Matched pair analysis for comparison of survival outcome of alternative regimens to standard three-weekly cisplatin-based concurrent chemoradiation of head and neck cancer.

BACKGROUND: To compare head and neck cancer (HNC) patients treated with three-weekly versus weekly cisplatin-based or other chemotherapy-based concurrent chemoradiation (CRT) and CRT with versus without induction chemotherapy (ICT) to investigate differences in overall survival (OS) and cancer-specific survival (CSS).
METHODS: HNC patients treated with definitive or adjuvant CRT at Roswell Park Comprehensive Cancer Center between 2003 and 2017 were retrospectively reviewed. Propensity score matching was performed to obtain three sets of balanced matched pairs: three-weekly and weekly cisplatin CRT, three weekly and non-cisplatin CRT, CRT with and without ICT. Multivariate Cox regression and Kaplan-Meier analyses were used to estimate and compare survival outcomes.
RESULTS: A total of 623 patients received either definitive (81%) or post-operative (19%) RT. Of these, 283 patients concurrently received three-weekly cisplatin (45%); 189 patients (30%) received weekly cisplatin; 151 patients (24%) received non-cisplatin regimen. Median follow-up was 55.4 months (interquartile range, 38.0-88.7). Patients who received CRT alone and those who received ICT and CRT had no difference in 5-year OS (51.5% and 41.0% respectively, P=0.53) and CSS (64.9% and 49.7% respectively, P=0.21). Compared to patients who received three-weekly cisplatin, patients who received weekly cisplatin had no difference in 5-year OS (59.3% vs. 54.1%, P=0.35) and CSS (70.3% vs. 62.4%, P=0.09); patients who received non-cisplatin CRT also had no difference in 5-year OS (54.5% vs. 58.3%, P=0.51) and CSS (67.5% vs. 64.7%, P=0.45).
CONCLUSIONS: No significant difference in OS and CSS was observed in any of the three pairs of CRT regimens. ICT prior to CRT did not improve survival of CRT alone. Non-cisplatin and weekly cisplatin regimens did not prove to be inferior to the standard three-weekly cisplatin. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

Concurrent chemoradiation (CRT) with high-dose cisplatin (HDC) (three cycles of 100 mg/m2 given once every 3 weeks) produces better locoregional control (LRC) and overall survival (OS) compared to radiotherapy (RT) alone in several randomized trials of head and neck cancer (HNC) (1-7). Compared to RT alone, however, CRT with HDC reports roughly 20–40% greater rate of high-grade toxicity that significantly deters up to 40% of patients from completing the aggressive treatment (3-8).

Low-dose cisplatin (LDC) (30–50 mg/m2 given once weekly) regimens show promising LRC, OS, and cancer-specific survival (CSS) with acceptable rate of severe acute toxicity (9-11). Several retrospective analyses suggest comparable efficacy and improved toxicity profile with LDC (12-15). Three randomized trials, however, report conflicting data that HDC versus LDC (16,17) and LDC versus HDC (18) is superior in the post-operative setting.

Multiple non-cisplatin based regimens have been tested but have not supplanted HDC (19-24). Induction chemotherapy (ICT), despite its efficacy compared to RT alone in larynx preservation (25,26) and reduction of distant metastases (1,4,25,26), does not improve survival compared to CRT alone (27,28).

In this study, we aimed to compare HNC patients treated with (I) ICT and CRT vs. CRT alone, (II) HDC- vs. LDC-based CRT, (III) HDC- vs. other chemotherapy-based CRT to determine whether there is a significant difference in OS and CSS.

We present the following article in accordance with the STROBE reporting checklist (available at http://dx.doi.org/10.21037/atm-20-5032).

Methods

Patient population

An institutional database of HNC patients treated with definitive or post-operative CRT between 2003 and 2017 at Roswell Park Comprehensive Cancer Center was retrospectively reviewed. Patients who received CRT were included regardless of dosing schedule and chemotherapy agent; those who received RT alone or treatment with non-curative intent were excluded. Patients who received ICT followed by CRT were included; those who received ICT alone or ICT with RT were excluded. Length of follow-up, for those still alive, was defined as length of time between date of diagnosis to last date of follow-up visit.

Statistical analysis

Multivariate (MVA) logistic regression analysis was performed using backward selection (α<0.20) of potential confounders to identify patient factors and treatment factors associated with survival. All P values were two-sided and factors with P values ≤0.05 were considered statistically significant. MVA Cox regression analysis was performed to identify factors that are associated with OS and CSS. Kaplan-Meier analysis was used to estimate survival of matched cohorts.

Propensity score matching was performed in patients with (I) CRT with and without ICT, (II) HDC-CRT and LDC-CRT, (III) HDC-CRT and non-cisplatin CRT. Survival outcomes were compared. Baseline characteristics, including age, gender, pre-RT weight, smoking status, p16 status, tumor staging, primary tumor site, and treatments received were matched to construct well-balanced pairs. Propensity score matching was performed using the nearest neighbor matching without replacement method in 1:1 ratio with a caliper width of 0.1 of the standard deviation of the logit (29). SAS (SAS Institute, Cary, NC, USA) and R (version 3.6.1, R Project for Statistical Computing, Vienna, Austria) software were used.

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the institutional review board of Roswell Park Comprehensive Cancer Center (EDR-103707) and individual consent for this retrospective analysis was waived.

Results

Baseline characteristics

We analyzed a total of 623 patients, of whom 506 were males (81%) and 117 were females (19%) with a median age at time of diagnosis of 61 years [interquartile range (IQR), 50.4–66.6]. About 95% [593] of these HNC patients had squamous cell carcinoma. The most common site of primary tumor was oropharynx (42%). Median follow-up was 55.4 months (IQR, 38.0–88.7 months). All patients received either definitive (81%) or post-operative (19%) RT, median dose 70 Gy (IQR, 70–70 Gy) for all cohorts.

Standard regimen of three-weekly HDC was concomitantly used to treat 283 patients (45%); weekly LDC (30–50 mg/m2 once weekly) was given to 189 patients (30%); 151 patients (24%) received other chemotherapy regimen which included weekly cetuximab, weekly carboplatin, platinum regimen not otherwise specified, and crossover to carboplatin or cetuximab. Total or mean cumulative dose in each cohort is unknown due to incomplete information in the database. Prior to matching, median OS for HDC cohort was 40.3 months (IQR, 26.2–63.5), LDC was 42.4 months (IQR, 19.8–78.4), and other chemotherapy was 37.9 months (IQR, 13.8–69.2). Taxane-based ICT was given to 70 patients (11%). Median OS before matching for this cohort was 34.6 months (IQR, 16.5–65.3). The baseline patient and treatment characteristics before matching are summarized in .

Table 1

Baseline characteristics before matching

Variable	Cis q3wk			Cis q1wk			Other		P
	N	%		N	%		N	%
Gender									0.29
Male	234	83		156	83		116	77
Female	49	17		33	17		35	23
Total	283	100		189	100		151	100
Age (yrs)									<0.001
<61	173	61		91	48		60	40
≥61	110	39		98	52		91	60
Total	283	100		189	100		151	100
Smoker									0.008
Never	71	25		37	20		26	17
Former	147	52		87	46		93	62
Current	65	23		65	34		32	21
Total	283	100		189	100		151	100
HPV									0.007
Negative	47	17		50	26		33	22
Positive	121	43		59	31		42	28
NA	115	41		80	42		76	50
Total	283	100		189	100		151	100
Comorbidity (no.)									0.05
0	63	22		42	22		17	11
1	87	31		51	27		41	27
2	69	24		48	25		44	29
3+	64	23		48	25		49	32
Total	283	100		189	100		151	100
T stage									<0.001
X	1	0		0	0		2	1
0-2	145	51		79	42		68	45
3-4	136	48		109	58		71	47
NA	1	0		1	1		10	7
Total	283	100		189	100		151	100
N stage									<0.001
0-1	78	28		61	32		67	44
2-3	203	72		127	67		74	49
NA	2	1		1	1		10	7
Total	283	100		189	100		151	100
M stage									<0.001
0	277	98		182	96		135	89
1	5	2		4	2		5	3
NA	1	0		3	2		11	7
Total	283	100		189	100		151	100
Primary site									0.03
NA	23	8		27	14		26	17
Oral cavity	25	9		22	12		9	6
Nasopharynx	14	5		2	1		2	1
Oropharynx	130	46		70	37		63	42
Hypopharynx	15	5		14	7		9	6
Glottis	33	12		21	11		19	13
Salivary	5	2		3	2		2	1
Other	0	0		3	2		5	3
Unknown	22	8		15	8		9	6
Multiple	16	6		12	6		7	5
Total	283	100		189	100		151	100
Histology									<0.001
Squamous	278	98		182	96		133	88
Other	5	2		7	4		18	12
Total	283	100		189	100		151	100
RT type									0.63
Definitive	230	81		149	79		125	83
Adjuvant	53	19		40	21		26	17
Total	283	100		189	100		151	100
RT total dose (Gy)									0.15
Median	70			70			70
IQR	70–70			70–70			70–70
RT start year									0.01
<2011	72	25		50	26		58	38
≥2011	211	75		139	74		93	62
Total	283	100		189	100		151	100
RT complete									0.02
No	3	1		10	5		8	5
Yes	278	98		178	94		141	93
NA	2	1		1	1		2	1
Total	283	100		189	100		151	100
Treatment response									<0.001
No response	6	2		7	4		11	7
Partial	237	84		142	75		92	61
Complete	23	8		22	12		30	20
NA	17	6		18	10		18	12
Total	283	100		189	100		151	100
Surgery									0.51
No	225	80		147	78		125	83
Yes	58	20		42	22		26	17
Total	283	100		189	100		151	100
Induction chemo									0
No	264	93		156	83		133	88
Yes	19	7		33	17		18	12
Total	283	100		189	100		151	100
Nutrition support									0.07
No	122	43		62	33		55	36
Yes	161	57		127	67		96	64
Total	283	100		189	100		151	100
Hospitalized									0.97
No	215	76		142	75		112	74
Yes	67	24		46	24		39	26
NA	1	0		1	1		0	0
Total	283	100		189	100		151	100
WBC count									0.11
Normal	242	86		153	81		113	75
Low	5	2		5	3		5	3
High	25	9		20	11		17	11
NA	11	4		11	6		16	11
Total	283	100		189	100		151	100
Hemoglobin (g/dL)									<0.001
≥12	229	81		132	70		86	57
<12	43	15		46	24		49	32
NA	11	4		11	6		16	11
Total	283	100		189	100		151	100

