab vs. RT (no placebo)

(Gp A n = 208; Gp B n = 212)

Mucositis

All grades: n = 193 (93%);

Grades 3‐5: n = 116 (56%)

All grades: n = 199 (94%);

Grades 3‐5: n = 110 (52%)

Acneiform rash

All grades: n = 181 (87%);

Grades 3‐5: n = 35 (17%)

All grades: n = 21 (10%);

Grades 3‐5: n = 2 (1%)

Radiation dermatitis

All grades: n = 179 (86%);

Grades 3‐5: n = 48 (23%)

All grades: n = 191 (90%);

Grades 3‐5: n = 38 (18%)

Weight loss

All grades: n = 175 (84%);

Grades 3‐5: n = 23 (11%)

All grades: n = 153 (72%);

Grades 3‐5: n = 15 (7%)

Xerostomia

All grades: n = 150 (72%);

Grades 3‐5: n = 10 (5%)

All grades: n = 151 (71%);

Grades 3‐5: n = 6 (3%)

Dysphagia

All grades: n = 135 (65%);

Grades 3‐5: n = 54 (26%)

All grades: n = 134 (63%);

Grades 3‐5: n = 64 (30%)

Koh 2013

Comparison: mAb

(CRT + cetuximab vs. CRT (no placebo))

Gp A n = 48; Gp B n = 44

Partial narrative reporting for Gp A toxicities relating to attrition rationale in abstract:

"Reason for incompletion in CDP arm included hypersensitivity (1), septic shock (1), skin rash (1), seizure (1), arterial thrombosis (1), unexplained death (1), unsatisfactory response (1), and withdrawal of informed consent (1)". No detail reported for Gp B. However, poster presented alongside abstract indicates similar frequency of Grade 3‐4 toxicities in both Gps

Neutropenia

CRT induction

Concurrent CRT

Grades 3‐4: n = 13 (27%)

Grades 3‐4: n = 5 (10%)

Grades 3‐4: n = 5 (11%)

Grades 3‐4: n = 4 (9%)

Anorexia

CRT induction

Concurrent CRT

Grades 3‐4: n = 3 (6%)

Grades 3‐4: n = 6 (13%)

Grades 3‐4: n = 4 (9%)

Grades 3‐4: n = 4 (9%)

Mucositis

CRT induction

Concurrent CRT

Grades 3‐4: n = 4 (8%)

Grades 3‐4: n = 9 (19%)

Grades 3‐4: n = 0 (0%)

Grades 3‐4: n = 3 (7%)

Febrile neutropenia

CRT induction

Concurrent CRT

Grades 3‐4: n = 7 (%)

Grades 3‐4: n = 3 (6%)

Grades 3‐4: n = 4 (9%)

Grades 3‐4: n = 0 (0%)

Skin toxicity

CRT induction

Concurrent CRT

Grades 3‐4: n = 3 (6%)

Grades 3‐4: n = 4 (8%)

Grades 3‐4: n = 0 (0%)

Grades 3‐4: n = 1 (2%)

Diarrhoea

CRT induction

Concurrent CRT

Grades 3‐4: n = 4 (8%)

Grades 3‐4: n = 0 (0%)

Grades 3‐4: n = 3 (7%)

Grades 3‐4: n = 0 (0%)

Reddy 2014

Comparison: mAb

CRT + nimotuzumab vs. CRT vs. RT + nimotuzumab vs. RT

(Gp A n = 23; Gp B n = 23; Gp C n = 23; Gp D n = 23)

**Note: control arm mucositis events Gp D n = 27 (+4 than Gp n), therefore, all grade mucositis data not used in analysis

Mucositis

All grades: n = 44 (96%);

Grade 3: n = 21 (46%)

All grades: n = 49 (106%)** ;

Grade 3: n = 21 (46%)

Skin reaction

All grades: n = 37 (80%);

Grade 3: n = 0 (0%)

All grades: n = 35 (76%);

Grade 3: n = 2 (4%)

Nausea/vomiting

All grades: n = 34 (74%);

Grade 3: n = 1 (2%)

All grades: n = 30 (65%);

Grade 3: n = 0 (0%)

Salivary gland disorder

All grades: n = 27 (59%);

Grade 3: n = 0 (0%)

All grades: n = 29 (63%);

Grade 3: n = 2 (4%)

Dysphagia

All grades: n = 16 (35%);

Grade 3: n = 5 (%)

All grades: n = 16 (35%);

Grade 3: n = 1 (2%)

Candidiasis

All grades: n = 13 (28%);

Grade 3: n = 5 (11%)

All grades: n = 19 (41%);

Grade 3: n = 7 (15%)

Rodriguez 2010

Comparison: mAb

RT + nimotuzumab vs. RT + placebo

(Gp A n = 54; Gp B n = 52)

Any adverse event

n = 38 (70%)

n = 30 (58%)

Mucositis

n = 11 (20%)

n = 9 (17%)

Dry mouth

n = 9 (17%)

n = 12 (23%)

Dry radio‐dermatitis

n = 6 (11%)

n = 6 (12%)

Odynophagia

n = 4 (7%)

n = 6 (12%)

Methods

Location of trial: Germany

Number of centres: 1

Trial ID: prior to trial registrations

Recruitment period: period of 3 years ‐ dates not reported

Funding source: Grant no. Bi 230/2, awarded by the Deutsche Forschungsgemeinschaft

Participants

Inclusion criteria: participants had histologically confirmed HNSCC with tumour stage T1/2, N0‐2M0

Exclusion criteria: not reported

Sex (M/F): 14/10

Age at baseline (years): mean 68

OC: Gp A: n = 10; Gp B: n = 8

OP: Gp A: n = 0; Gp B: n = 2

Lip cancers: Gp A: n = 2; Gp B: n = 2

Number randomised: 24

Number evaluated: 24

Interventions

Comparison: BCG‐CWP+ surgery versus surgery alone

Gp A (n = 10): 3 mL BCG‐CWP was injected under local anaesthesia in a fan like manner over the entire tumour very slowly. Surgery was then performed when the inflammatory infiltrate had subsided (mean 24 d, range 16‐43 d)

Gp B (n = 10): surgery only

Outcomes

Primary: recurrence, survival

Secondary: changes in clinical, laboratory‐chemical and immunological findings pre and post BCG‐CWP therapy

