Early mortality after diagnosis of cancer of the head and neck - A population-based nationwide study.

BACKGROUND: Cancers of the head and neck have a high mortality rate, and roughly 10% of the patients die within six months of diagnosis. To our knowledge little has been written about this group. We wished to identify risk factors for early death, to predict and monitor patients at risk better and, if possible, avoid unjustified major treatment. METHODS AND
FINDINGS: This population-based nationwide study from the Swedish Head and Neck Cancer Register (SweHNCR) included data from 2008-2015 and 9733 patients at potential risk of early death. A total of 925 (9.5%) patients died within six months. For every year older the patients became, the risk of early death increased by 2.3% (p<0.001). The relative risk of death was 3.37 times higher (237%) for patients with WHO score 1 compared with WHO score 0. A primary tumour in the hypopharynx correlated with a 24% increased risk over the oral cavity (p<0.024). Patients with stage IV disease had a 3.7 times greater risk of early death than those with stage I (p<0.001). As expected, a 12 times increased risk of early death was noted in the palliative treatment group, compared to the curative group. Limitations to this study were that the actual cause of death was not recorded in the SweHNCR, and that socioeconomic factors, alcohol consumption, smoking habits, and HPV status, were not reported in SweHNCR until 2015. However, the fact that this is a population-based nationwide study including 9733 patients compensates for some of these limitations.
CONCLUSIONS: Identification of patients at increased risk of early death shows that older patients with advanced disease, increased WHO score, primary tumour in the hypopharynx, and those given palliative treatment, are more likely than the others to die from head and neck cancer within six months of diagnosis.

Introduction

Sweden has a population of about 10 million inhabitants, and in 2016 the overall incidence of cancer was 64 000 in a population of just over 60 000. About 1400 new cases of cancer of the head and neck are reported annually, and according to the Swedish Cancer Society, this corresponds to 2.3% of all cancers [1]. Cancer of the head and neck is the sixth most common type worldwide, and most cases (60%) are locally advanced at the time of diagnosis (stage III or IV) [2]. Many of the tumours are aggressive, and the median survival without treatment is reported to be three to five months [3-6]. Late stages of the disease are linked to poor overall survival, but even patients with early cancers (for example of the mobile tongue) can have dismal prognoses [7, 8]. The management of metastatic and recurrent cancers of the head and neck is challenging, and typically involves combined treatments.

Both surgical and oncological treatments usually have acute side effects, but may also result in long-term consequences, and even death. Apart from our own recent review [9], the issue of early death (within six months of diagnosis) among patients with cancer of the head and neck remains poorly explored. Patients with improved survival are often HPV positive, of working age, and have only a few coexisting conditions, and for these patients reductions in treatment have been discussed [10, 11]. However, only a few authors have focussed on patients with the poorest prognosis–the older patients with coexisting conditions and high WHO scores [12, 13]. Treatment-induced side effects [14] [15] can be bearable as long as substantial duration of survival may be expected, but if it were possible to identify patients for whom it was short, resources could be more effectively spent on palliative, or at best supportive, care.

The primary aim of this study was to find out which patients were at risk of early death during the first six months after diagnosis in a large, nationwide, register-based group of patients with cancers of the head and neck. Secondly, we wanted to know whether there were any common denominators that could help identify those patients who were at risk of early death. We also aimed to increase awareness about this subgroup of patients and achieve our ultimate goal of improving the management of cancer of the head and neck.

Many studies have dealt with survival but, as far as we know, no population-based studies have focused on the patients with cancer of the head and neck who die within six months of diagnosis.

Patients, material, and methods

We obtained our data from the Swedish Head and Neck Cancer Register (SweHNCR) (Ethics Committee approval; Gothenburg, number 299–14, T230-17), which is funded by the Swedish government, and covers 98.5% of all Swedish patients with cancers of the head and neck, when cross-referenced with Sweden’s National Board of Health and Welfare [16]. The total number of consecutive affected Swedish patients during the period 2008–2015 in the SweHNCR with one-year follow-up was 9733 (Fig 1, Table 1).

Fig 1

Flow chart for treatment and early death for included patients.

Table 1

Prevalence.

	Prevalence n (%) Total number = 9733
Diagnosis
Lip	863 (8.9)
Oral cavity	2714 (27.9)
Oropharynx	2524 (25.9)
Nasopharynx	234 (2.4)
Hypopharynx	455 (4.7)
Larynx	1300 (13.4)
Nose/sinuses	477 (4.9)
Salivary glands	795 (8.2)
CUP*	355 (3.6)
Incomplete diagnosis	16 (0.2)
Sex
Male	6244 (64.2)
Female	3489 (35.8)
Age continuous (mean (sd))	66.8 (13.3)
Age groups
0–39	295 (3.0)
40–49	644 (6.6)
50–59	1751 (18.0)
60–69	3165 (32.5)
70–79	2217 (22.8)
80-	1661 (17.1)
Stage
I	2564 (26.3)
II	1638 (16.8)
III	1265 (13.0)
IV	3711 (38.1)
Missing	200 (2.1)
(CUP lacks stage)	355 (3.6)
WHO score
0	6223 (63.9)
1	1279 (13.1)
2	608 (6.2)
3	344 (3.5)
4	133 (1.4)
Missing	1146 (11.8)
Treatment intent
Curative	8751 (89.9)
Palliative	882 (9.0)
Missing	100 (1.0)
Given treatment
Surgery	2370 (24.4)
Surgery+postoperative RT**	1295 (13.3)
Preoperative RT+ Surgery	306 (3.1)
Surgery +RT	356 (3.7)
RT	2625 (27.0)
CRT***	1169 (12.0)
Other	818 (8.4)
No treatment	245 (2.5)
Missing	549 (5.6)

*CUP = Cancer of unknown primary

**RT = Radiotherapy

***CRT = Chemoradiotherapy

Nine sites of tumours were included: lip (C00.0–2, C00.6, C00.8, C00.9), oral cavity (C00.3, C00.4, C02, C03, C04, C05, C06), oropharynx (C01.9, C05.1, C05.2, C05.8, C05.9, C09, C10), nasopharynx (C11), hypopharynx (C12, C13), larynx (C10.1, C32), nose (C30.0) and nasal sinuses (C31), salivary glands (C07, C08), and head and neck cancer of unknown primary (C77.0), Table 1. Malignant tumours located in the thyroid, the parathyroid glands, or the oesophagus were not included. Data that are reported to SweHNCR include: incidence, diagnosis, time to treatment, sex, age, survival, WHO score, TNM classification, stage, follow up, and recurrence. The Eastern Cooperative Oncology Group score, also called the WHO score runs from 0 to 5, with 0 denoting perfect health and 5 death [17]. The WHO score was rated 0–4, Table 2, indicating the physical performance of the patient. An overall increase in incidence of 22% was noted from 2008 (n = 1211) to 2015 (n = 1473) for head and neck cancer in Sweden, which corresponds to a yearly increase of 2.9% [16].