Cis, cisplatin; q3wk, once every 3 weeks; Q1wk, once every week; HPV, human papilloma virus; NA, not available; RT, radiotherapy; IQR, interquartile range; Chemo, chemotherapy; WBC, white blood cell.

Survival outcome

MVA showed no significant association between alternative chemotherapy regimen and survival. In comparison to standard HDC dosing, neither non-cisplatin nor weekly LDC was associated with any significant change in OS [hazard ratio (HR) 0.95, 95% confidence interval (CI), 0.70–1.28, P=0.72] or CSS (HR 1.14, 95% CI, 0.79–1.65, P=0.48). ICT was also not associated with any significant change in OS (HR 1.28, 95% CI, 0.89–1.82, P=0.18) or CSS (HR 1.41, 95% CI, 0.93–2.13, P=0.10). The results of MVA on survival outcome are organized in .

Table 2

Cox regression analysis of survival outcome

Variable	Overall survival								Cancer-specific survival
	UVA				MVA				UVA				MVA
	HR	95% CI	P		HR	95% CI	P		HR	95% CI	P		HR	95% CI	P
Chemo
Cis q3wk	1	Ref			1	Ref			1	Ref			1	Ref
Cis q1wk	1.36	1.03–1.79	0.03		0.97	0.72–1.30	0.82		1.69	1.21–2.36	0.002		1.17	0.81–1.70	0.4
Others	1.79	1.36–2.37	<0.001		1.04	0.75–1.42	0.83		2.03	1.44–2.87	<0.001		1.05	0.70–1.57	0.81
ICT
No	1	Ref			1	Ref			1	Ref			1	Ref
Yes	1.54	1.12–2.12	0.007		1.29	0.90–1.85	0.17		1.8	1.25–2.59	0.002		1.46	0.96–2.23	0.07
Gender
Male	1	Ref							1	Ref
Female	1.13	0.85–1.50	0.41						1.08	0.76–1.53	0.66
Age (yrs)
<61	1	Ref			1	Ref			1	Ref			1	Ref
≥61	1.57	1.24–1.97	<0.001		1.56	1.22–1.99	<0.001		1.57	1.19–2.08	0.001		1.68	1.25–2.26	<0.001
Smoker
Never	1	Ref			1	Ref			1	Ref			1	Ref
Former	1.92	1.36–2.72	<0.001		1.51	1.04–2.18	0.03		1.91	1.25–2.91	0.003		1.31	0.84–2.06	0.23
Current	2.53	1.75–3.67	<0.001		2.23	1.51–3.30	<0.001		2.49	1.59–3.90	<0.001		1.98	1.23–3.17	0.005
HPV
Negative	1	Ref			1	Ref			1	Ref			1	Ref
Positive	0.51	0.37–0.71	<0.001		0.92	0.65–1.30	0.63		0.58	0.39–0.86	0.006		1.2	0.79–1.82	0.39
NA
Comorbidity (no.)
0	1	Ref							1	Ref
1	0.8	0.56–1.13	0.21						0.74	0.49–1.12	0.15
2	1.1	0.78–1.55	0.59						0.94	0.63–1.41	0.78
3+	1.23	0.88–1.73	0.23						0.99	0.66–1.48	0.97
T stage
X	2.43	0.34–17.43	0.38						3.75	0.52–27.16	0.19
0–2	1	Ref			1	Ref			1	Ref			1	Ref
3–4	2.32	1.81–2.97	<0.001		1.87	1.44–2.41	<0.001		2.95	2.15–4.05	<0.001		2.33	1.69–3.21	<0.001
N stage
0–1	1	Ref							1	Ref
2–3	0.95	0.74–1.22	0.69						1.14	0.84–1.55	0.41
NA
M stage
0	1	Ref			1	Ref			1	Ref			1	Ref
1	3.89	2.17–6.95	<0.001		1.02	0.49–2.14	0.96		3.87	1.98–7.59	<0.001		0.77	0.35–1.70	0.52
NA
Histology
Squamous	1	Ref			1	Ref			1	Ref			1	Ref
Other	2	1.31–3.06	0.001		0.87	0.46–1.63	0.66		2.3	1.42–3.74	<0.001		0.98	0.43–2.21	0.96
Primary site
NA	1	Ref			1	Ref			1	Ref			1	Ref
OC	0.84	0.54–1.31	0.44						0.93	0.56–1.54	0.77
NP	0.68	0.32–1.44	0.32						0.77	0.34–1.73	0.53
OP	0.43	0.30–0.61	<0.001		1.13	0.73–1.73	0.59		0.39	0.26–0.59	<0.001		1.08	0.63–1.83	0.78
HP	0.95	0.59–1.54	0.84						0.89	0.50–1.58	0.68
Glottis	0.8	0.52–1.22	0.3						0.68	0.41–1.15	0.15
Salivary	0.78	0.35–1.73	0.54						0.96	0.40–2.27	0.92
Other	0.33	0.08–1.34	0.12						0.46	0.11–1.91	0.29
Unknown	0.39	0.22–0.69	0.001		1.07	0.54–2.12	0.84		0.29	0.13–0.61	0.001		0.86	0.35–2.12	0.75
Multiple	1.15	0.70–1.87	0.59						0.93	0.51–1.70	0.81
RT total dose (Gy)
<70	1	Ref			1	Ref			1	Ref			1	Ref
≥70	0.66	0.50–0.86	0.002		0.82	0.60–1.11	0.2		0.57	0.42–0.78	<0.001		0.72	0.51–1.03	0.07
RT start year
<2011	1	Ref							1	Ref
≥2011	0.91	0.71–1.17	0.49						0.97	0.72–1.31	0.84
RT complete
No	1	Ref			1	Ref			1	Ref			1	Ref
Yes	0.16	0.10–0.25	<0.001		0.53	0.33–0.86	0.01		0.12	0.08–0.20	<0.001		0.44	0.27–0.71	0.001
Response
None	1	Ref			1	Ref			1	Ref			1	Ref
Partial	0.08	0.05–0.12	<0.001		0.08	0.05–0.13	<0.001		0.05	0.03–0.08	<0.001		0.05	0.03–0.08	<0.001
Complete	0.5	0.31–0.81	0.005		0.45	0.27–0.75	0.002		0.46	0.28–0.76	0.002		0.46	0.28–0.77	0.003
Surgery
No	1	Ref							1	Ref
Yes	0.86	0.64–1.16	0.34						0.92	0.65–1.31	0.66
Nutrition support
No	1	Ref			1	Ref			1	Ref
Yes	1.29	1.01–1.65	0.04		0.95	0.72–1.26	0.74		1.29	0.96–1.74	0.09
Hospitalized
No	1	Ref			1	Ref			1	Ref			1	Ref
Yes	1.65	1.28–2.12	<0.001		1.52	1.17–1.98	0.002		1.49	1.09–2.02	0.01		1.31	0.95–1.81	0.1
WBC count
Normal	1	Ref			1	Ref			1	Ref			1	Ref
Low	2.55	1.39–4.68	0.002		1.61	0.85–3.08	0.15		2.94	1.50–5.76	0.002		1.84	0.89–3.80	0.1
High	2.04	1.47–2.82	<0.001		1.27	0.89–1.80	0.18		2.35	1.62–3.41	<0.001		1.36	0.90–2.06	0.14
Hemoglobin (g/dL)
≥12	1	Ref			1	Ref			1	Ref			1	Ref
<12	2.38	1.86–3.06	<0.001		1.36	1.05–1.77	0.02		2.5	1.86–3.36	<0.001		1.05	0.75–1.47	0.78

Chemoradiation with and without ICT

A total of 51 pairs were matched, with all variables well balanced (). Median overall follow-up was 71.7 months (IQR, 43.4–98.4). Median OS was 35.3 months (IQR, 14.3–74.2) and 36.0 months (IQR, 17.7–60.0) for non-ICT and ICT cohorts, respectively (P=0.53). OS at 5 years was 51.5% (95% CI, 39.1–67.7%) for patients who did not receive ICT and 41.0% (95% CI, 28.8–58.5%) for patients who received ICT (P=0.53, ). CSS at 5 years was 64.9% (95% CI, 52.2–80.7%) for non-ICT cohort and 49.7% (95% CI, 36.2–68.3%) for ICT cohort (P=0.21, ).