Duration of follow‐up: 1 year

Notes

Adverse events: included in Table 9

Toxicity evaluation assessment tool: not reported

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

"randomised". Method of sequence generation not described

Allocation concealment (selection bias)

Unclear risk

Allocation concealment not mentioned

Blinding of participants (performance bias)

Unclear risk

Not reported, no placebo. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms; however, primary outcomes considered to be objective

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Not reported, no placebo

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All randomised participants included in outcomes

Selective reporting (reporting bias)

Low risk

Planned outcomes reported

Other bias

Low risk

No other sources of bias identified

Methods

Location of trial: international; US (73 centres), Australia (5), South Africa (5), France (5), Israel (3), UK (3), Belgium (2), Spain (2), Poland (2), Switzerland (2), Sweden (1), Germany (1), New Zealand (1)

Number of centres: 105

Trial ID:

Recruitment period: April 1999 to March 2002

Funding: ImClone Systems, New York, Merck (Darmstadt Germany) and National Institutes of Health grant (CA06294)

Participants

Inclusion criteria: patients with stage III or IV, non‐metastatic measurable SCC of oropharynx, hypopharynx or larynx. Medically suitable, KPS > 60 and normal haematopoietic, renal and hepatic function

Exclusion criteria: previous cancer or chemotherapy in past 3 years, or previous surgery or RT for head and neck cancer

Sex (M/F): Gp A: 171/40, Gp B: 169/44

Age at baseline: Gp A: median 56 years, Gp B median 58 years

OC: none

OP: Gp A: 118/211(56%), Gp B: 135/213 (63%)

Number randomised: 424

Number evaluated: 424

Interventions

Comparison: cetuximab + RT vs. RT alone

Gp A (n = 211): high‐dose RT + cetuximab

Gp B (n = 213): high‐dose RT alone

On study registration all investigators selected 1 of 3 RT regimens:

• 2.0 Gy/fraction, 5 fractions/wk for 7 wk (total radiation dose: 70.0 Gy in 35 fractions)

• 1.2 Gy/fraction, 2 fractions/d, for 6‐6.5 wk (total radiation dose: 72.0‐76.8 Gy in 60‐64 fractions)

• 1.8 Gy/fraction, 5 fractions/wk for 3.6 wk (32.4 Gy) (total radiation dose: 72.0 Gy in 42 fractions)

then twice daily; mornings: 1.8 Gy/fraction 5 fractions/wk for 2.4 wk

afternoons: 1.5 Gy/fraction, 5 fractions/wk for 2.4 wk

Duration of follow‐up: 5 years

Outcomes

LRC, OS, PFS, ORR, safety

Notes

Sample size: to detect an increase in the rate of LRC at 1 year, from 44% to 57% it was calculated the 208 participants per treatment group were required to give 90% power at the 5% significance level with use of a 2‐sided log‐rank test

Adverse events note: different percentages presented in later publication (2010; Table 2)

Selected reported adverse events: most frequently reported (top 6) included in Table 8 8: for full listing of reported adverse effects, see original study reference's Table 4)

Toxicity evaluation assessment tool: toxicity criteria of the Radiation Therapy Oncology Group (1995)

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Randomisation was stratified for KPS (60‐80 vs 90‐100), nodal involvement (N0 vs N+) tumour stage (T1‐3 vs T4) and radiation fractionation regimen (1x/day vs 2x/day vs concom boost). A minimisation method was used in the randomisation of patients to receive RT + cetuximab or RT alone)"

Allocation concealment (selection bias)

Low risk

RT regimens were chosen prior to randomisation based on preset clinical criteria. Allocation was obtained using an Interactive Voice Response System, which informed investigators of study number and treatment group

Blinding of participants (performance bias)

Unclear risk

Open‐label, no placebo. Knowledge of allocated treatment may have influenced participants' care.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Investigator‐generated data were submitted for blind review by an independent committee of experts

Incomplete outcome data (attrition bias)
All outcomes

Low risk

ITT analysis for efficacy included all randomised participants

Selective reporting (reporting bias)

Low risk

Planned and expected primary outcomes reported in full

Other bias

Low risk

No other biases identified

Methods

Location of trial: Torino, Italy

Number of centres: 5 centres

Trial ID: not given

Recruitment period: May 1990 to December 1996

Funding: Italian Association for Cancer Research and from the Italian Ministry for the Universities and Scientific and Technological Research

Participants

Inclusion criteria: patients with histologically confirmed resectable stage T2‐4, M0, N0‐N3, SCC of OC or oropharynx. KPS at least 70%, aged 18‐75 years, medically fit for surgery, WBC > 4000/mm³, creatinine < 1.4 mg/dL, platelets > 150,000/mm³, AAT and ALP levels < 1.5 normal

Exclusion criteria: patients with infiltration of the pterygopalatine fossa and carotid artery were classed as inoperable. Patients with tumours of the anterior floor of the mouth and the base of the tongue were excluded due to their high risk of cervical metastasis and need for bilateral neck dissection. Patients with previous treatment for this or any other malignancy and the need for steroid management were excluded

Age at baseline (years): Gp A: median 62, Gp B: median 61

OC: Gp A: n = 64; Gp: B n = 61 (125 evaluated)

OP: Gp A: n = 36; Gp B: n = 40 (76 evaluated)

OC + OP: 201 evaluated

Number randomised: 220 (110 and 110)

Number evaluated: 201

Interventions

Comparison: rIL‐2 + surgery (± RT) versus surgery (± RT)

Gp A (n = 100): daily course of rIL‐2 (5000 units in 1 mL phosphate buffered saline/albumin) in the ipsilateral cervical lymph node chain for 10 d prior to surgery. 2 administration sites in each person. Surgery ± RT. Further 5‐d courses were given monthly in the residual contralateral chain for 1 year after surgery or after RT when this was necessary

Gp B (n = 101): surgery ± RT

Duration of follow‐up: median 64 months (minimum of 2 years)

Outcomes

Primary: OS

Secondary: DFS, recurrence, new primary tumours

Notes

Sample size: planned number of participants was 250, because a 5‐year DFS rate of 55% was presumed for the control group, and a 10% increase in survival was presumed in the rIL‐2 group

Adverse events: included in Table 9 9

Toxicity evaluation assessment tool: not reported

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were assigned randomly to the 2 treatment groups by a central office (Chiron Medical Department, Amsterdam, The Netherlands). Randomisation stratified by participant gender and date of birth, tumour site (OC or oropharynx), TNM classification, and KPS score. Method of sequence generation not stated