Table 2

WHO score.

Grade	Explanation of activity
0	Fully active, able to carry on all pre-disease performance without restriction
1	Restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature, e.g., light house work, office work
2	Ambulatory and capable of all selfcare but unable to carry out any work activities. Up and about more than 50% of waking hours
3	Capable of only limited selfcare, confined to bed or chair more than 50% of waking hours
4	Completely disabled. Cannot carry on any selfcare. Totally confined to bed or chair
5	Dead

Treatment of these cancers in Sweden is centralised to the university hospitals, but a few second-level hospitals are able to give oncological treatment, and National Healthcare and Social Security systems are offered equally to all inhabitants. All patients were examined by either a head and neck surgeon or an oncologist after completion of treatment to evaluate its efficacy.

Statistical methods

The primary outcome in this study was early death. The relative risk of death within 6 months of diagnosis between different groups was estimated using univariable and multivariable Poisson regression [18]. Robust variance estimates were used with Poisson models to obtain valid confidence intervals. No censuring occurred within the first 12 months. Observed survival was calculated using the Kaplan-Meier method [19]. Survival time was calculated from date of diagnosis to date of death if death occurred before 24 months of diagnosis, censored 24 months of diagnosis or earlier if end of follow-up. The patients were diagnosed in the period 2008 to June 2015 and had a follow-up until June 2016. In the survival analyses, showing probability of survival for patients with HNC based on treatment intent, tumour site, and treatment modality (Figs 2–4), patients diagnosed after June 2014 were censored at end of follow-up before reaching a 24 months follow-up if they still lived. To perform tests of differences in age, stage and WHO score between patients dead within 0–6 months, and 7–12 months compared to those alive after 12 months (Table 3), the two-sample t-test and a nonparametric test for trend across ordered groups developed by Cuzick [20] were used. Exact binominal confidence intervals were estimated for proportions [21]. A p-value of < 0.05 was considered to be statistically significant. All statistical analyses were performed using Stata 15.1 [22].

Fig 2

Probability of survival for patients with HNC based on treatment intent.

Fig 4

Probability of survival based on treatment modality.

Table 3

WHO/Stage and survival during the first year after diagnosis.

	Patients dying within:				Alive 12 months after diagnosis
	0–6 months	p-value*	7–12 months	p-value*
Age at diagnosis, years; mean (SD)	75.7 (11.1)	<0.001	71.8 (12.2)	<0.001	65.3 (13.1)
WHO		<0.001		<0.001
0	147 (15.9)		333 (41.0)		5743 (71.8)
1	166 (18.0)		161 (19.8)		952 (11.9)
2	158 (17.1)		130 (16.0)		320 (4.0)
3	167 (18.0)		63 (7.8)		114 (1.4)
4	107 (11.6)		15 (1.8)		11 (0.1)
Unknown	180 (19.5)		111 (13.6)		855 (10.7)
Stage; n (%)		<0.001		<0.001
I	52 (5.6)		64 (7.9)		2448 (30.6)
II	70 (7.6)		101 (12.4)		1467 (18.4)
III	124 (13.4)		119 (14.6)		1022 (12.8)
IV	587 (63.5)		477 (58.7)		2647 (33.1)
Missing	92 (10.0)		52 (6.4)		411 (5.1)

*p-value for dead within 6 months as compared to those alive after 12 months

Results

The risk for death within six months has been divided into patient-, tumour and treatment- related factors. Of the whole group of 9733 patients, 925 died within six months of diagnosis (9.5%). Among the 9633 patients with a treatment decision at a multidisciplinary tumour board meeting, 9.1% died within six months, leaving a high rate of early death among patients for whom details of treatment were missing (n = 51, 51%) (Fig 1).

Patient-related factors

Sex: 6244 patients were male and 3489 female, Table 1. There was no significant survival difference between sexes where 589 men (9.4%) and 336 women (9.6%), died within six months of diagnosis (p = 0.75), Table 4.

Table 4

Univariate analysis, targeted variable is death within 6 months.