Table 3

Baseline characteristics after matching

Variable	Cis q3wk			Cis q1wk		P	Cis q3wk			Other		P	No ICT			ICT		P
	N	%		N	%		N	%		N	%		N	%		N	%
Gender						0.89						0.86						0.36
Male	153	84		151	83		74	79		76	81		47	92		43	84
Female	30	16		32	17		20	21		18	19		4	8		8	16
Total	183	100		183	100		94	100		94	100		51	100		51	100
Age (yrs)						0.30						0.88						1
<61	101	55		90	49		46	49		44	47		33	65		33	65
≥61	82	45		93	51		48	51		50	53		18	35		18	35
Total	183	100		183	100		94	100		94	100		51	100		51	100
Smoker						0.40						0.45						0.42
Never	45	25		35	19		16	17		14	15		14	27		8	16
Former	83	45		85	46		63	67		58	62		17	33		20	39
Current	55	30		63	34		15	16		22	23		20	39		23	45
Total	183	100		183	100		94	100		94	100		51	100		51	100
HPV						0.20						0.93						0.82
Negative	37	20		49	27		18	19		20	21		12	24		15	29
Positive	73	40		59	32		30	32		31	33		18	35		17	33
NA	73	40		75	41		46	49		43	46		21	41		19	37
Total	183	100		183	100		94	100		94	100		51	100		51	100
Comorbidity (no.)						0.82						0.89						1
0	41	22		41	22		11	12		10	11		12	24		13	25
1	57	31		49	27		27	29		26	28		17	33		16	31
2	43	23		47	26		30	32		27	29		11	22		10	20
3+	42	23		46	25		26	28		31	33		11	22		12	24
Total	183	100		183	100		94	100		94	100		51	100		51	100
T stage						0.46						0.61						1
X	0	0		0	0		0	0		1	1		0	0		0	0
0–2	89	49		77	42		46	49		50	53		17	33		17	33
3–4	93	51		105	57		47	50		43	46		32	63		33	65
NA	1	1		1	1		1	1		0	0		2	4		1	2
Total	183	100		183	100		94	100		94	100		51	100		51	100
N stage						0.71						0.66						1
0–1	50	27		58	32		42	45		39	41		10	20		11	22
2–3	132	72		124	68		51	54		55	59		39	76		39	76
NA	1	1		1	1		1	1		0	0		2	4		1	2
Total	183	100		183	100		94	100		94	100		51	100		51	100
M stage						1						1						1
0	178	97		178	97		91	97		91	97		47	92		48	94
1	4	2		4	2		2	2		3	3		2	4		2	4
NA	1	1		1	1		1	1		0	0		2	4		1	2
Total	183	100		183	100		94	100		94	100		51	100		51	100
Primary site						0.97						0.99						1
NA	18	10		26	14		11	12		9	10		6	12		7	14
Oral cavity	22	12		21	11		9	10		7	7		2	4		2	4
Nasopharynx	2	1		2	1		3	3		2	2		1	2		2	4
Oropharynx	77	42		70	38		47	50		46	49		24	47		19	37
Hypopharynx	12	7		14	8		4	4		4	4		5	10		7	14
Glottis	20	11		20	11		11	12		13	14		2	4		3	6
Salivary	2	1		3	2		1	1		2	2		1	2		1	2
Other	0	0		0	0		0	0		0	0		1	2		1	2
Unknown	17	9		15	8		5	5		7	7		4	8		4	8
Multiple	13	7		12	7		3	3		4	4		5	10		5	10
Total	183	100		183	100		94	100		94	100		51	100		51	100
Histology						1						1						1
Squamous	178	97		177	97		90	96		91	97		49	96		50	98
Other	5	3		6	3		4	4		3	3		2	4		1	2
Total	183	100		183	100		94	100		94	100		51	100		51	100
RT type						0.90						1						NA
Definitive	147	80		145	79		79	84		79	84		51	100		51	100
Adjuvant	36	20		38	21		15	16		15	16		0	0		0	0
Total	183	100		183	100		94	100		94	100		51	100		51	100
RT total dose (Gy)						0.53						0.83						0.69
Median	70			70			70			70			70			70
IQR	70–70			70–70			70–70			70–70			70–70			70–70
RT start year						1						0.88						0.84
<2011	48	26		47	26		29	31		31	33		25	49		23	45
≥2011	135	74		136	74		65	69		63	67		26	51		28	55
Total	183	100		183	100		94	100		94	100		51	100		51	100
RT complete						0.08						0.62						1
No	2	1		9	5		2	2		1	1		2	4		2	4
Yes	179	98		173	95		91	97		93	99		49	96		49	96
NA	2	1		1	1		1	1		0	0		0	0		0	0
Total	183	100		183	100		94	100		94	100		51	100		51	100
Treatment response						0.65						0.86						0.97
No response	5	3		6	3		6	6		4	4		4	8		3	6
Partial	148	81		138	75		69	73		69	73		35	69		37	73
Complete	17	9		21	11		9	10		12	13		8	16		8	16
NA	13	7		18	10		10	11		9	10		4	8		3	6
Total	183	100		183	100		94	100		94	100		51	100		51	100
Surgery						1						0.71						1
No	142	78		142	78		75	80		78	83		50	98		51	100
Yes	41	22		41	22		19	20		16	17		1	2		0	0
Total	183	100		183	100		94	100		94	100		51	100		51	100
ICT						0.16						1						NA
No	164	90		154	84		85	90		84	89		51	100		0	0
Yes	19	10		29	16		9	10		10	11		0	0		51	100
Total	183	100		183	100		94	100		94	100		51	100		51	100
Chemotherapy						NA						NA						1
Cis q3wk	183	100		0	0		94	100		0	0		18	35		18	35
Cis q1wk	0	0		183	100		0	0		0	0		22	43		22	43
Other	0	0		0	0		0	0		94	100		11	22		11	22
Total	183	100		183	100		94	100		94	100		51	100		51	100
Nut support						0.38						0.46						1
No	68	37		59	32		40	43		34	36		21	41		21	41
Yes	115	63		124	68		54	57		60	64		30	59		30	59
Total	183	100		183	100		94	100		94	100		51	100		51	100
Hospitalized						0.86						0.62						1
No	140	77		136	74		71	76		67	71		45	88		44	86
Yes	42	23		46	25		23	24		27	29		6	12		7	14
NA	1	1		1	1		0	0		0	0		0	0		0	0
Total	183	100		183	100		94	100		94	100		51	100		51	100
WBC count						0.64						0.32						0.79
Normal	157	86		149	81		81	86		76	81		37	73		40	78
Low	1	1		3	2		3	3		2	2		1	2		2	4
High	16	9		20	11		5	5		12	13		8	16		5	10
NA	9	5		11	6		5	5		4	4		5	10		4	8
Total	183	100		183	100		94	100		94	100		51	100		51	100
Hgb (g/dL)						0.37						0.86						0.83
≥12	140	77		128	70		71	76		69	73		28	55		31	61
<12	34	19		44	24		18	19		21	22		18	35		16	31
NA	9	5		11	6		5	5		4	4		5	10		4	8
Total	183	100		183	100		94	100		94	100		51	100		51	100

Cis, cisplatin; q3wk, once every 3 weeks; Q1wk, once every week; IC, induction chemotherapy; HPV, human papilloma virus; NA, not available; RT, radiotherapy; IQR, interquartile range; ICT, induction chemotherapy; Nut, nutrition; WBC, white blood cell; Hgb, hemoglobin

Figure 1

Overall survival for chemoradiation with vs. without induction after matching.

Figure 2

Cancer-specific survival for chemoradiation with vs. without induction after matching.