Allocation concealment (selection bias)

Low risk

Allocation conducted by central office. Concealment not specifically reported

Blinding of participants (performance bias)

Unclear risk

Not reported, no placebo. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms; however, primary outcomes considered to be objective

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Not reported

Incomplete outcome data (attrition bias)
All outcomes

High risk

10 participants lost from rIL‐2 group (due to interruption of planned treatment after randomisation) and 9 lost from control group (4 died due to myocardial infarction, 5 refused to take part). Reasons different in each group and probably related to allocated treatment

Selective reporting (reporting bias)

Low risk

Planned and expected outcomes reported

Other bias

Low risk

Unplanned interim analysis resulted in early closure of study. Guidance from Higgins 2011 is that simulation evidence suggests that inclusion of stopped early trials in meta‐analyses will not lead to substantial bias

Study ID

Specified adverse effect

Intervention arm

Control arm

De Stefani 2002

Comparison: interleukin

rIL‐2 + RT + surgery vs. RT + surgery (no placebo)

(Gp A n = 100; Gp B n = 101)

Narrative commentary only:

"Complication and toxicity rates for the surgery arm and the surgery plus radiotherapy arm were the same in the control group and in the rIL‐2 group.

Postoperative radiotherapy originates well‐known side effects, but they were independent of the preoperative rIL‐2 treatment. [...]

Neoadjuvant rIL‐2 injections did not complicate the surgical treatment, and adjuvant rIL‐2 injections did not increase distant side effects due to previous surgery or radiotherapy"

Mantovani 1998

Comparison: interleukin

rIL‐2 + CRT ± surgery vs. CRT ± surgery (no placebo)

(Gp A n = 14; Gp B n = 16)

Nausea/vomiting

All grades: n = 13 (93%);

Grades 3‐4: n = 1 (7%)

All grades: n = 12 (75%);

Grades 3‐4: n = 0 (0%)

Stomatitis

All grades: n = 9 (64%);

Grades 3‐4: n = 3 (21%)

All grades: n = 7 (44%);

Grades 3‐4: n = 2 (13%)

Leukopenia

All grades: n = 4 (29%);

Grades 3‐4: n = 1 (7%)

All grades: n = 8 (50%);

Grades 3‐4: n = 3 (19%)

Fever

All grades: n = 6 (43%);

Grades 3‐4: n = 0 (0%)

All grades: n = 0 (0%);

Grades 3‐4: n = 0 (0%)

Diarrhoea

All grades: n = 2 (14%);

Grades 3‐4: n = 1 (7%)

All grades: n = 3 (19%);

Grades 3‐4: n = 1 (6%)

Anaemia

All grades: n = 3 (21%);

Grades 3‐4: n = 1 (7%)

All grades: n = 2 (13%);

Grades 3‐4: n = 0 (0%)

Bier 1981

Comparison:

BCG‐CWP + surgery vs. surgery (no placebo)

(Gp A n = 12; Gp B n = 12)

Increased temperature
(> 3 to ≤ 14 days post treatment)

n = 2 (17%)

n = 3 (25%)

Moderate‐severe chills

n = 5 (42%)

n = 0 (0%)

Gastrointestinal complaints (including nausea/vomiting)

n = 2 (17%)

n = 0 (0%)

Methods

Location of trial: International; Belgium, Czech Republic, Germany, India, Poland, Serbia, Taiwan, US

Number of centres: 26

Trial ID: NCT00229723

Recruitment period: November 2004 to June 2008

Funding source: AstraZeneca

Participants

Inclusion criteria: patients with previously untreated histologically confirmed stage III/IV non‐metastatic HNSCC with measurable disease (by Response Evaluation Criteria in Solid Tumours ‐ (RECIST)) , aged ≥ 18 years, WHO performance status 0‐1, life expectancy ≥ 12 wk

Exclusion criteria: buccal, mucosal carcinoma and post nasal space, thyroid or salivary gland tumours, stage III laryngeal carcinoma or metastatic disease, disease invading mandible, simultaneous primary tumours, abnormal renal or liver function, neutrophils < 1 x 10⁹/L, platelets < 100 x 10⁹/L, pregnancy or breastfeeding

Sex (M/F): 87.6%/12.4%

Age at baseline (mean (SD)) (years): Gp A: 53.1 (8.36), Gp B: 52.5 (12.37), Gp C: 53.4 (10.59), Gp D: 54.3 (8.56), Gp E: 50.8 (9.29), Gp F: 53.4 (6.22), Gp G: 54.9 (9.95)

OC: Gp A: 4; Gp B: 0; Gp C: 2; Gp D: 1; Gp E: 3; Gp F: 2; Gp G: 3; total 15

OP: Gp A: 26; Gp B: 12; Gp C: 15; Gp D: 19; Gp E: 13; Gp F: 15; Gp G: 10; total 110

OC + OP: 125/226 (55%)

Number randomised: 226

Number evaluated: 83 who completed study prior to interim analysis

Interventions

Comparisons: gefitinib 250 mg/d vs. gefitinib 500 mg/d vs. placebo (in concomitant and maintenance phases)

Gp A (n = 60): placebo (both concomitant and maintenance treatment phases)

Gp B (n = 24): gefitinib 250 mg/d orally concomitant and placebo maintenance phase

Gp C (n = 31): gefitinib 500 mg/d orally and placebo maintenance phase

Gp D (n = 31): gefitinib 250 mg/d orally concomitant and gefitinib 250 mg/d maintenance

Gp E (n = 24): gefitinib 500 mg/d orally concomitant and gefitinib 500 mg/d maintenance

Gp F (n = 34): placebo in concomitant phase and gefitinib 250 mg/d maintenance

Gp G (n = 22): placebo in concomitant phase and gefitinib 500 mg/d maintenance

In the concomitant phase, all participants received CRT comprising cisplatin (100 mg/m² IV every 21 d (3 cycles) plus 3D conformal RT 70 Gy in 35 fractions over 7 wk) plus intervention tablets administered daily from the day of randomisation to the end of RT (8‐9 wk in total)

Maintenance phase lasted up to 2 years

Duration of follow‐up: 2 years

Outcomes

Primary: LRC at 2 years

Secondary: LRC at 1 year, CRR, ORR, PFS, OS, safety and tolerability

Duration of follow‐up: 2 years

Notes

7 arm trial ‐ only 1 comparison (gefitinib concomitant vs. placebo (Gp B+C vs. A), included in this review's primary outcomes