Variable	Dead within 6 months/N	Dead within 6 months % (95% KI)	RR (95% KI)	P*
Diagnosis
Lip	22/863	2.5 (1.6–3.8)	0.23 (0.15–0.36)	<0.001
Oral Cavity	298/2714	10.9 (9.8–12.2)	1.0	-
Oropharynx	217/2524	8.6 (7.5–9.8)	0.78 (0.66–0.92)	0.004
Nasopharynx	10/234	4.3 (2.1–7.7)	0.39 (0.21–0.72)	0.003
Hypopharynx	106/455	23.3 (19.5–27.5)	2.12 (1.74–2.59)	<0.001
Larynx	105/1300	8.1 (6.6–9.7)	0.74 (0.59–0.91)	0.005
Nose/Sinuses	66/477	13.8 (10.9–17.3)	1.26 (0.98–1.61)	0.068
Salivary Gland	51/795	6.4 (4.8–8.3)	0.58 (0.44–0.78)	<0.001
CUP	46/355	13.0 (9.6–16.9)	1.18 (0.88–1.58)	0.263
Sex
Men	589/6244	9.4 (8.7–10.2)	1.0	-
Women	336/3489	9.6 (8.7–10.7)	1.02 (0.90–1.16)	0.750
Age (continuously)	925/9733	9.5 (8.9–10.1)	1.06 (1.05–1.06)	<0.001
Age groups
0–39	0/295	0.0 (0.0–1.2)	-	-
40–49	10/644	1.6 (0.7–2.8)	0.44 (0.23–0.85)	0.015
50–59	62/1751	3.5 (2.7–4.5)	1.0	-
60–69	226/3165	7.1 (6.3–8.1)	2.02 (1.54–2.67)	<0.001
70–79	266/2217	12.0 (10.7–13.4)	3.39 (2.59–4.44)	<0.001
80-	361/1661	21.7 (19.8–23.8)	6.14 (4.73–7.97)	<0.001
Stage
I	52/2564	2.0 (1.5–2.7)	1.0	-
II	79/1638	4.3 (3.3–5.4)	2.11 (1.48–3.00)	<0.001
III	124/1265	9.8 (8.2–11.6)	4.83 (3.52–6.63)	<0.001
IV	587/3711	15.8 (14.7–17.0)	7.80 (5.90–30.3)	<0.001
WHO score
0	147/6223	2.4 (2.0–2.8)	1.0	-
1	166/1279	13.0 (11.1–14.9)	5.49 (4.44–6.80)	<0.001
2	158/608	26.0 (22.5–29.7)	11.0 (8.93–13.6)	<0.001
3	167/344	48.5 (43.2–54.0)	20.6 (16.9–24.9)	<0.001
4	107/133	80.5 (72.7–86.8)	34.1 (28.4–40.8)	<0.001
Unknown	180/1146	15.7 (13.6–18.0)	6.65 (5.40–8.19)	<0.001
Treatment intention
Curative	396/8751	4.5 (4.1–5.0)	1.0	-
Palliative	478/882	54.2 (50.8–57.5)	12.0 (10.7–13.4)	<0.001
Given treatment
Surgery only	82/2370	3.5 (2.8–4.3)	1.0	-
Surgery+postoperative RT**	24/1295	1.8 (1.2–27.4)	0.54 (0.34–0.84)	0.007
Preoperative RT+surgery	3/306	1.0 (0.2–2.8)	0.28 (0.09–0.89)	0.031
Surgery+RT	5/356	1.4 (0.5–3.3)	0.41 (0.17–0.99)	0.049
RT onlyRT+Chemotherapy	300/262543/1169	11.4 (10.2–12.7)3.7 (2.7–4.9)	3.30 (2.60–4.19)1.06 (0.74–1.53)	<0.0010.741
Other treatment	38/818	4.8 (3.4–6.5)	1.34 (0.92–1.96)	0.125
No treatment	169/245	52.4 (48.9–55.9)	13.7 (10.9–17.3)	<0.001

*P-value is for Relative Risk

**RT = radiotherapy

Age: the mean (range) age at diagnosis was 67 (8–102) years, and the risk to die within six months after diagnosis increased with older age, Table 3, and with 2.3% for every year older the patient became (p<0.001), Tables 5 and 6.

Table 5

Multivariable analysis, targeted variable is death within 6 months (n = 9098).

	Curative and palliative treatment(n = 9098)		Curative treatment(n = 8314)
Variable	RR (95% CI)	P*	RR (95% CI)	P*
Diagnosis
Lip	0.76 (0.49–1.19)	0.227	0.69 (0.41–1.17)	0.167
Oral Cavity	1.0	-	1.0	-
Oropharynx	0.96 (0.82–1.13)	0.645	0.91 (0.68–1.20)	0.500
Nasopharynx	0.61 (0.33–1.12)	0.109	0.38 (0.09–1.52)	0.171
Hypopharynx	1.24 (1.03–1.50)	0.024	1.27 (0.90–1.78)	0.179
Larynx	1.09 (0.89–1.33)	0.424	1.09 (0.81–1.47)	0.583
Nose/Sinuses	0.94 (0.75–1.19)	0.624	1.02 (0.66–1.58)	0.930
Salivary Gland	0.59 (0.45–0.79)	<0.001	0.63 (0.41–0.99)	0.046
Sex
Men	1.0	-	1.0	-
Women	0.96 (0.85–1.09)	0.518	0.81 (0.65–1.01)	0.060
Age (continuous)	1.023 (1.017–1.029)	<0.001	1.039 (1.029–1.049)	<0.001
Stage
I	1.0	-	1.0	-
II	1.68 (1.17–2.40)	0.005	1.59 (1.03–2.48)	0.030
III	2.82 (2.02–3.93)	<0.001	3.22 (2.14–4.85)	<0.001
IV	3.71 (2.71–5.10)	<0.001	4.61 (3.17–6.70)	<0.001
WHO score
0	1.0	-	1.0	-
1	3.37 (2.67–4.25)	<0.001	3.25 (2.45–4.31)	<0.001
2	4.53 (3.54–5.80)	<0.001	4.82 (3.51–6.62)	<0.001
3	5.68 (4.40–7.32)	<0.001	8.29 (5.92–11.6)	<0.001
4	6.77 (5.22–8.78)	<0.001	16.8 (10.4–21.4)	<0.001
unknown	3.47 (2.73–4.40)	<0.001	3.08 (2.25–4.23)	<0.001
Treatment intention
Curative	1.0	-
Palliative	3.16 (2.68–3.72)	<0.001

*P-value is for relative risk

Table 6

Multivariate analysis, targeted variable is death within 6 months, including all tumour sites (n = 9098).