HDC and LDC chemoradiation

A total of 183 pairs were matched, with all variables well balanced (). Median overall follow-up was 60.4 months (IQR, 38.0–88.4). Median OS was 38.7 months (IQR, 24.8–61.5) and 44.8 months (IQR, 20.8–78.6) for HDC and LDC cohorts, respectively (P=0.35). OS at 5 years was 59.3% (95% CI, 51.9–67.7%) for patients treated with HDC and 54.1% (95% CI, 46.9–62.2%) for patients treated with LDC (P=0.35, ). CSS at 5 years was 70.3% (95% CI, 63.0–78.4%) for HDC cohort and 62.4% (95% CI, 55.2–70.5%) for LDC cohort (P=0.09, ).

Figure 3

Overall survival for three-weekly cisplatin (cis q3wk) vs. weekly cisplatin (cis q1wk) chemoradiation after matching.

Figure 4

Cancer-specific survival three-weekly cisplatin (cis q3wk) vs. weekly cisplatin (cis q1wk) chemoradiation after matching.

HDC and non-cisplatin chemoradiation

A total of 94 pairs were matched, with all variables well balanced (). Median overall follow-up was 52.3 months (IQR, 39.1–94.5). Median OS was 37.7 months (IQR, 21.4–60.5) and 42.6 months (IQR, 19.9–70.9) for HDC and other chemotherapy cohorts, respectively (P=0.51). OS at 5 years was 54.5% (95% CI, 44.7–66.4%) for patients treated with three-weekly HDC and 58.3% (95% CI, 48.6–69.8%) for patients treated with non-cisplatin CRT (P=0.51, ). CSS at 5 years was 67.5% (95% CI, 57.8–78.9%) for HDC cohort and 64.7% (95% CI, 55.0–76.0%) for non-cisplatin cohort (P=0.45, ).

Figure 5

Overall survival for three-weekly cisplatin (cis q3wk) vs. non-cisplatin (other) chemoradiation after matching.

Figure 6

Cancer-specific survival for three-weekly cisplatin (cis q3wk) vs. non-cisplatin (other) chemoradiation after matching.

Discussion

Analysis of well-balanced matched pairs of CRT regimens found: (I) ICT does not show to increase survival benefit of CRT alone, (II) HDC may not be the optimal dose as LDC shows insignificant difference in survival, and (III) non-cisplatin regimens fail to improve survival compared to HDC but may be an effective alternative for patients who are unfit to tolerate cisplatin. These findings are consistent with the literature.

We controlled for variables such as current smoking status, older age, advanced tumor stage, unexpected hospitalization, and nutrition support () that are known to be associated with worse survival in our patients as well as other variables by performing propensity score matching in three groups of patients and created well-balanced matched-pairs (). Compared to patients who received ICT prior to CRT, patients who did not receive ICT had no difference in 5-year OS (51.5% vs. 41.0%, P=0.53, ) and CSS (64.9% vs. 49.7%, P=0.21, ). Compared to patients who received HDC-CRT, patients who received LDC had no difference in 5-year OS (59.3% vs. 54.1%, P=0.35, ) and CSS (70.3% vs. 62.4%, p=0.09, ); patients who received non-cisplatin also had no difference in 5-year OS (54.5% vs. 58.3%, P=0.51, ) and CSS (67.5% vs. 64.7%, P=0.45, ).

ICT compared to chemoradiation alone

Despite its potential to reduce tumor burden and assist in the administration of and patient selection for adjuvant therapy, ICT in treatment of HNC remains debatable with unproven advantage over standard-of-care with CRT alone ().

Table 4

References for studies on induction chemotherapy

ICT	Study								Chemo		RT		OS				DFS
	Author, year	Design	Incl pd	Tx	Arms	Pts (no.)	Median f/u (mo.)		Regimen		Dose to 1' (Gy)		3-yr OS	5-yr OS	Diff in OS		3-yr DFS	5-yr DFS	Diff in DFS
ICT vs. Surg	Wolf et al. (VA Larynx), 1991 (25)	Prosp PIII, LC	–	Def	A. ICT then RT	166	33		PF		66–76		60%	–	p=0.98		58%	53%	P=0.12
					B. Surg then RT	166			–		50–50.4 +/− 10		63%	–			70%	65%
	Lefebvre et al. (EORTC 24891), 1996 (26)	Prosp PIII, PSC	1990–1992	Def	A. ICT then RT	100	51		PF		50 +/− 20		57%	30%	None reported		43%	25%	None reported
					B. Surg then RT	94			–		50 +/− 14		43%	35%			32%	27%
	Forastiere et al. (RTOG 91–11), 2003 (3)	Prosp PIII, LC	1992–2000	Def	A. ICT then RT	173	45.6		Cis–5FU		70		76% (2yr)	55%	None reported		52% (2–yr)	38%	A>C, P=0.02; B>C, P=0.006
					B. CRT	172			HDC		70		74% (2-yr)	54%			61% (2-yr)	36%
					C. RT alone	173			–		70		75% (2-yr)	56%			44% (2-yr)	27%
ICT vs. CRT alone	Haddad et al. (PARADIGM), 2013 (27)	Prosp PIII	2004–2008	Def	A. ICT then CRT	70	49		TPF–> Doce/Carbo		72/70		73%	67%	P=0.77		67% (PFS)	–	P=0.82
					B. CRT alone	75			HDC		72		78%	70%			69% (PFS)	–
	Cohen et al. (DeCIDE), 2014 (28)	Prosp PIII	2004–2009	Def	A. ICT tbe CRT	138	30		TPF–> DFHX		74–75		–	67%	P=0.68		–	–	–
					B. CRT alone	135			DFHX		74–75		–	67%			–	–	–
	Ghi et al., ABSTRACT only (30)	Prosp PIII	2008–2014	Def	A. ICT then CRT	207	41.3		TPF –> Cis–5FU or Cetux		70		58%	–	A>B, P=0.025		47% (PFS)	–	A>B, P=0.015
					B. CRT alone	208			Cis–5FU or Cetux		70		46%	–			37% (PFS)	–
	Stokes et al., 2017 (31)	Retrosp	2003–2011	Def	A. ICT then CRT	1569	29.7		NA		>66		–	–	p=0.35		–	–	–
					B. CRT alone	6462			NA		>66		–	–			–	–	–
	Chen et al., 2016 (32)	Retrosp	2002–2011	Def	A. CRT alone	7986	50		Pt–based		70		50%	44%	A>B/C, P<0.0001		–	46%	A>B/C, P<0.0001
					B. ICT +/− RT/CRT	503			Docetaxel–based		70		38%	30%			–	41%
					C. ICT +/− RT/CRT	2232			Pt–based		70		38%	30%			–	38%
	Ock et al., 2016 (33)	Retrosp	2005–2013	Def	A. ICT then CRT	144	52.4		Varied		60		77%	–	A>B, P=0.017 (matched)		65% (PFS)	–	P=0.06
					B. CRT alone	80			Varied		60		57%	–			54% (PFS)	–
	Merlano et al., Ongoing (34)	Prosp PIII	Ongoing	Def	A. ICT then CRT	–	–		TPF- > Cetux		70		–	–	–		–	–	–
					B. CRT alone	–	–		HDC		70		–	–	–		–	–	–
	Yang et al., 2019 (35)	Prosp PIII, NPC	2008–2015	Def	A. ICT then CRT	238	82.6		PF- > HDC		–		89%	81%	A>B, P=0.04		81%	73%	A>B, P=0.007
					B. CRT alone	238			HDC		–		88%	77%			74%	63%
	Zhang et al., 2019 (36)	Prosp PIII, NPC	2013–2016	Def	A. ICT then CRT	242	42.7		Gem-Cis –> HDC		70		95%	–	A>B, HR 0.43		85%	–	A>B, P=0.001
					B. CRT alone	238			HDC		70		90%	–			77%	–

Chemo, chemotherapy; RT, radiotherapy; CRT, chemoradiation; OS, overall survival; DFS, disease–free survival; PFS, progression–free survival; ICT, induction chemotherapy; Incl pd, inclusion period; Tx, treatment; Pts, patients; f/u, follow–up; 1’, primary; Diff, difference; Surg, surgery; Pros, prospective; PIII, phase III; Retrosp, retrospective; LC, laryngeal cancer; PSC, pyriform sinus cancer; NPC, nasopharyngeal cancer; TPF, docetaxel-Platinum-5-Fluorouracil (Docetaxel 75 mg/m2 on day 1, cisplatin 75–100 mg/m2 on day 1, 5-fluorouracil 750–1,000 mg/m2 on days 1–4 as continuous infusion; 3–4 cycles on 21–day interval); PF, Platinum-5-Fluorouracil (Cisplatin 80–100 mg/m2 given as rapid intravenous infusion followed by 5-fluorouracil 800–1,000 mg/m2/day continuous 24-hour infusion for 5 days; 2–4 cycles on 21-day interval); Def, definitive; Cis-5FU, Cisplatin-5-Fluorouracil (Cisplatin 75–100 mg/m2 bolus then 5–fluorouracil 1 g continuous infusion for 2–3 cycles); HDC, high dose cisplatin (80–100 mg/m2 3–weekly ×2–3 cycles); Doce, Docetaxel (20 mg/m2 weekly for 4 cycles); Carbo, Carboplatin (weekly); DFHX, Docetaxel-fluorouracil-hydroxyurea; Cetux, Cetuximab (initial dose 400 mg/m2 during the week before radiotherapy followed by maximum of 7 doses of 250 mg/m2 during radiotherapy; Pt, Platinum; Gem-Cis, Gemcitabine-Cisplatin (Gemcitabine 1 g/m2 on days 1 and 8, cisplatin 80 mg/m2 on days 1, 22, 43).