Note: data are held by the pharmaceutical company that produces gefitinib, unable to interrogate reported HRs

Adverse events: included in Table 10 10 (CRT + gefitinib (250/500 mg) vs. CRT + placebo concomitant phase only: Gp A: n = 110 (Gps B+C+D+E); Gp B: n = 116 (Gps A+F+G))

Toxicity evaluation assessment tool: National Cancer Institute Common Terminology Criteria for Adverse Events (version 3)

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Randomisation scheme was prepared by Astra Zeneca statistical personnel …. and codes were allocated strictly sequentially as patients entered the study"

Allocation concealment (selection bias)

Low risk

Central randomisation schedule; however, concealment not specifically reported

Blinding of participants (performance bias)

Low risk

"Each patient received an identical kit of two matching bottles of tablets and took one tablet from each bottle every day

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

"Placebo controlled double‐blind"

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Outcomes were not reported for each randomised group and it was difficult to determine the numbers of participants excluded from the analysis

Selective reporting (reporting bias)

High risk

Planned outcomes of LRC, CRR and ORR reported but, but only for combinations of groups. Data on PFS and OS not found

Other bias

Low risk

No other sources of bias identified

Study ID

Specified adverse effect

Intervention arm

Control arm

Gregoire 2011

Comparison: TKI

(CRT + gefitinib (250/500 mg) vs.
CRT + placebo) (concomitant phase
only: Gp A n = 110 (Gps B+C+D+E);

Gp B n = 116 (Gps A+F+G))

Mucositis

All grades: n = 96 (87%);

Grades 3‐5: n = 51 (%)

All grades: n = 98 (84%);

Grades 3‐5: n = 42 (36%)

Nausea

All grades: n = 50 (45%);

Grades 3‐5: n = 4 (4%)

All grades: n = 55 (47%);

Grades 3‐5: n = 3 (3%)

Vomiting

All grades: n = 56 (51%);

Grades 3‐5: n = 6 (5%)

All grades: n = 52 (45%);

Grades 3‐5: n = 9 (8%)

Dysphagia

All grades: n = 29 (26%);

Grades 3‐5: n = 5 (5%)

All grades: n = 43 (37%);

Grades 3‐5: n = 13 (11%)

Dry mouth (xerostomia)

All grades: n = 33 (30%);
Grades 3‐5: n = 1 (1%)

All grades: n = 30 (26%);

Grades 3‐5: n = 2 (2%)

Radiation skin injury

All grades: n = 29 (26%);

Grades 3‐4: n = 2 (2%)

All grades: n = 29 (25%);

Grades 3‐5: n = 4 (3%)

Harrington 2013

Comparison: TKI

(CRT + lapatinib vs. CRT + placebo)

Gp A n = 34; Gp B n = 33

Diarrhoea

All grades: n = 16 (46%);

Grade 3: n = 2 (6%)

All grades: n = 1 (3%);

Grade 3: n = 0 (0%)

Rash

All grades: n = 10 (29%);

Grade 3: n = 3 (9%)

All grades: n = 5 (16%);

Grade 3: n = 1 (3%)

Other skin reactions

All grades: n = 15 (43%);

Grade 3: n = 2 (6%)

All grades: n = 7 (23%);

Grade 3: n = 5 (16%)

Martins 2013
Comparison: TKI (CRT + erlotinib vs.
CRT (no placebo))

(Gp A n = 95; Gp B n = 96)

Pain

All grades: n = 50 (53%);

Grades 3‐4: n = 18 (19%)

All grades: n = 54 (56%);

Grades 3‐4: n = 18 (19%)

Gastrointestinal

Grades 3‐4: n = 46 (48%)

Grades 3‐4: n = 41 (43%)

Rash

All grades: n = 65 (68%);

Grade 3: n = 12 (13%)

All grades: n = 10 (10%);

Grade 3: n = 2 (2%)

Serious adverse events*

n = 38 (40%)

n = 32 (33%)

Haematological

Grades 3‐4: n = 15 (16%)

Grades 3‐4: n = 25 (26%)

Metabolic

Grades 3‐4: n = 7 (7%)

Grades 3‐4: n = 5 (5%)

Singh 2013

Comparison TKI

RT + gefitinib vs. RT (no placebo)

(Gp A n = 30; Gp B n = 30)

Mucositis

All grades: n = 30 (100%);

Grades 3‐4: n = 21 (70%)

All grades: n = 30 (100%);

Grades 3‐4: n = 19 (63%)

Skin reaction

All grades: n = 30 (100%);

Grades 3‐4: n = 12 (40%)

All grades: n = 30 (100%);

Grades 3‐4: n = 11 (36%)

Methods

Location of trial: international (9 countries)

Number of centres: unclear

Trial ID: NCT00387127, EudraCT#: 2005‐003767‐23

Recruitment period: November 2006 to January 2009

Funding source: GlaxoSmithKline

Participants

Inclusion criteria: eligible patients had histologically confirmed, unresected stage III/IVA/IVB HNSCC of at least 1 of the following sites: OC, oropharynx, hypopharynx or larynx. Other inclusion criteria included: European Cooperative Oncology Group Performance Status 0, 1 or 2; adequate renal, hepatic and haematological function and normal left ventricular ejection fraction

Exclusion criteria: patients with T1N1, T2N1 or metastatic disease and patients who had received any prior or current treatment for invasive HNSCC were excluded. Initially, only patients with EGFR over‐expression (3+ by immunohistochemistry) were included. This inclusion criterion was later amended to allow inclusion of all participants regardless of EGFR status

Sex (M/F): 90%/10%

Age at baseline (years): Gp A: range 44‐66, median 56; Gp B: range 34‐69, median 57

OC: 5 (7%)

OP: 43 (64%)

OC + OP: 48

Number randomised: 67

Number evaluated: 67

Interventions

Comparison: lapatinib + CRT vs. CRT alone

Gp A (n = 34): concomitant lapatinib 1500 mg/d + CRT + lapatinib only in maintenance phase

Gp B (n = 33): CRT + placebo + placebo during maintenance phase

Both groups received CRT comprising cisplatin 100 mg/m² on d 1, 22 and 43 of the RT course. RT was delivered once daily (dose < 2.5 Gy/fraction to a total dose of 65 Gy (IMRT) or 70 Gy (2D or 3D RT) over 6.5‐7 wk to the gross site with 50 Gy to elective nodal areas