Variable	RR (95% CI)	P*
Diagnosis
Lip	0.76 (0.49–1.19)	0.227
Oral Cavity	1.0	-
Oropharynx	0.96 (0.82–1.13)	0.645
Nasopharynx	0.61 (0.33–1.12)	0.109
Hypopharynx	1.24 (1.03–1.50)	0.024
Larynx	1.09 (0.89–1.33)	0.424
Nose/Sinuses	0.94 (0.75–1.19)	0.624
Salivary Gland	0.59 (0.45–0.79)	<0.001
Sex
Men	1.0	-
Women	0.96 (0.85–1.09)	0.518
Age (continuous)	1.023 (1.017–1.029)	<0.001
Stage
I	1.0	-
II	1.68 (1.17–2.40)	0.005
III	2.81 (2.01–3.92)	<0.001
IV	3.74 (2.71–5.13)	<0.001
WHO score
0	1.0	-
1	3.37 (2.67–4.25)	<0.001
2	4.53 (3.54–5.80)	<0.001
3	5.68 (4.40–7.32)	<0.001
4	6.77 (5.22–8.78)	<0.001
unknown	3.47 (2.73–4.40)	<0.001
Treatment intention
Curative	1.0	-
Palliative	3.16 (2.68–3.72)	<0.001

*P-value for relative risk of death within 6 months

-

-

WHO score: only 147 patients (2.4%) with a WHO score of 0 died within 6 months, compared with 166 of patients (13%) with a WHO score of 1. Of 133 patients with a WHO score of 4, 107 died within 6 months (80.5%). The relative risk of death was 3.37 times higher for patients with a WHO score of 1 (237%) compared with those who scored 0, Table 6.

Tumour-related factors

Site: the probability of survival varied depending on the site of the primary tumour (Fig 3). The worst prognosis was found among patients with hypopharyngeal cancer, of whom 106 died within six months (23%), Table 4. Patients with hypopharyngeal cancers had an increased relative risk for early death of 1.24 compared with those with cancer of the oral cavity (p = 0.024), Table 6.

Fig 3

Probability of survival based on site of tumor.

Salivary gland cancer is morphologically considered to be a separate entity of head and neck cancer. Therefore, we performed analyses excluding salivary gland cancer in order to evaluate differences in risk of death within six months. The exclusion of salivary gland cancer did not significantly affect the outcome, Tables 6 and 7.

Table 7

Multivariate analysis, targeted variable is death within 6 months, excluding salivary gland cancer (n = 8332).

Variable	RR (95% CI)	P*
Diagnosis
Lip	0.77 (0.49–1.21)	0.255
Oral Cavity	1.0	-
Oropharynx	0.97 (0.83–1.13)	0.663
Nasopharynx	0.60 (0.33–1.11)	0.103
Hypopharynx	1.24 (1.03–1.50)	0.023
Larynx	1.09 (0.89–1.34)	0.383
Nose/Sinuses	0.94 (0.75–1.19)	0.619
Sex
Men	1.0	-
Women	0.99 (0.87–1.12)	0.863
Age (continuous)	1.022 (1.016–1.028)	<0.001
Stage
I	1.0	-
II	1.75 (1.21–2.52)	0.003
III	2.88 (2.05–4.07)	<0.001
IV	3.72 (2.68–5.16)	<0.001
WHO score
0	1.0	-
1	3.40 (2.68–4.31)	<0.001
2	4.55 (3.54–5.86)	<0.001
3	5.59 (4.30–7.26)	<0.001
4	6.48 (4.96–8.45)	<0.001
unknown	3.41 (2.66–4.37)	<0.001
Treatment intention
Curative	1.0	-
Palliative	3.28 (2.77–3.88)	<0.001

*P-value is for relative risk

-

Stage: we found an association between tumour stage and the risk of death within six months, Table 4. Among the 925 patients who died within six months, 587 had stage IV disease (64%), and only 52 patients (6%) who died within six months had stage I tumours. The distribution between tumour stages also varied depending on the site of the primary tumour with for example, 699 of all cancers of the lip (81%) being stage I, and only 17 (2%) stage IV. Stage IV disease was found at diagnosis in two-thirds of patients with oropharyngeal (n = 1691, 67%) and hypopharyngeal (n = 305, 66%) cancers. Most of the 152 patients with nasopharyngeal cancer (65%) were diagnosed with stage III or IV disease. In total, a patient with stage IV disease had a 3.7 times higher relative risk of death within six months than patients with stage I disease (p<0.001), Table 6.

TNM class

T: low T class correlated with a better prognosis. Only 77 patients with a T1 tumour (2.4%) died within six months, while the six-month mortality among T4 patients was 455 (23%).

N: 5777 of patients (59%) had no neck metastases, whereas 3791 did (39%). Data about N class were missing in 2% of patients. A total of 399 patients with N negative necks (7%) died within six months, as did 482 patients (13%) with N positive necks.

M: Of patients without distant metastases 753 (8%) died within six months. A total of 256 patients (3%) had distant metastasis at the time of diagnosis, and of these 125 (49%) died within six months.

Treatment-related factors

Curative compared with palliative intent

Treatment with curative intent was recommended to 8751 patients (90.8%), and palliative treatment to 882 (9.2%) (Fig 2). Among those whose treatment was potentially curative, 396 (4.5%) died within six months (Fig 1, Table 4). Risk of early death increased among patients with curative treatment decision and high WHO score and/or high stage, Table 5. Of those for whom palliative treatment or “best supportive care” was recommended, 478 died within six months (54.2%). During the period 2008–2014 where 8271 patients were studied and a two-year follow up was available interestingly 104 patients (14%) for whom palliative treatment had been recommended survived for more than two years.

Treatment: Modality

A total of 2370 patients (24%) were treated by primary surgery alone, and primary surgery was combined with postoperative radiotherapy (RT) in 1295 (16%) (Fig 4).

In the group who had primary surgery alone, 82 (3.5%) died within six months. When it was combined with postoperative RT, 24 (1.8%) died within six months. Preoperative RT or chemoradiotherapy (CRT) together with surgery resulted in the death of eight patients (1%) within six months. Patients given either RT or CRT alone had a mortality of 343 within six months (9%), Table 4.

Treatment: Outcome

Patients in whom locoregional control was achieved after primary treatment had a low risk of early death, and only 52 died within six months (0.7%). Among those for whom primary treatment failed, 381 (35%) died within six months. For a total of 1236 patients (12.7%), data about locoregional control after treatment were missing, and for these patients the mortality was 493 within six months (40%).

Patient–tumour, and treatment-related factors

Taken together, patient-, tumour and treatment- related factors all influenced the risk of death within six months. In order to clarify the extent of how vastly it affects this risk, a comparison between a low-risk patient (55-year old, oral cavity cancer, WHO 0, stage 1, curative treatment) and a high-risk patient (75-year old, hypohryngeal cancer, WHO 2, stage III, curative treatment) was made and showed a 24.9 fold risk increase to die within six months for the high-risk patient (95% CI: 16.3–38.0, p<0.001). If the low-risk patient however, had a stage III disease instead of stage I, the risk of early death was still 8.8 times higher for the 75-year old man (95% CI: 6.47–12.06, p<0.001).