Table 5

References for studies on induction chemotherapy

ICT	Study			LRC			DC				Response
	Author	Arms		2-yr LRC	Diff in LRC		3-yr DC	5-yr DC	Diff in DC		Overall Resp	Diff in Resp
ICT vs. Surg	Wolf et al. (25)	A. ICT –> RT		92%, ns	B>A, P=0.0005		89%, ns	–	A>B, P=0.016		–	None reported
		B. Surg –> RT		95%, ns			83%, ns	–			–
	Lefebvre et al. (26)	A. ICT –> RT			None reported		72%	65%	A>B, P=0.041		–	None reported
		B. Surg –> RT					57%	52%			–
	Forastiere et al. (3)	A. ICT –> RT		61% [54–69]	B>A, P=0.003		91%, 2-yr DC	85%	B>C, P=0.03		–	–
		B. CRT		78% [72–85]	A vs. C, P=0.16		92%, 2-yr DC	88%			–	–
		C. RT alone		56% [48–63]	B>C, P<0.001		84%, 2-yr DC	78%			–	–
ICT vs. CRT alone	Haddad et al. (27)	A. ICT –> CRT		84%, ns	None reported		93%, ns	–	None reported		–	–
		B. CRT alone		85%, ns			89%, ns	–			–	–
	Cohen et al. (28)	A. ICT –> CRT		70%	P=0.16		–	62%	P=0.37		74%	P=0.45
		B. CRT alone		60%			–	60%			79%
	Ghi et al., Abstract (30)	A. ICT –> CRT		–	–		–	–	–		–	–
		B. CRT alone		–	–		–	–	–		–	–
	Stokes et al. (31)	A. ICT –> CRT		–	–		–	–	–		–	–
		B. CRT alone		–	–		–	–	–		–	–
	Chen et al. (32)	A. CRT alone		–	–		–	–	–
		B. ICT +/− RT/CRT		–	–		–	–	–
		C. ICT +/− RT/CRT		–	–		–	–	–
	Ock et al. (33)	A. ICT –> CRT		69% (3-yr)	P=0.11		89%	–	P=0.85		76%	A>B, P=0.005 (matched)
		B. CRT alone		59% (3-yr)			87%	–			53%
	Merlano et al., ongoing (34)	A. ICT –> CRT		–	–		–	–	–		–	–
		B. CRT alone		–	–		–	–	–		–	–
	Yang et al. (35)	A. ICT –> CRT		88% (5-yr)	P=0.21		86%	83%	A>B, P=0.014		–	–
		B. CRT alone		85% (5-yr)			82%	73%			–	–
	Zhang et al. (36)	A. ICT –> CRT		92% (3-yr)	None reported		91%	–	None reported		97%	None reported
		B. CRT alone		91% (3-yr)			84%	–			97%

LRC, locoregional control; DC, distant control; ICT, induction chemotherapy; CRT, chemoradiation; RT, radiotherapy; Diff, difference; Resp, response; Surg, surgery; TPF, docetaxel-platinum-5-fluorouracil (docetaxel 75 mg/m2 on day 1, cisplatin 75–100 mg/m2 on day 1, 5-fluorouracil 750–1,000 mg/m2 on days 1–4 as continuous infusion; 3–4 cycles on 21-day interval); PF, platinum-5-fluorouracil (cisplatin 80–100 mg/m2 given as rapid intravenous infusion followed by 5-fluorouracil 800–1,000 mg/m2/day continuous 24-hour infusion for 5 days; 2–4 cycles on 21-day interval).

Table 6

References for studies on induction chemotherapy (continued)

ICT	Study			Severe acute toxicity (grade 3–5)
	Author	Arms		Pts (no.)	N/V	Mucositis	Dysphagia	Leukopenia	Neutropenia	Tbcp	Anemia	Infection	Renal	Neuro	Skin	Ototoxic	Total toxic deaths	All	Diff in acute toxicity
ICT vs. Surg	Wolf et al. (25)	A. ICT –> RT		–	–	–	–	–	–	–	–	–	–	–	–	–	5	–	None reported
		B. Surg –> RT		–	–	–	–	–	–	–	–	–	–	–	–	–	3	–
	Lefebvre et al. (26)	A. ICT –> RT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. Surg –> RT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Forastiere et al. (3)	A. ICT –> RT		168	14%	20%	19%	52% (hematologic)				5%	2%	4%	10%	–	5	66%	None reported
		B. CRT		171	20%	43%	35%	47% (hematologic)				4%	4%	5%	7%	–	9	77%
		C. RT alone		171	0	24%	19%	3% (hematologic)				1%	0	0	9%	–	5	47%
ICT vs. CRT alone	Haddad et al. (27)	A. ICT –> CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Cohen et al. (28)	A. ICT –> CRT		124	6%/3%	51%	12%	26%	3%	3%	7%	11%	–	–	18%	–	–	47%	A>B, P=0.002
		B. CRT alone		133	5%/2%	47%	15%	11%	2%	2%	3%	14%	–	–	24%	–	–	24%
	Ghi et al., Abstract (30)	A. ICT –> CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Stokes et al. (31)	A. ICT –> CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Chen et al. (32)	A. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. ICT +/− RT/CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
		C. ICT +/− RT/CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Ock et al. (33)	A. ICT –> CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Merlano et al., Ongoing (34)	A. ICT –> CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		B. CRT alone		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Yang et al. (35)	A. ICT –> CRT		No report on acute adverse events. Eye damage significantly higher with CRT alone than ICT–>CRT (16% vs. 10%, P=0.03)
		B. CRT alone
	Zhang et al. (36)	A. ICT –> CRT		239	23%	29%	–	26%	28%	11%	10%	0	3%	–	2%	0	–	76%	None reported
		B. CRT alone		237	14%	32%	–	20%	11%	1%	1%	0	0%	–	4%	0	–	56%

ICT, induction chemotherapy; Pts, patients; N/V, nausea or vomiting; Tbcp, thrombocytopenia; Neuro, neurological; Diff, difference; Surg, surgery; TPF, docetaxel-platinum-5-fluorouracil (docetaxel 75 mg/m2 on day 1, cisplatin 75–100 mg/m2 on day 1, 5–fluorouracil 750–1,000 mg/m2 on days 1–4 as continuous infusion; 3–4 cycles on 21-day interval); PF, platinum-5-fluorouracil (cisplatin 80–100 mg/m2 given as rapid intravenous infusion followed by 5-fluorouracil 800–1,000 mg/m2/day continuous 24-hour infusion for 5 days; 2–4 cycles on 21-day interval); CRT, chemoradiation; RT, radiotherapy.

Pignon et al. in a meta-analysis of 87 trials found that CRT had a greater mortality benefit than ICT, though ICT offered a significant reduction of distant metastasis (DM) risk (HR 0.73, 95% CI, 0.61–0.88, P=0.001) (1).

Three recent phase III randomized trials compared docetaxel, cisplatin, and 5-flurouracil (TPF) ICT followed by CRT against CRT alone in patients with locally advanced HNC (28,29,34). Two of these trials fell short of their target accrual (145 of targeted 330 patients in PARADIGM, 285 of targeted 400 patients in DeCIDE trial) and failed to show significant difference in OS between the two arms (HR 1.09, 95% CI, 0.59–2.03, P=0.77 in PARADIGM; HR 0.91, 95% CI 0.59–1.41, P=0.70 in DeCIDE), and both showed 3-year OS rates over 20% higher than the expected 50–55% in the two arms (28,29). Although we reviewed a heterogeneous patient population including those with advanced as well as earlier stage disease and used taxane-based of ICT, analysis of our matched pairs supports the lack of improvement in survival with taxane-based induction before CRT. Similarly, a retrospective analysis of over 8,000 patients in the National Cancer Data Base (NCDB) by Stokes et al. reports that ICT does not offer significant survival advantage when compared to CRT alone (HR 0.96, 95% CI, 0.88–1.05, P=0.35) while making it more likely for patients to receive lower (<66 Gy) RT doses (P<0.01); subgroup analysis on advanced disease also did not show difference in survival with ICT (31). Chen et al. retrospectively analyzed over 10,000 HNC patients in Taiwan who were treated with either CRT alone or ICT (docetaxel- or platinum-based) preceding locoregional treatment and also showed superior survival rate (P<0.0001) with CRT alone (32).