Duration of follow‐up: 40 months

Outcomes

Primary: CRR at 6 months

Secondary: PFS, OS, LRC, distant relapse, biomarkers and safety

Notes

Small, short study. Primary outcome CRR at 6 months (to indicate whether further trials are justified) and likely insufficient power to detect a clinically important benefit of lapatinib if this is present

Adverse events: included in Table 10 10

Adverse events note: selectively reported numerical data only provided for adverse events not comparable in incidence during concomitant phase. All adverse events otherwise presented graphically, and data not reliably extractable

Toxicity evaluation assessment tool: not reported

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Phase II randomised controlled trial sponsored by GlaxoSmithKline. Method of sequence generation not clearly reported but randomisation stratified by nodal status (N0‐1 vs. N2‐3) and tumour location

Allocation concealment (selection bias)

Low risk

Central randomisation schedule; however, concealment not specifically reported. There were nine centres so it is likely that the randomisation was done centrally by GSK.

Blinding of participants (performance bias)

Low risk

Double blind ‐ placebo used

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

"All scans and clinical data for the primary end‐point were reviewed centrally by an independent review board (BioClinica, Newtown, PA, USA), and all readers were blinded to treatment"

Incomplete outcome data (attrition bias)
All outcomes

Low risk

ITT analysis for efficacy

Selective reporting (reporting bias)

Unclear risk

Planned clinical outcomes reported in full; adverse events selectively reported numerically (all otherwise available graphically)

Other bias

Unclear risk

GSK who funded the study were involved in the preparation of the trial report.

Methods

Location of trial: Korea

Number of centres: multicentre, Seoul National University Hospital; Clinical Research Center for Solid Tumor [sic]

Trial ID: NCT0623558; CRSCT‐L0002

Recruitment period: not reported. April 2008‐?

Funding source: Merck KGaA Darmstadt, Germany

Participants

Inclusion criteria: > 18 years old; either sex; unresectable LN‐HNSCC due to either tumour fixation, involvement of basal skull or lymph node fixation or LA‐HNSCC with low surgical curability on the basis of advanced disease (T3‐T4) or regional lymph node extension (N2‐3, except for T1N2), or LA‐HNSCC, which is a candidate for organ preservation (primary tumour to be in OC, oropharynx, hypopharynx or larynx); ECOG performance status 0 or 1; informed consent; ALT/AST < 2.5 x ULN; serum albumin level > 3.0 g/dL; serum ALP < 2.5 x ULN; bilirubin < 1.5 mg/dL; serum creatinine < 1.5 x ULN; WBC > 3000/mm³; absolute neutrophil count > 1500/mm³; haemoglobin > 9 g/dL

Exclusion criteria: previous chemotherapy or RT for LA‐HNSCC; targeted‐lesions RT with 6 months; previous EGFR pathway‐targeting therapy; prior cancer surgery (excluding diagnostic biopsy in 4 wk prior to recruitment; distant metastatic disease; heart failure/coronary artery disease/myocardial infarction in prior 6 months; known allergy to study treatment; pregnancy/lactation; investigational agent receipt in prior 28 d; previous malignancy in prior 5 years (excluding adequately treated in‐situ cervical cancer/non‐melanoma skin cancer; legal incapacity; diagnosis of another cancer within 5 years, peripheral neuropathy or hearing disorder of > grade 2 by NCI‐CTCAE, co‐morbidity or biological disorder contraindicating administration of systemic chemotherapy or RT

Sex (M/F): overall: 82/10; Gp A: 46/2; Gp B: 36/8

Age at baseline (years): overall: median 59, range 29‐73; Gp A: median 58, range 40‐73; Gp B: median 61, range 29‐73

OC: overall: n = 14 (15%); Gp A: n = 4 (8%); Gp B: n = 10 (23%)
OP: overall: n = 41 (45%); Gp A: n = 25 (52%); Gp B: n = 16 (36%)
OC + OP: overall: n = 55 (60%); Gp A: n = 29 (60%); Gp B: n = 26 (59%)

Number randomised: 92 (Gp A: n = 48; Gp B: n = 44)
Number evaluated: 92

Interventions

Comparison: RT + docetaxel + cisplatin + cetuximab vs. RT + docetaxel + cisplatin

Gp A (n = 48): RT (weekly for 9 wk) + docetaxel (3 x d 1 of 3 wk cycles; 75 mg/m²) + cisplatin (3 x d 1 of 3 wk cycles; 75 mg/m²; followed by weekly 30 mg/m² for 9 wk) + cetuximab (weekly for 9 wk; first weekly dose 400 mg/m²; followed by weekly 250 mg/m² for 9 wk)
Gp B (n = 44): RT (weekly for 9 wk) + docetaxel (3 x d 1 of 3 wk cycles; 75 mg/m²) + cisplatin (3 x d 1 of 3 wk cycles; 75 mg/m²; followed by weekly 30 mg/m² for 9 wk)
Duration of follow‐up: 60 months

Outcomes

Primary: ORR

Secondary: completion rate of induction and CCRT, PFS, OS and toxicities

Notes

Adverse events: published as conference poster although not legible from PDF copy

Socioeconomic status: not reported, full study report currently unpublished other than abstract and poster

Sample size calculation: not reported.

Method of analysis: ITT

Lost‐to‐follow‐up: incompletion rationale provided only for Gp A (n = 8), not indicated for Gp B (n = 4)
Gp A: hypersensitivity n = 1; septic shock n = 1; skin rash n = 1; seizure n = 1; arterial thrombosis n = 1; unexplained death n = 1; unsatisfactory response n = 1; informed consent withdrawn n = 1

Adverse events note: toxicity data split by CRT induction or concurrent CRT

Selected reported adverse events: most frequently reported (top 6) included in Table 8 8: for full listing of reported adverse effects, see conference poster

Toxicity evaluation assessment tool: National Cancer Institute Common Terminology Criteria for Adverse Events (version 3)

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Central computerised randomisation.