Recurrence

A total of 8437 of the patients (87%) had no recurrence within the first 12 months.

Discussion

Studies on the incidence and aetiology of early death among patients with cancers of the head and neck are still scarce, and to our knowledge this is not only the first official report to describe the SweHNCR data on 9733 such patients, but also the first attempt in Europe to investigate early death among patients in a nationwide population-based series. A previous Swedish study based on the data from 6785 such patients during the period 2008–2013 indicated a risk of death within six months of 665 (9.8%) [9]. In the present study we found a slightly reduced risk of early death of 925/9733 patients during the extended period 2008–2015 (9.5%). If data from the group who were given palliative treatment were excluded, the remainder (those given curative treatment) had a six-month mortality of 396 (4.5%). Our results suggest that extra precautions should be taken for patients with tumours located in the hypopharyngeal area, stage IV disease, older age, and increased WHO score.

Five-year mortality from cancers of the head and neck is high, even though mortality in general has decreased in Northern Europe. One reason for this is the increase in the number of HPV-induced tumours [23, 24]. Patients in the SweHNCR have an overall relative five-year survival rate of 6521 (67%), which includes all sites [16]. It has been reported that patients with early stage cancers of the head and neck can have five-year survival rates of up to 80%, whereas those with late-stage disease have reduced five-year survival of about 20% [25]. Early mortality (within six months of diagnosis) has been studied in other cancers [26, 27] but to our knowledge still remains unexplored for the head and neck.

Our results indicate that the risk of early death was associated with patient- related and treatment-related factors as well as tumour-related ones. Eskiizmir et al in 2017 showed differences in survival by sex in early-stage laryngeal cancer [21], but we could not confirm a sex difference in stage I-II laryngeal cancer in our series as 336 (9.6%) women and 590 (9.4%) men died within six months (p = 0.518).

Our results showed that older patients with cancers of the head and neck had a poorer prognosis than younger ones, and that the risk of early death increased with older age. Other authors have also found that mortality increases for older patients after resection and CRT [22, 23]. A Danish study by Johansen et al, found that survival improved from 1980 to 2012 for all age groups up to 79 years of age, but this effect was less pronounced for patients over 80 [24]. Prediction of survival among older patients must be based on available demographic information, and the details that are recorded differ among countries. We found a total of 1661 patients over 80 years old, of whom 361 (22%) died within six months. Multivariate analysis showed that patients over 80 were more likely to have coexisting conditions, but up to 75 years of age most patients who had a WHO score of 0–1 were less likely to die early. Older patients with cancer given curative treatment, therefore, should be particularly carefully monitored to minimise complications and failure of treatment, and ultimately, the risk of early death.

A high WHO score has been reported to be an indicator of a poor prognosis in both leukaemia and colon cancer [25, 26]. A total of 107 patients with cancer of the head and neck and a WHO score of 4 (80.5%) died within six months, Table 4. Among patients who had a WHO score of 4, 109 (82%) were given palliative treatment because of the severity of their disease, or coexisting conditions, or both.

Cancers of the head and neck are a heterogeneous group of malignancies with varied biology. The aggressiveness of the tumour, coexisting conditions, and presenting symptoms can all affect the outcome. For example, patients who presented with easily-detected symptoms such as those of lip cancer (801 of 836 tumours (93%) were diagnosed in stages I-II) or laryngeal cancer (819 of 1280 tumours, (63%) were diagnosed in stages I-II), had better survival than those with cancers at sites associated with more subtle symptoms, such as the hypopharynx. Patients with hypopharyngeal tumours tended to be detected at later stages (293 of 341 tumours (86%) were diagnosed at stages III-IV). If the tumours were discovered later, patients were at higher risk of early death, findings confirmed elsewhere [28]. Stage IV cancer had a 3.7 fold risk of death within six months, compared with those with stage I (p<0.001).

Patients with advanced tumours had the shortest survival, independent of the site of the tumour. In accordance with a number of previous publications [29, 30], our results showed that advanced tumours and metastatic disease were both associated with an increased risk of early death. However, there was a significant difference in the risk of early death depending on the site of the tumour–for example, a patient with hypopharyngeal cancer had a 24% increased risk compared to patients with tumours in the oral cavity (p<0.024). This is in line with the results from more than 34 000 patients from Denmark in 2014, among whom the probability of five-year survival for those with hypopharyngeal cancer was only 22.8% compared with 62.4% for all other sites in the head and neck [31].

Patients being given treatment with curative intent lived longer than those who were given palliation, who were at a 12-fold risk of death within six months (p<0.001). However, it should be noted that 223 of patients who had palliative treatment (25%) lived for longer than 12 months, and 104 (14%) for 24 months or more, which is longer than could be expected. This is particularly interesting as other authors have found that without any treatment the median survival was usually less than four months [3]. These data suggest that even if the intention of treatment is not to cure but to palliate, palliative treatment can be effective for a substantial subgroup, and even prolong time to death.

Patients who were operated on primarily had smaller tumours (1975 of them (87%) had stage I or II disease). They also had tumours in earlier stages than patients who were treated by radiotherapy or CRT together with resection, and were therefore more likely to survive. It has been reported that higher doses of chemotherapy and radiotherapy together with extensive resection will increase the risk of lethal complications [32]. Patients with advanced cancers of the head and neck who were admitted to hospital for more than five days during or after radiotherapy had worse locoregional control, progression-free survival, and overall survival [33].

In the group of patients (n = 245) who were given no active treatment, 89 (36%) had been recommended for curative treatment at a multidisciplinary tumour board meeting, and 43 (48%) of them died within six months. It could be speculated that these patients chose not to undergo treatment, were too ill to start, or died before treatment could begin.