In contrast, Ghi et al. in Italy randomized 420 patients to receive TPF then CRT (with cisplatin or cetuximab) or CRT (with cisplatin or cetuximab) alone and demonstrated 3-year OS (57.6% vs. 45.7%, HR 0.72, 95% CI, 0.55–0.96, P=0.025) and PFS (46.8% vs. 36.7%, HR 0.73, 95% CI, 0.57–0.94, P=0.015) favoring TPF over CRT alone (30). The 3-year OS rates fall within the expected range but may reflect compromised survival due to use of cetuximab- rather than solely cisplatin-based CRT. A retrospective single-center analysis with propensity score matching by Ock et al. was similar to our study but had different results that too showed survival benefit with taxane-based ICT, which improved 3-year OS (77.4% vs. 56.7%, HR 0.48, 95% CI, 0.26–0.87, P=0.017) as well as complete response rates (75.7% vs. 52.9%, P=0.005) compared to CRT alone (33). Subgroup analysis was also done and showed that male patients with N2-3 oropharyngeal cancer had improved OS with ICT followed by CRT.

The ongoing phase III trial (INTERCEPTOR) by Gruppo Oncologico del Nord-Ovest comparing TPF followed by cetuximab-CRT and HDC-CRT alone (clinicaltrials.gov, NCT00999700) will stratify patients by HPV status (34).

ICT may also play a role in locoregionally advanced nasopharyngeal cancer (LA-NPC.) Yang et al. recently reported on 476 patients with LA-NPC that demonstrated long-term OS (81% vs. 77%, P=0.04) and disease-free survival (DFS) (73% vs. 63%, P=0.007) benefits with ICT preceding standard CRT (35). Zhang et al. randomized a similarly sized cohort of LA-NPC patients to receive gemcitabine and cisplatin-based ICT or CRT alone and also reported improved 3-year OS (HR 0.43, 95% CI, 0.24–0.77) and 3-year DFS (85% vs. 77%, P=0.001) with ICT (36).

With the possible exception of LA-NPC, routine use of ICT may not be advised given increased toxicity and no clear survival benefit. In our patients, we no longer routinely use ICT except: (I) on a clinical trial or (II) if required to achieve 30 days of smoking cessation in current smokers. Current smokers are known to have significantly reduced survivals that can be effectively ameliorated by 30 days of smoking cessation (37). This OS benefit justifies use of ICT; moreover, ICT allows initiation of treatment without significant delay from the time of diagnosis which can also reduce survival (38).

Optimal cisplatin dose for chemoradiation

In addition to studies showing the OS and LRC benefits of HDC (1-7), as shown in , 2 of 3 studies of LDC with RT showed significant OS benefit over RT alone and the third showed significant improvement in LRC.

Table 7

References for studies on cisplatin–based chemoradiotherapy regimen

CIS-CRT	Study								Chemo		RT		OS					DFS
	Author, year	Design	Incl pd	Tx	Arms	Pts (no.)	Median f/u (mo.)		Regimen		Dose to 1’ (Gy)		3-yr OS	5-yr OS		Diff in OS		3-yr DFS	5-yr DFS	Diff in DFS
HDC vs. RT	Pignon et al. (MACH–NC), 2009 (1)	Retrosp	1965–2000	Def/Adj	CRT vs. LRT	Total 17,346	67.2		Varied		Varied		CRT has 6.5% 5-yr absolute OS benefit, P<0.001					CRT improves DFS, P<0.0001
					ICT vs. LRT				Varied		Varied
	Adelstein et al., 2003 (2)	Prosp PIII	1992–1999	Def	A. RT alone	95	41		–		70		23%		14%	B>A, P=0.014		33%	24%	B>A, P=0.01
					B. CRT	87			HDC		70		37%		26%			51%	45%
					C. Split CRT	89			Cis–5FU		60–70		27%		21%			41%	41%
	Cooper et al. (RTOG 9501), 2004 (5)	Prosp PIII	1995–2000	Adj	A. RT alone	210	45.9		–		60–66		45%		40%	P=0.19		38%	25%	B>A, P=0.04
					B. CRT	206			HDC		60–66		55%		45%			48%	35%
	Bernier et al. (EORTC 22931), 2004 (6)	Prosp PIII	1994–2000	Adj	A. RT alone	167	60		–		66		48%		40%	B>A, P=0.02		40% (PFS)	36% (PFS)	B>A, P=0.04
					B. CRT	167	60		HDC		66		60%		53%			55% (PFS)	47% (PFS)
LDC vs. RT	Bachaud et al., 1996 (9)	Prosp PIII	1984–1988	Adj	A. RT alone	44	36		–		65–74		46% (2-yr)		13%	B>A, P<0.01		44% (2-yr)	23%	B>A, P<0.02
					B. CRT	39	36		LDC		65–74		72% (2-yr)		36%			68% (2-yr)	45%
	Sharma et al., 2010 (10)	Prosp PII, OP/NPC	2003–2005	Def	A. RT alone	76	22		–		70		42%		–	B>A, P=0.024		42% (PFS)	–	P=0.88
					B. CRT	77	22		LDC		70		62%		–			37% (PFS)	–
	Ghosh-Laskar et al., 2016 (11)	Prosp PIII	2000–2007	Def	A. RT alone	57	48		–		66–70		–		36%	P=0.11			25%	B>A, P=0.03
					B. CRT	65	48		LDC		66–70		–		56%				39%
HDC vs. LDC	Lee et al., 2018 (12)	Retrosp	2007–2012	Def	A. HDC-CRT	65	–		HDC		66.4, mean		81%		–	P=0.34		64% (PFS)	–	P=0.81
					B. LDC-CRT	155	–		LDC		68.4, mean		67%		–			60% (PFS)	–
	Helfenstein et al., 2019 (13)	Retrosp	2008–2015	Def/Adj	A. HDC-CRT	127	40.6		HDC		69–72/60–66		Cum. cis dose >200 mg/m2 on OS, P=0.10					Cum cis dose >200 mg/m2 on PFS, P=0.97
					B. LDC-CRT	187	40.6		LDC		69–72/60–66
	Bauml et al., 2019 (14)	Retrosp	2000–2014	Def	A. HDC-CRT	2200	–		HDC		–		–	–		P=0.44		–	–	–
					B. LDC-CRT	701	–		LDC		–		–	–				–	–	–
	Szturz et al., 2017 (15)	Retrosp	–	Def/Adj	A. HDC-CRT	31 studies	–		HDC		66–70/60–66		–	–		No diff in OS in Def/Adj setting		Unable to merge data from studies
					B. LDC-CRT	17 studies	–		LDC		66–70/60–66		–	–
	Tsan et al., 2012 (16)	Prosp PIII, OC	2008–2010	Adj	A. HDC-CRT	26	12		HDC		66		79% (1-yr)	–		P=0.98		–	–	–
					B. LDC-CRT	24	12		LDC		66		72% (1-yr)	–				–	–	–
	Noronha et al., 2018 (17)	Prosp PIII	2013–2017	Def/Adj	A. HDC-CRT	150	22		HDC		60 or 70		53%	–		P=0.48		48% (2-yr PFS)	–	P=0.21
					B. LDC-CRT	150	22		LDC		60 or 70		53%	–				55% (2-yr PFS)	–

Chemo, chemotherapy; RT, radiotherapy; OS, overall survival; DFS, disease-free survival; PFS, progression-free survival; Cis, cisplatin; CRT, chemoradiation; Incl pd, inclusion period; Tx, treatment; f/u, follow-up; 1’, primary; Diff, difference; Retrosp, retrospective; Prosp, prospective; PII/III, phase II/III; Def, definitive; Adj, adjuvant; LRT, locoregional treatment; ICT, induction chemotherapy; HDC, high dose cisplatin (80–100 mg/m2 3-weekly, 2–3 cycles); Cis-5FU, cisplatin-5-fluorouracil (Cisplatin 75–100 mg/m2 bolus then 5-fluorouracil 1 g continuous infusion for 2–3 cycles); LDC, low dose cisplatin (30–50 mg/m2 weekly, 6–9 cycles); OP, oropharyngeal cancer; NPC, nasopharyngeal cancer; OC, oral cavity squamous cell cancer.