Allocation concealment (selection bias)

Low risk

Not reported

Blinding of participants (performance bias)

Unclear risk

Not reported, no placebo used. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms; however, primary outcomes considered to be objective

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Not reported

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Loss to follow‐up rationale not reported for Gp B

Selective reporting (reporting bias)

Low risk

Planned outcomes reported

Other bias

Low risk

Compliance (completion of therapy) reported: Gp A: n = 40; Gp B: n = 40

Methods

Location of trial: Italy

Number of centres: multicentre (number of centres unclear)

Trial ID: not reported

Recruitment period: August 1995 to August 1996

Funding source: CNR Rome, AP "clinical applications of oncological research" contract 96.00588.PF39

Participants

Inclusion criteria: adults aged 18‐75 years, with histologically confirmed technically resectable or unresectable stage III‐IV HNSCC, with measurable disease, ECOG performance status 0‐1 and a life expectancy of at least 6 months, WBC count > 4000/ml³, platelet count > 100,000/ml³, serum bilirubin < 2 mg/dL, creatinine clearance > 60 mL/minute. Participants must have adequate nutritional and liquid supply, at least a 4‐wk interval from whatever surgery and 2 wk from the taking of biopsy specimen

Exclusion criteria: previous treatment for the actual disease, any infection requiring antibiotic treatment, distant metastases, a history or the presence of a second malignancy except for basal cell carcinoma of the skin and cervical carcinoma in situ. Serious hearing impairment, clinical evidence of congestive heart failure grade III‐IV (New York Heart Classification), drug‐uncontrolled angina pectoris, drug‐uncontrolled clinically significant cardiac arrhythmia, drug or chronic alcohol addiction, women of childbearing age without a negative pregnancy test, previous or present hypercalcaemia

Sex (M/F): 29/4

Age at baseline (years): mean 54.5, range 38‐72

OC: Gp A: n = 2; Gp B: n = 2

OP: Gp A: n = 6; Gp B: n = 9

OC + OP: total n = 19

Number randomised: 33

Number evaluated: 28 (3 participants refused treatment, 2 died during treatment)

Interventions

Comparison: rIL‐2 + chemotherapy vs. chemotherapy alone

Gp A (n = 16): 3 cycles of chemotherapy (Al Sarraf regimen) + 4.5 MIU/d rIL‐2, subcutaneously on d 8‐12 and 15‐19 of chemotherapy, repeated 3 times for each cycle

Gp B (n = 17): 3 cycles of chemotherapy (Al Sarraf regimen)

Both groups received the same chemotherapy regimen; 100 mg/m² cisplatin IV as a 60‐minute infusion on d 1, with a standard pre‐ and post‐hydration protocol with forced diuresis by 250 mL 18% mannitol, plus 1000 mg/m²/d 5‐fluorouracil on d 1‐5 (120 h) as a continuous infusion by peripheral vein. 3 cycles planned unless disease progression or unacceptable toxicity necessitated a break

Outcomes

Primary: objective tumour regression (complete or partial), OS and time to disease progression

Secondary: toxicity

Duration of follow‐up: > 18 months

Notes

The sample size necessary to demonstrate a 20% difference between the 2 regimens at the significance level of α = 0.05 with a power of β = 0.80 was calculated to be 95 participants per arm

An interim analysis, carried out on the first 30 participants, is presented in this study

Selected reported adverse events: most frequently reported (top 6) included in Table 9 9: for full listing of reported adverse effects, see original study reference (Table 4)

Toxicity evaluation assessment tool: WHO Recommendations for Grading of Acute and Subacute Toxicity (1981)

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Randomly assigned". Centralised. Method of sequence generation not described

Allocation concealment (selection bias)

Low risk

"The randomisation was centralized in the Division of Medical Oncology, S. Giovanni Hospital, Turin." Concealment not reported specifically

Blinding of participants (performance bias)

Unclear risk

Open label, no placebo. Knowledge of allocated treatment may have influenced participants' care receipt.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Blinding of outcome assessment not reported, but used a centralised system.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Attrition was very high ‐ although rationale was provided and a balance of participants lost‐to‐follow‐up reported across arms

Selective reporting (reporting bias)

Low risk

Planned outcomes of response to treatment and toxicity reported in full. OS and time to disease progression also mentioned

Other bias

Low risk

No other sources of bias identified

Methods

Location of trial: US

Number of centres: 10 centres

Trial ID: NCT00410826

Recruitment period: December 2006 to October 2011

Funding source: supported by University of Washington and grants from Genentech and Astellas Pharma Global Development

Participants

Inclusion criteria: biopsy‐confirmed stage III, IVA or IVB HNSCC with identified primary tumour in the nasopharynx (WHO type I), OC, oropharynx, hypopharynx or larynx; ECOG performance status of 0‐2, calculated creatinine clearance > 55 mL/minute; adequate liver function; platelet count > 100,000/dL; and neutrophil count > 1250/dL

Exclusion criteria: prior chemotherapy or EGFR‐directed therapy, prior attempt at complete surgical resection, prior RT to the head and neck, prior history of HNSCC, pregnant or breastfeeding women, people with cardiac history precluding hydration required for cisplatin. Prior second malignancy was allowed if person was without evidence of disease

Sex (M/F): 175/28

Age at baseline: not stated

OC: 15 (7%)

OP: 137 (67%)

OC + OP: 152 (75%)

Number randomised: 204

Number evaluated: 204

Interventions

Comparison: erlotinib + CRT vs. CRT alone

Gp A (n = 99): erlotinib 150 mg/d (orally or through endogastric tube), started 1 wk prior to CRT, and continued throughout

Gp B (n = 105): CRT only, no placebo used

Both groups received cisplatin 100 mg/m² on d 1, 22 and 43 given with hydration. Cisplatin was only administered during RT, which was either 66‐70 Gy in 30‐35 fractions (either iIMRT or 3D) of 2‐2.2 Gy/fraction, 5 x weekly. Sites of potential/occult disease received 50 Gy

Outcomes

Primary: CR (definitions provided)

Secondary: PFS, OS, adverse effects

Duration of follow‐up: 60 months

Notes

Sample size calculation: the sample size of 204 randomly assigned participants was estimated to have 80% power to detect a difference between a 60% CRR in arm B and a 40% CRR in arm A, by a 2‐sided test of independence with 5% type I error, assuming a 5% drop‐off rate

Adverse events notes: partial reporting of toxicities by grade

Serious adverse events*: bespoke category of "Serious adverse events (SAEs)" reported without explanation of criteria/grades included ‐ not originating from National Cancer Institute assessment tool