For the 1236 patients (12.7%) with missing data about locoregional control at first follow-up, the mortality within six months was high 493 (40%). At the first check-up after treatment it can be difficult to decide whether a patient has a residual tumour, or just divergent anatomy/mucosa after intense treatment. As mortality is so high in this group, clinicians could be more cautious, and suspect recurrence as soon as there is any doubt.

A limitation of this study was that the actual cause of death is not recorded in the SweHNCR. It is recorded if the patient is considered to be free of tumour, but the immediate cause of death is not reported. Questions about whether death was caused by the cancer itself, by treatment-related complications, coexisting conditions, or causes unrelated to the cancer, could therefore not be answered by this study. Other limiting factors include the lack of information about socioeconomic factors, alcohol consumption, smoking habits, and HPV status, which were not reported in the SweHNCR until 2015, but thereafter alcohol consumption, smoking habits, and HPV status have been included. Sawabe et al showed that a high alcohol consumption leads to significantly shorter overall survival for these patients [34]. Socioeconomic status, other life-style factors, and depression have also been suggested as risk factors for early death [35-39]. As these data were not reported in the SweHNCR during the study period, we were unable to evaluate them.

Smoking is decreasing in Sweden, and roughly 9% of the total population now smoke daily compared with 15% 10 years ago (according to the Public Health Agency in Sweden). One could speculate that over time this would result in a decrease in mortality as well. The incidence of HPV-related cancer in the oropharynx is increasing, which can influence and decrease mortality [40] [41]. However, the large group of patients, its homogeneity in gross socioeconomic and health-related factors from a well-defined geographical area, and the degree of coverage of 98.5%, compensate for some of the limiting factors.

We found that older age, advanced stage, and a hypopharyngeal tumour, together with higher WHO score and palliative treatment were significant risk factors for early death. These findings contribute to the knowledge of a risk profile for early death among patients with cancers of the head and neck, and we think it is important to identify these patients early to optimise treatment, control symptoms, and reduce the number who die early. This study could be used as a base for future studies of subgroups within the palliative and curative treatment groups, so that we are able to understand better which patients are at greatest risk of early death.

Taken together, this study indicates that from clinical relevant information, risk profiles can be identified, and a high-risk patient had a considerable increased risk to die within six months compared to a low-risk patient. Understanding and identifying high-risk patients for early death is therefore important for physicians when recommending treatment to patients, curative or palliative.

Conclusion

We identified a risk profile for early death in this population-based register study of 9733 patients with cancers of the head and neck. The risk depended on patient’s age, WHO score, site of primary tumour, stage, and intention to treat. Our most important findings were that the risk of early death increased with 2.3% for every year older the patient became (p<0.001). The relative risk of death was 3.37 times higher for patients with a WHO score of 1, compared with a WHO score of 0 (p<0.001). Patients with stage IV disease had a 3.7 times greater risk of death within six months than those with stage I disease (p<0.001), and the group given palliative treatment had a 12 times higher risk of early death (p<0.001) than the others. If the tumour was located in the hypopharynx the risk of dying within six months increased by 24% compared with that of a tumour in the oral cavity (p = 0.024). Even if the intention of treatment was not cure but palliation, palliative treatment was shown to have an effect for a substantial group, and possibly prolong time to death.

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

10 Jul 2019

PONE-D-19-15950

Early mortality after diagnosis of cancer of the head and neck – A population-based nationwide study

PLOS ONE

Dear MD, PhD Farnebo,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

For acceptance the revision will need to include modification of the methods section and improved statistical review as documented by reviewer one.  I also agree with reviewer 2 that salivary gland cancers can have entirely different histologies and therefore prognosis from SCC and therefore would consider excluding them or doing a sub-analysis excluding them that is reported in the manuscript.  Finally, reviewer 2 mentions that it may be helpful to do a sub-analysis of those receiving curative treatment to see who does not benefit - while not mandatory for publication, I agree could be an interesting finding.

We would appreciate receiving your revised manuscript by Aug 24 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Jessica D McDermott, MD, MSCS

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files

3. Please ensure that you refer to Figure 4 in your text as, if accepted, production will need this reference to link the reader to the figure.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Overall an interesting article which some gaps in the methodological presentation. Also, the presentation of the results could benefit from being more precise because otherwise they sound in part trivial, which is not really the case.

Methods:

The description of the statistical analysis is presented in four lines. Given that by using this method all results have been obtained this looks odd. More details need to be added and also relevant literature needs to be cited. So far only a citation to Stata is provided which is a software package. Please cite original papers introducing the used methods.

It seems no Cox proportional hazard model has been used. Why is it not appropriate in this case?

What statistical test has been used to identify a difference in the survival times? Why has this test been selected (discuss and compare with alternatives)?

A discussion about censoring is completely missing. How has this been handeled? What type of censoring is present?

Results:

The figures of the survival curves do not include information about censoring.

For all statistical test, please add the actual p-values and not just p<0.xyz.

Conclusions:

Identification of patients at i 52 ncreased risk of early death shows that older

53 patients with advanced disease, increased WHO score, primary tumour in the hypopharynx,

54 and those given palliative treatment, are more likely than the others to die from head and neck

55 cancer within six months of diagnosis.

This does not sound surprising. What aspect of this is none trivial? Adding quantitive results would make this actually interesting.

We think that it would be

327 beneficial to identify patients at risk early and bear this information in mind when planning

328 their treatment and follow up.

I wonder if there is any disease for which this would NOT be beneficial? If there is none, this is trivial.

Reviewer #2: The authors reported that old age, advanced stage, poor performance status and no curative treatment were the clinical risk factors associated with early death 6 months after diagnosis of head and neck cancer using a nationwide data although the incidence and case number of head and neck cancer are relatively smaller than those of other countries. It is really helpful for clinicians to think about who are not the candidate to recieve the curative treatment, especially in old people or people with poor performance so that will be better if this study describes or analysis the outcome of these people with curative treatment.