Table 8

References for studies on cisplatin–based chemoradiotherapy regimen (continued)

CIS-CRT	Study			LRC				DC				Response
	Author	Arms		2-yr LRC		Diff in LRC		3-yr DC	5-yr DC	Diff in DC		Overall resp	Diff in resp
HDC vs. RT	Pignon et al. (1)	CRT vs. LRT		CRT improves LRC, P=0.04				ICT improves DC, P=0.001				–	–
		ICT vs. LRT										–	–
	Adelstein et al. (2)	A. RT alone		–		–		–	–	–		27.4% (CR)	C>A, P=0.002; A vs. B, P=0.07
		B. CRT		–		–		–	–	–		40.2% (CR)
		C. Split CRT		–		–		–	–	–		49.4% (CR)
	Cooper et al. (5)	A. RT alone		72%		B>C, P=0.01		77%, ns		P=0.46		–	–
		B. CRT		82%				80%, ns				–	–
	Bernier et al. (6)	A. RT alone		69%, 5-yr LRC		B>A, P=0.007		–	75%	P=0.61		–	–
		B. CRT		88%, 5-yr LRC				–	79%			–	–
LDC vs. RT	Bachaud et al. (9)	A. RT alone		59%		B>A, P=0.05		81% (2-yr)	49%	None reported		–	–
		B. CRT		84%				73% (2-yr)	58%			–	–
	Sharma et al. (10)	A. RT alone		53%, overall		P=0.26		97%, overall	–	P=0.05		67% (CR)	B>A, P=0.04
		B. CRT		66%, overall				90%, overall	–			81% (CR)
	Ghosh-Laskar et al. (11)	A. RT alone		32%, 5-yr LRC		B>A, P=0.01		–	–	None reported		–	–
		B. CRT		49%, 5-yr LRC				–	–			–	–
HDC vs. LDC	Lee et al. (12)	A. HDC-CRT		–		–		–	–	–		92%	P=0.81
		B. LDC-CRT		–		–		–	–	–		91%
	Helfenstein et al. (13)	A. HDC-CRT		–		–		–	–	–		–	–
		B. LDC-CRT		–		–		–	–	–		–	–
	Bauml et al. (14)	A. HDC-CRT		–		–		–	–	–		–	–
		B. LDC-CRT		–		–		–	–	–		–	–
	Szturz et al. (15)	A. HDC-CRT		Unable to merge data				Unable to merge data				80.0%	No diff
		B. LDC-CRT										89.0%
	Tsan et al. (16)	A. HDC-CRT		71% (1-yr)	P=0.81			–	–	–		–	–
		B. LDC-CRT		60% (1-yr)				–	–	–		–	–
	Noronha et al. (17)	A. HDC-CRT		73%	A>B, P=0.014			–	–	–		–	–
		B. LDC-CRT		59%				–	–	–		–	–

Cis, cisplatin; CRT, chemoradiation; LRC, locoregional control; DC, distant control; Diff, difference; Resp, response; HDC, high dose cisplatin (80–100 mg/m2 3-weekly, 2–3 cycles); LDC, low dose cisplatin (30–50 mg/m2 weekly, 6–9 cycles); RT, radiotherapy; LRT, locoregional treatment; CR, complete response; ns, not specified.

Table 9

References for studies on cisplatin–based chemoradiotherapy regimen (continued)

CIS-CRT	Study			Severe acute toxicity (grade 3–5)
	Author	Arms		Pts (no.)	N/V	Mucositis	Dysphagia	Leukopenia	Neutropenia	Tbcp	Anemia	Infection	Renal	Neuro	Skin	Oto–toxic	Total toxic deaths	All	Diff in acute toxicity
HDC vs. RT	Pignon et al. (1)	CRT vs. LRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	Not assessed
		ICT vs. LRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Adelstein et al. (2)	A. RT alone		98	6%	33%		1%	–	0	0	–	1%	–	13%	–	2	52%	B>A, P<0.0001
		B. CRT		95	16%	45%		42%	–	3%	18%	–	8%	–	7%	–	4	89%	B>C, P=0.02
		C. Split CRT		94	9%	47%		31%	–	3%	19%	–	0	–	2%	–	2	77%	C>A, P<0.001
	Cooper et al. (5)	A. RT alone		210	0	18%	15%	0			0	0	0	0	10%	–	0	34%	B>A, P<0.001
		B. CRT		206	19%	30%	24%	38% (hematologic)			3%	6%	2%	5%	7%	–	4	77%
	Bernier et al. (6)	A. RT alone		167	–	21%	12%	–	–	–	–	–	–	–	1%	–	1	–	B>A, P=001
		B. CRT		167	12%	41%	10%	16%	13%	–	–	–	–	–	2%	–	1	–
LDC vs. RT	Bachaud et al. (9)	A. RT alone		44	–	–	–	–	–	–	–	–	–	–	–	–	0	18%	None reported
		B. CRT		39	–	–	–	–	–	–	–	–	–	–	–	–	0	41%
	Sharma et al. (10)	A. RT alone		76	–	–	–	–	–	–	–	–	–	–	–	–	0	20%	B>A, P=0.015
		B. CRT		77	–	–	–	–	–	–	–	–	–	–	–	–	0	40%
	Ghosh-Laskar et al. (11)	A. RT alone		57	–	21%	–	–	–	–	–	–	–	–	26%	–	0	–	None reported
		B. CRT		65	–	35%	–	–	–	–	–	–	–	–	23%	–	2	–
HDC vs. LDC	Lee et al. (12)	A. HDC-CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
		B. LDC-CRT		–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	Helfenstein et al. (13)	A. HDC-CRT		127	–	–	–	–	–	–	–	–	33%	–	–	–	–	–	A>B, P=0.02
		B. LDC-CRT		187	–	–	–	–	–	–	–	–	21%	–	–	–	–	–
	Bauml et al. (14)	A. HDC-CRT		2200	HDC: higher rate of AKI (HR 1.72, P<0.001), neutropenia (HR 2.21, P=0.005), dehydration (HR 1.15, P=0.04), hearing loss (HR 1.34, P=0.004)
		B. LDC-CRT		701
	Szturz et al. (15)	A. HDC-CRT		25,6 studies	16, 10%	42, 37%	26, 20%	19, 19%	18, 14%	4, 2%	8, 6%	5, 11%	5, 3%	2, 5%	11, 6%	3, 2%	3, 2%	–	Bold = sig diff
		B. LDC-CRT		14,3 studies	3, 16%	25, 51%	8, 54%	1, 12%	5, 9%	1, 2%	4, 3%	8, NA%	1, 2%	1, NA%	14, 12%	NA, 2%	2, 1%	–
	Tsan et al. (16)	A. HDC-CRT		26	12%	39%	54%	0	0	0	4%		–	–	8%	–	–	81%	B>A, P=0.02
		B. LDC-CRT		24	21%	75%	54%	13%	4%	0	4%	–	–	–	8%	–	–	92%
	Noronha et al. (17)	A. HDC-CRT		149	7%	18%	39%	16%	13%	2%	5%	34%	0%	0%	8%	13%	10	85%	A>B, P=0.006
		B. LDC-CRT		148	1%	17%	42%	3%	1%	3%	2%	21%	0%	0%	7%	5%	5	72%

Cis, cisplatin; CRT, chemoradiation; Pts, patients; N/V, nausea or vomiting; Tbcp, thrombocytopenia; Neuro, neurological; Diff, difference; HDC, high dose cisplatin (80–100 mg/m2 3-weekly, 2–3 cycles); LDC, low dose cisplatin (30–50 mg/m2 weekly, 6–9 cycles); RT, radiotherapy; LRT, locoregional treatment; AKI, acute kidney injury; HR, hazard ratio.

Szturz et al. in a meta-analysis of 59 prospective studies with over 5,200 locally advanced HNC patients found no significant difference in OS was observed between HDC- and LDC-CRT. In the definitive setting, LDC had greater compliance (88% vs. 71%, P=0.0017) and less toxicity such as myelosuppression (leukopenia P=0.0083, neutropenia P=0.0024), severe nephrotoxicity (P=0.01) and severe nausea and/or vomiting (P<0.0001) compared to HDC (15).

Our findings are consistent with those of Szturz et al.; LDC appears to be equivalent to HDC in terms of OS. Although in our study we controlled for adverse events such as hospitalizations and risk factors such as comorbidities to create matched pairs, we are aware that in the early years of this analysis many patients were given LDC specifically because they were felt to be unable to tolerate HDC. Thus, we suspect that there remained a bias that favored HDC despite our attempts to correct with match pairing. This may explain the non-significant decrease in CSS with LDC.

As shown in , the majority of publications appear to reveal greater cisplatin-related toxicity with HDC than LDC. This warrants future study.

Non-cisplatin agents for chemoradiation

Non-platinum agents such as cetuximab (IgG1 monoclonal antibody against epidermal growth factor receptor) and other platinum agents such as carboplatin (second generation platinum drug) have been investigated (). Two randomized phase III trials by Gillison et al. (RTOG 1016) and Mehanna et al. (De-ESCALATE) examined the outcome of cetuximab (400 mg/m2 loading dose then seven weekly 250 mg/m2 doses) versus HDC given concomitantly with RT (70 Gy in standard fractions over six weeks) used to treat patients with HPV-positive oropharyngeal cancer; both trials failed to demonstrate non-inferiority of cetuximab over cisplatin (19,20). RTOG 1016 showed that cetuximab neither met the non-inferiority criteria for OS (P=0.51) nor improved acute (P=0.16) or late (P=0.19) severe toxicity profile of cisplatin while exhibiting inferior 5-year progression-free survival (67.3% vs. 78.4%, P=0.0002). De-ESCALATE study also showed that cetuximab did not reduce overall severe toxicity (P=0.98) while showing worse 2-year OS (89.4% vs. 97.5%, P=0.001) and 2-year recurrence (16.1% vs. 6.0%, P=0.001) compared to cisplatin. Our non-cisplatin cohort included patients who received weekly cetuximab and those who received modified regimen of crossover to cetuximab, but we did not directly compare cisplatin- with cetuximab-based CRT in our retrospective review of a heterogeneous patient population. Although we showed that patients who received non-cisplatin regimen had no difference in OS with those who received cisplatin, this may be the result of including other platinum-based regimen in our non-cisplatin cohort. It appears safe to assume that cetuximab is neither less toxic nor equally as effective as cisplatin, which thus cannot be replaced in treating HPV-positive oropharyngeal cancer.