Selected reported adverse events: most frequently reported (top 6) included in Table 10 10: for full listing of reported adverse effects, see original study reference (Table 2)

Toxicity evaluation assessment tool: National Cancer Institute Common Terminology Criteria for Adverse Events (version 3)

Adverse events analysis method: per‐protocol

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Random assignment was stratified by centre, performance status (ECOG 0/1 vs 2), and degree of nodal involvement (N0/1 vs N2/3)." Method of sequence generation not reported

Allocation concealment (selection bias)

Low risk

Central allocation. Concealment not specifically reported though

Blinding of participants (performance bias)

Unclear risk

Open label, no placebo used. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms; however, primary outcomes considered to be objective

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Not reported for assessors. Refers to "Central review of all clinical and pathological outcomes was conducted without knowledge of treatment group"

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Numbers and reasons for withdrawals clearly described and efficacy analysis by ITT

Selective reporting (reporting bias)

Unclear risk

Planned outcomes of this Phase II trial reported; partial reporting of adverse events

Other bias

Low risk

No other sources of bias identified

Methods

Location of trial: India

Number of centres: 3 centres

Trial ID: h‐R3/SCCHN/001/IND

Recruitment period: September 2004‐July 2005

Funding source: Biocon Ltd. (Indian pharmaceutical company, Bangalore, India)

Participants

Inclusion criteria: aged 18‐70 years; histologically confirmed stage III or IVA (T1‐T4a, N0‐N2) HNSCC; suitable for concurrent CRT/RT; KPS of P60%; life expectancy of > 6 months; adequate haematological function (WBC count > 4000/mm³; absolute neutrophil count 1500/mm³; platelets > 100,000/mm³; total bilirubin 61.2 mg/dL; AST; ALT 62.5 x normal limit (37 and 40 U/L, respectively); serum creatinine < 1.4 mg/dL); people contraindicated for surgery owing to prohibitive morbidity or compromised quality of life

Exclusion criteria: nasopharyngeal carcinoma; history of malignancy other than non‐melanoma skin cancer or carcinoma‐in situ of the cervix; evidence of distant metastases/concurrent secondary malignancy or T4b lesion; chemotherapy within 3 months before enrolment; prior RT to the head and neck; prior immunotherapy; increased risk of lethal infections and pharmacokinetic interactions with nimotuzumab; history of allergy with similar biological compounds; pregnant/lactating women; people on other investigational drugs/devices

Age at baseline (years): Gp A: mean 49.9 (SD 10.6); Gp B: mean 53.7 (SD 9.0); Gp C: mean 58.7 (SD 8.2); Gp D: mean 58.7 (SD 8.1) (P value not reported)

Sex (M/F): overall: 82/10 (89%/11%); Gp A: 20/3; Gp B: 21/2; Gp C: 21/2; Gp D: 20/3 (P value not reported)

OC: overall: 19.6% (n = 18); Gp A: 22% (n = 5); Gp B: 17% (n = 4); Gp C: 13% (n = 3); Gp D: 26% (n = 6) (P value not reported)
OP: overall: 54.3% (n = 50); Gp A: 44% (n = 10); Gp B: 52% (n = 12); Gp C: 78% (n = 18); Gp D: 44% (n = 10) (P value not reported)

Number randomised: 92

Number evaluated: 92 (n = 68 (74%) attrition). Gp A: lost to follow‐up n = 4, died n = 10; Gp B: lost to follow‐up n = 2, died n = 17; Gp C: withdrew (serious adverse events) = 1, died n = 14; Gp D: lost to follow‐up n = 3, died n = 17)

Analysis: ITT

Interventions

Comparison: RT + cisplatin + nimotuzumab vs. RT + cisplatin and RT + nimotuzumab vs. RT

Gp A: (n = 23) RT (60‐66 Gy; 2 Gy/fraction/d; 5 d/wk for 6‐6.5 wk) + cisplatin (50 mg dilution in 1 L saline; by IV over 2 h; 1 d/wk for 6 wk) + nimotuzumab (4 x 50 mg dilution in 0.9% saline; by IV over 1 h; 1 d/wk for 6 wk)

Gp B: (n = 23) RT (60‐66 Gy; 2 Gy/fraction/d; 5 d/wk for 6‐6.5 wk) + cisplatin (50 mg dilution in 1 L saline; by IV over 2 h; 1 d/wk for 6 wk)

Gp C: (n = 23) RT (60‐66 Gy; 2 Gy/fraction/d; 5 d/wk for 6‐6.5 wk) + nimotuzumab (4 x 50 mg dilution in 0.9% saline; by IV over 1 h; 1 d/wk for 6 wk)

Gp D: (n = 23) RT (60‐66 Gy; 2 Gy/fraction/d; 5 d/wk for 6‐6.5 wk)

Duration of follow‐up: 60 months

Outcomes

OS, PFS, adverse events

Duration of follow‐up: 60 months

Notes

Sample size calculation: Quote "the sample size was 17 patients per arm, calculated by assuming P < 0.05 (5%), β < 0.2 (20%), power = 0.80 (80%), and a confidence interval (CI) of 95%"

Adverse events notes: to compare intervention (adjunctive nimotuzumab) with control (standard therapy: CRT or RT), adverse events presented as follows.

Intervention: Gp A+C (CRT/RT + nimotuzumab) n = 46

Control: Gp B+D (CRT/RT) n = 46

Control arm mucositis events Gp D n = 27 (+4 than Gp n), therefore all grade mucositis data not used in analysis

"Nimotuzumab was administered 3 d after cisplatin in order to identify AEs [adverse events] related to either intervention separately"

Selected reported adverse events: most frequently reported (top 6) included in Table 8 8: for full listing of reported adverse effects, see original study reference's Table 3

Toxicity evaluation assessment tool: Common Toxicity Criteria of the National Cancer Institute of Canada Clinical Trials Group (NCIC CTG)

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Each group was further randomised in a 1:1 fashion ‐ Group I into CRT + nimotuzumab or CRT and Group II into RT + nimotuzumab or RT." Method not reported

Allocation concealment (selection bias)

Unclear risk

Not reported

Blinding of participants (performance bias)

Unclear risk

Open label, no placebo control. Knowledge of allocated treatment may have influenced patients' care receipt and perception of symptoms.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Not reported

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Attrition explained, and data analysed as ITT

Selective reporting (reporting bias)