Salivary gland cancer is also a totally different cancer from other head and neck cancer so the story is possibly different if salivary gland cancer is excluded.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Aug 2019

Responses to the Reviewers

Thank you for the received Reviewer reports and the opportunity to revise our manuscript. We have now improved the manuscript to include a more thorough Methods section and we also upgraded the description of the statistical review. Furthermore, we included a new sub-analysis excluding salivary gland cancers, and included a cohort with only curative treatment intent, as recommended by Reviewer 2. In order to meet the request on a sub-analysis of those receiving curative treatment to see who does not benefit, we also added examples of a low-risk patient and a patient with high risk of six-month mortality, in order to emphasize the conclusion in the manuscript.

Our detailed responses to the Reviewers are listed below in blue, and corrections of the manuscript appear highlighted in yellow in the revised version.

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files

Response: Tables are now included in the revised version of the main manuscript, and individual files removed.

3. Please ensure that you refer to Figure 4 in your text as, if accepted, production will need this reference to link the reader to the figure.

Response: This has now been done accordingly and we refer to Figure 4 in the Results section (page 10, paragraph 2, row 198).

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Response: We have improved the Methods section accordingly and we included new statistical calculations to improve the manuscript, page 7.

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

Response: We have taken measures to improve the clarity in the Methods section. Statistical analyses have been described and explained in more detail and new calculations have been added, and are highlighted in yellow on pages 7, 9, 10, 11 and 16.

________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Response: According to the approval of this study by the Swedish Ethical Review Authority we are not allowed to make data available in any form other than aggregated data.

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

________________________________________

5. Review Comments to the Author

Reviewer #1: Overall an interesting article which some gaps in the methodological presentation. Also, the presentation of the results could benefit from being more precise because otherwise they sound in part trivial, which is not really the case.

Response: We have improved the methodological presentation in text and tables. In order to avoid the Results section from sounding trivial, we have included an example showing difference in RR between a low-risk patient and a high-risk patient, page 11. We hope this addition emphasizes the value of the conclusions in this article.

Methods:

The description of the statistical analysis is presented in four lines. Given that by using this method all results have been obtained this looks odd. More details need to be added and also relevant literature needs to be cited. So far only a citation to Stata is provided which is a software package. Please cite original papers introducing the used methods.

Response: The statistical analyses have been explained in more detail and relevant citations (see below) are now included in the revised version of our paper.

1. Schmidt CO, Kohlmann T. When to use the odds ratio or the relative risk? Int J Public Health 53 (2008) 165–167.

2. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. Journal of the American Statistical Association 53 (1958) 457–481.

3. Cuzick J. 1985. A Wilcoxon-type test for trend. Statistics in Medicine 4 (1985) 87-90.

4. Clopper CJ and Pearson ES. The use of confidence or fiducial limits illustrated in the case of binomial. Biometrika 26 (1934) 404-413.

5. StataCorp. 2017. Stata: Release 15. Statistical Software. College Station, TX: StataCorp LLC.

It seems no Cox proportional hazard model has been used. Why is it not appropriate in this case? What statistical test has been used to identify a difference in the survival times? Why has this test been selected (discuss and compare with alternatives)? A discussion about censoring is completely missing. How has this been handeled? What type of censoring is present?

Response: Cox proportional hazard model is mainly recommended when censoring needs handling. We have no censoring within the first 12 months of follow-up, which is why we instead used the Poisson-regression to analyze mortality within six months follow-up, a method suitable when no censoring needs handling.

The primary aim of this study was to analyze early mortality within 6 months. Since we had no censuring within 12 months follow-up the results were mainly presented as relative risk of death within 6 months follow-up.

Differences in the 24 months survival have been presented as an illustrative example as Kaplan-Meier-curves in Figure 2-4 but have not been analyzed further. There are censuring after 12 months for cases due to end of follow-up which have been taken care of by the KM-method. There are no ‘lost to follow-up cases’ within 24 months.

Results:

The figures of the survival curves do not include information about censoring.For all statistical test, please add the actual p-values and not just p<0.xyz.

Response: Information about censoring is to our knowledge, generally not presented in Kaplan-Meier diagrams. For clarity for the reader, we have not presented p-values of less than 0.001, we do however show 95%-confidence intervals, numbers that can give a deeper understanding than the p-value.

Conclusions:

Identification of patients at i 52 ncreased risk of early death shows that older

53 patients with advanced disease, increased WHO score, primary tumour in the hypopharynx,

54 and those given palliative treatment, are more likely than the others to die from head and neck

55 cancer within six months of diagnosis.

This does not sound surprising. What aspect of this is none trivial? Adding quantitive results would make this actually interesting.

Response: Comparison between a high-risk patient, a 75-year old with hypopharyngeal cancer WHO 2, stage III , curative treatment, and a low-risk patient, a 55-year old man with oral cavity cancer, WHO 0, stage I, curative treatment:

------------------------------------------------------------------------------

ded_6 | IRR Std. Err. z P>|z| [95% Conf. Interval]

-------------+----------------------------------------------------------------

(1) | 24.89902 5.375931 14.89 0.000 16.30797 38.01584

------------------------------------------------------------------------------

The relative risk RR of death within six months = 24.9 (95% CI: 16.3-38.0), p<0.001, for the high-risk patient compared to the low-risk.

If the 55-year old man instead had a stage III disease the RR = 8.83 (95% CI: 6.47 – 12.06), p<0.001

------------------------------------------------------------------------------

ded_6 | IRR Std. Err. z P>|z| [95% Conf. Interval]

-------------+----------------------------------------------------------------

(1) | 8.835795 1.402317 13.73 0.000 6.4737 12.05976

------------------------------------------------------------------------------

This means that a 75-year old hypopharyngeal cancer patient with WHO 2, stage III, had a 24.9 fold risk increase to die within six months as compared to a 55-year old with oral cavity cancer, WHO 0, stage 1. If the 55-year old however had a stage III disease, the risk of early death was 8.8 times higher for the 75-year old patient.

We think that it would be

327 beneficial to identify patients at risk early and bear this information in mind when planning

328 their treatment and follow up.

I wonder if there is any disease for which this would NOT be beneficial? If there is none, this is trivial.

We agree, this sentence is trivial, and it has now been removed from the manuscript and replaced with the high-risk/low-risk patient example above.