Table 10

References for studies on non–cisplatin chemoradiotherapy regimen

Study								Chemo		RT		OS				DFS
Author, year	Design	Incl pd	Tx	Arms	Pts (no.)	Median f/u (mo.)		Regimen		Dose to 1’ (Gy)		3-yr OS	5-yr OS	Diff in OS		3-yr DFS	5-yr DFS	Diff in DFS
Bonner et al., 2006 (19)	Prosp PIII	1992–2002	Def	A. RT alone	213	54		–		70–76.8		45%	32%	B>A, P=0.03		31% (PFS)	–	B>A, P=0.04
				B. Cetux-CRT	211	54		Cetux		70–76.8		55%	40%			42% (PFS)	–
Gillison et al. (RTOG 1016), 2019 (20)	Prosp PIII, HPV-OP	2011–2014	Def	A. HDC-CRT	406	54		HDC		70		90%	85%	A>B, P=0.016		82%	78%	A>B, P=0.0002
				B. Cetux-CRT	399	54		Cetux		70		87%	78%			72%	67%
Mehanna et al. (DeESCALATE), 2019 (21)	Prosp PIII, HPV-OP	2012–2016	Def	A. HDC-CRT	166	25.9		HDC		70		98% (2-yr)	–	A>B, P=0.0012		–	–	–
				B. Cetux-CRT	168	25.9		Cetux		70		89% (2-yr)	–			–	–	–
Shapiro et al., 2014 (22)	Retrosp	2002–2008	Def	A. HDC-CRT	259	53.1		HDC		70		–	87% (4-yr)	A>C, P<0.0001		–	–	–
				B. Carbo-CRT	52	53.1		Carbo-5-FU		70		–	70% (4-yr)	B>C, P=0.002		–	–	–
				C. Cetux-CRT	49	53.1		Cetux		70		–	41% (4-yr)	A vs. B, P=0.35		–	–	–
Denis et al., 2004 (23)	Prosp PIII, OP	1994–1997	Def	A. RT alone	113	66		–		70		–	16%	B>A, P=0.05		–	15%	B>A, P=0.01
				B. Carbo-CRT	109	66		Carbo-5FU		70		–	22%			–	27%
Tao et al. (REACH), ABSTRACT only (24)	Prosp PIII	Ongoing	Def	A. HDC-CRT	–	–		HDC		70		–	–	–		–	–	–
				B. Cetux/Ave–RT	–	–		Cetux-Ave		70		–	–	–		–	–	–

Chemo, chemotherapy; RT, radiotherapy; OS, overall survival; DFS, disease–free survival; PFS, progression–free survival; CIS, cisplatin; CRT, chemoradiation; Incl pd, inclusion period; Tx, treatment; Pts, patients; f/u, follow-up; 1’, primary; Diff, difference; Prosp PIII, prospective phase III; Retrosp, retrospective; Def, definitive; HDC, high-dose cisplatin (80–100 mg/m2 3-weekly, 2–3 cycles); Carbo-5FU: carboplatin-5-Fluorouracil (carboplatin 70 mg/m2 and 5-fluorouracil 600 mg/m2/day continuous infusion for 4 days, 3 cycles on 21-day interval); Cetux: cetuximab (initial dose 400 mg/m2 during the week before radiotherapy followed by maximum of 7 doses of 250 mg/m2 during radiotherapy); Ave: avelumab (10 mg/kg intravenous infusion over 1 hour every 2 weeks during RT and for 12 months following radiotherapy); HPV, human papilloma virus; OP, oropharyngeal cancer.

Table 11

References for studies on non–cisplatin chemoradiotherapy regimen (continued)

Study, author	Arms	LRC			DC				Response
		2-yr LRC	Diff in LRC		3-yr DC	5-yr DC	Diff in DC		Overall Resp	Diff in Resp
Bonner et al. (19)	A. RT alone	41%	B>A, P=0.005		83%	–	None reported		64%	B>A, P=0.02
	B. Cetux-CRT	50%			84%	–			74%
Gillison et al. (20)	A. HDC-CRT	90% (5-yr)	A>B, P=0.0005		–	–	P=0.09		–	–
	B. Cetux-CRT	83% (5-yr)			–	–			–	–
Mehanna et al. (21)	A. HDC-CRT	94%	A>B, P=0.0007		97%, ns	–	A>B, P=0.01		–	–
	B. Cetux-CRT	84%			91%, ns	–			–	–
Shapiro et al. (22)	A. HDC-CRT	94% (4-yr)	C>A, P<0.0001		–	88% (4-yr)	None reported		–	–
	B. Carbo-CRT	90% (4-yr)			–	82% (4-yr)			–	–
	C. Cetux-CRT	60 (4-yr)			–	71% (4-yr)			–	–
Denis et al. (23)	A. RT alone	25% (5-yr)	B>A, P=0.002		83%, ns	–	None reported		–	–
	B. Carbo-CRT	48% (5-yr)			82%, ns	–			–	–
Tao et al., Abstract only (24)	A. HDC-CRT	–	–		–	–	–		–	–
	B. Cetux/Ave-RT	–	–		–	–	–		–	–

LRC, locoregional control; DC, distant control; Cis, cisplatin; CRT, chemoradiation; Diff, difference; Resp, response; RT, radiotherapy; Cetux, cetuximab; Carbo, carboplatin; HDC, high dose cisplatin; Cetux/Ave, cetuximab or avelumab.

Table 12

References for studies on non-cisplatin chemoradiotherapy regimen (continued)

Study, author	Arms	Severe acute toxicity (grade 3–5)
		Pts (no.)	N/V	Mucositis	Dysphagia	Leukopenia	Neutropenia	Tbcp	Anemia	Infection	Renal	Neuro	Skin	Ototoxic	Total toxic deaths	All	Diff in acute toxicity
Bonner et al. (19)	A. RT alone	212	6%	52%	30%	–	–	–	6%	1%	–	–	1%	–	–	–	No sig diff
	B. Cetux-CRT	208	4%	56%	26%	–	–	–	1%	1%	–	–	17%	–	–	–
Gillison et al. (20)	A. HDC-CRT	398	19%	42%	37%	12%	15%	–	11%	–	3%	–	8%	3%	6	82%	P=0.16
	B. Cetux-CRT	394	8%	46%	32%	0%	1%	–	0%	–	0%	–	12%	0%	6	77%
Mehanna et al. (21)	A. HDC-CRT	162	–	–	–	12% (hematologic)				12%	7%	6%	4%	2%	–	–	P=0.49
	B. Cetux-CRT	165	–	–	–	1% (hematologic)				13%	0%	10%	30%	2%	–	–
Shapiro et al. (22)	A. HDC-CRT	No data on acute toxicity. Late toxicity was highest with 5FU/carboplatin (25%) vs. cisplatin (8%) vs. cetuximab (7.7%).															B>A, P=0.05; B>C, P=0.02
	B. Carbo-CRT
	C. Cetux-CRT
Denis et al. (23)	A. RT alone	No data on acute toxicity. Late toxicity no significant difference between arms
	B. Carbo-CRT
Tao et al., Abstract only (24)	A. HDC-CRT	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–
	B. Cetux/Ave-RT	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–	–

Cis, cisplatin; CRT, chemoradiation; Pts, patients; N/V, nausea or vomiting; Tbcp, thrombocytopenia; Neuro, neurological; Diff, difference; RT, radiotherapy; Cetux, cetuximab; HDC, high dose cisplatin; Carbo, carboplatin; Cetux/Ave, cetuximab or avelumab; 5FU, 5-fluorouracil.

There is an unmet need for an alternative CRT regimen for patients who cannot tolerate or risk the sequelae of severe toxicity with cisplatin. Prospective studies on alternative chemotherapy schedules and agents as well as immune checkpoint inhibitors (23) are warranted.

Conclusions

Survivals in our cohort were similar regardless of use of ICT, LDC, or non-cisplatin regimens. In the absence of a clear survival benefit, we only use ICT on clinical trial or as a temporizing maneuver for a patient trying to quit smoking. Patients unable to tolerate HDC should know that their survival may not be significantly impacted.

The article’s supplementary files as