High risk

All outcomes reported, including adverse events; however, derived HR values from graphs do not correspond with textually reported HRs. Confidence Intervals and P values also inconsistent with significance statements reported

Other bias

Low risk

No other apparent biases

Methods

Location of trial: Cuba

Number of centres: unclear

Trial ID: not reported

Recruitment period: July 2002 to February 2007

Funding source: not reported

Participants

Inclusion criteria: epithelial tumours of head and neck confirmed by cytological or histological techniques, in stages III or IV (advanced locoregional disease), who were suitable candidates for RT, were recruited. Measurable lesions, aged ≥ 18 years, ECOG performance status ≤ 2, life expectancy > 6 months and normal functioning of the organs and of the bone marrow as defined by absolute neutrophil count ≥ 1.5 x 10⁹/L, platelet count ≥ 100 x 10⁹/L, serum creatinine level ≤ ULN and ALT and AST level < 2.5 x ULN

Exclusion criteria: prior RT or chemotherapy, concurrent active cancer, any uncontrolled intercurrent illness and pregnancy or lactation

Sex (M/F): 81/24

Age at baseline (years): Gp A: mean 59.48, Gp B: mean 65.88

OC: 28 (26%)

OP: 72 (69%)

OC + OP: 100

Number randomised: 106

Number evaluated: 106

Interventions

Comparison: nimotuzumab + RT vs. placebo + RT

Gp A (n = 54): 6 doses of nimotuzumab IV in 250 mL saline + planned RT of 2 Gy/fraction, 5 fractions/wk to total dose of 60‐66 Gy

Gp B (n = 52): 6 doses of placebo IV in 250 mL saline + planned RT of 2 Gy/fraction, 5 fractions/wk to total dose of 60‐66 Gy

No modified fractionation, IMRT or concomitant boost RT was carried out

Outcomes

Primary: CRR

Secondary: OS

Duration of follow‐up: 24 months

Notes

Poorly reported. Information presented for some of the participants format was inconsistent and data were incomplete

Adverse events note: derived numbers do not result in reported percentages (equating instead to incomplete patient integers). Corresponding whole participant numbers used

Selected reported adverse events: narrative and partial reporting of adverse in publication. Included in Table 8 8

Toxicity evaluation assessment tool: Common Toxicity Criteria of the US National Cancer Institute

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Random assignment was performed centrally through a validated simple randomisation system (ASAL) version 1.2

Allocation concealment (selection bias)

Low risk

Random assignment was performed centrally through a validated simple randomisation system (ASAL) version 1.2

Blinding of participants (performance bias)

Low risk

Double‐blinded

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Not reported for assessors. Refers to "the final determination of response was done by a centralized independent committee composed by radiologists, head and neck surgeons and pathologists who performed a blinded revision of the investigators generated data" ‐ does not indicate if original assessors were blinded

Incomplete outcome data (attrition bias)
All outcomes

Low risk

ITT analysis

Selective reporting (reporting bias)

Low risk

All outcomes reported

Other bias

High risk

The quality of the report was poor and we were unable to eliminate the presence of possible other biases. May be due to translation into English. There was incomplete reporting of some data e.g. EGFR expression, and why some participants had biopsies and some did not.

Methods

Location of trial: India

Number of centres: 1

Trial ID: not given

Recruitment period: December 2008 to August 2010

Funding source: not given

Participants

Inclusion criteria: adult patients aged 40‐65 years, male or female, previously untreated, with measurable and evaluable disease, irrespective of EGFR status, KPS score > 70, biopsy‐confirmed SCC, locally advance OC cancer, normal haematology parameters, renal function and liver function tests normal before recruitment

Exclusion criteria: Patients who were previously treated with either chemotherapy or RT

Sex (M/F): overall: 45/15; Gp A: 25/5; Gp B: 23/7

Age at baseline (years): Gp A: median 55; Gp B: median 53

Gp A: alveolus 3, buccal mucosa 13, tongue 7, retromolar trigone 5, hard palate 2

Gp B: alveolus 3, buccal mucosa 15, tongue 11, retromolar trigone 1, hard palate 0

Number randomised: 60 (Gp A 30, Gp B 30)

Number evaluated: 60

Interventions

Comparison: gefitinib + RT vs. RT alone

Gp A: RT (70 Gy over 7 wk) + gefitinib (250 mg orally daily for 7 d before starting RT and continued up to 90 d)

Gp B: RT (70 Gy over 7 wk) alone (no placebo)

Outcomes

Primary: CR, PR, no response and disease progression

Secondary: grade of mucositis, skin reaction, haematological toxicity, incidence and grade of diarrhoea and vomiting

20‐month follow‐up shown in survival data but no follow‐up time stated in text

Notes

Very poorly written paper. Incomplete outcome data, selective reporting. No calculation of HR

Adverse events note: quote: "The acute toxicities in both the arms are not significantly different. [...] The common side effects expected due to Gefitinib like diarrhoea, nausea, and vomiting were not significantly increased in the concurrent Gefitinib arm compared with the RT only arm. There is no significant difference in hematological, hepatic, and renal toxicity"

Selected reported adverse events: narrative and partial reporting of adverse events in publication. Mucositis and skin reactions presented in Tables 3 and 4 in original study reference. Included in Table 10 10

Toxicity evaluation assessment tool:
Acute systemic toxicity: WHO Recommendations for Grading of Acute and Subacute Toxicity (1981)

Acute and late radiation reactions: toxicity criteria of the Radiation Therapy Oncology Group (1995)

Adverse events analysis method: ITT

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

The method of randomisation is not given, the authors only stated, "A total of 60 patients were randomised by simple randomisation to two treatment arms." The term 'imbalanced randomisation' was used but no explanation of this term was given

Allocation concealment (selection bias)

Unclear risk

Concealment not specifically reported

Blinding of participants (performance bias)

Unclear risk

No placebo was used so participants could not be blinded. Knowledge of allocated treatment may have influenced participants' care receipt and perception of symptoms.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

"The participants and the outcome adjudicators were blinded about allocation to treatment arms"

Incomplete outcome data (attrition bias)
All outcomes

High risk

Incomplete information about number of participants included in outcome assessment. Numbers and reasons for withdrawals not given

Selective reporting (reporting bias)

High risk

Some outcomes only partially reported e.g. adverse events

Other bias

Unclear risk

The quality of the report was poor and we were unable to eliminate the presence of possible other biases