Reviewer #2: The authors reported that old age, advanced stage, poor performance status and no curative treatment were the clinical risk factors associated with early death 6 months after diagnosis of head and neck cancer using a nationwide data although the incidence and case number of head and neck cancer are relatively smaller than those of other countries. It is really helpful for clinicians to think about who are not the candidate to recieve the curative treatment, especially in old people or people with poor performance so that will be better if this study describes or analysis the outcome of these people with curative treatment.

Response: Table 5 has now been extended to include this information in Table 5a. Please find the new Table 5a, b, and c below. It can now be clearly visualized that higher WHO score, age, and stage all contribute to increased risk of death within 6 months in the patient group with curative treatment intent, information that clinicians should bear in mind at the multi-disciplinary tumour board meeting.

Table 5a. Multivariable analysis, targeted variable is death within 6 months (n=9098)

Curative and palliative treatment

(n=9098) Curative treatment

(n=8314)

Variable RR (95% CI) P*

RR (95% CI) P*

Diagnosis

Lip

Oral Cavity

Oropharynx

Nasopharynx

Hypopharynx

Larynx

Nose/Sinuses

Salivary gland

0.76 (0.49-1.19)

1.0

0.96 (0.82-1.13)

0.61 (0.33-1.12)

1.24 (1.03-1.50)

1.09 (0.89-1.33)

0.94 (0.75-1.19)

0.59 (0.45-0.79)

0.227

-

0.645

0.109

0.024

0.424

0.624

<0.001

0.69 (0.41-1.17)

1.0

0.91 (0.68-1.20)

0.38 (0.09-1.52)

1.27 (0.90-1.78)

1.09 (0.81-1.47)

1.02 (0.66-1.58)

0.63 (0.41-0.99)

0.167

-

0.500

0.171

0.179

0.583

0.930

0.046

Sex

Male

Female

1.0

0.96 (0.85-1.09)

-

0.518

1.0

0.81 (0.65-1.01)

-

0.060

Age (continuous) 1.023 (1.017-1.029) <0.001 1.039 (1.029-1.049) <0.001

Stage

I

II

III

IV

1.0

1.68 (1.17-2.40)

2.82 (2.02-3.93)

3.71 (2.71-5.10)

-

0.005

<0.001

<0.001

1.0

1.59 (1.03-2.48)

3.22 (2.14-4.85)

4.61 (3.17-6.70)

-

0.030

<0.001

<0.001

0

1

2

3

4

unknown

1.0

3.37 (2.67-4.25)

4.53 (3.54-5.80)

5.68 (4.40-7.32)

6.77 (5.22-8.78)

3.47 (2.73-4.40)

-

<0.001

<0.001

<0.001

<0.001

<0.001

1.0

3.25 (2.45-4.31)

4.82 (3.51-6.62)

8.29 (5.92-11.6)

16.8 (10.4-21.4)

3.08 (2.25-4.23)

-

<0.001

<0.001

<0.001

<0.001

<0.001

Treatment intent

Curative

Palliative

1.0

3.16 (2.68-3.72)

-

<0.001

Salivary gland cancer is also a totally different entity from other head and neck cancers so the story is possibly different if salivary gland cancer is excluded.

Response: We have now made an alternative analysis and excluded salivary gland cancer from the study cohort. It can be noted that no drastic statistical changes were seen when the patients with salivary gland cancer were excluded (see below Fig 5b, and 5c).

Table 5b. Multivariate analysis, targeted variable is

death within 6 months (n=9098)

Variable RR (95% KI) P*

Diagnosis

Lip

Oral Cavity

Oropharynx

Nasopharynx

Hypopharynx

Larynx

Nose/Sinuses

Salivary gland

0.76 (0.49-1.19))

1.0

0.96 (0.82-1.13)

0.61 (0.33-1.13)

1.24 (1.03-1.50)

1.09 (0.89-1.33)

0.94 (0.75-1.19)

0.59 (0.45-0.79)

0.227

-

0.645

0.109

0.024

0.424

0.624

<0.001

Sex

Male

Female

1.0

0.96 (0.85-1.09)

-

0.518

Age (continuous) 1.023 (1.017-1.029) <0.001

Stage

I

II

III

IV

1.0

1.68 (1.17-2.40)

2.81 (2.01-3.92)

3.74 (2.71-5.13)

-

0.005

<0.001

<0.001

0

1

2

3

4

missing

1.0

3.37 (2.67-4.25)

4.53 (3.54-5.80)

5.68 (4.40-7.32)

6.77 (5.25-8.83)

3.47 (2.73-4.40)

-

<0.001

<0.001

<0.001

<0.001

<0.001

Treatment intent

Curative

Palliative

1.0

3.16(2.68-3.72)

-

<0.001

Table 5c. Multivariate analysis, targeted variable is death within 6 months,

excluding salivary gland cancer patients (n=8332)

Variable RR (95% KI) P*

Diagnosis

Lip

Oral Cavity

Oropharynx

Nasopharynx

Hypopharynx

Larynx

Nose/Sinuses

0.77 (0.49-1.21)

1.0

0.97 (0.83-1.13)

0.60 (0.33-1.11)

1.24 (1.03-1.50)

1.09 (0.89-1.34)

0.94 (0.75-1.19)

0.255

-

0.663

0.103

0.023

0.383

0.619

Sex

Male

Female

1.0

0.99 (0.87-1.12)

-

0.863

Age (continuous) 1.022 (1.016-1.028) <0.001

Stage

I

II

III

IV

1.0

1.75 (1.21-2.52)

2.88 (2.05-4.07)

3.72 (2.68-5.16)

-

0.003

<0.001

<0.001

0

1

2

3

4

missing

1.0

3.40 (2.68-4.31)

4.55 (3.54-5.86)

5.59 (4.30-7.26)

6.48 (4.96-8.45)

3.41 (2.66-4.37)

-

<0.001

<0.001

<0.001

<0.001

<0.001

Treatment intent

Curative

Palliative

1.0

3.28 (2.77-3.88)

-

<0.001

________________________________________

6. PLOS authors have the option to publish the peer review history of their article. If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Submitted filename: Response_to_reviewers190813.docx

Click here for additional data file.

16 Sep 2019

Early mortality after diagnosis of cancer of the head and neck – A population-based nationwide study

PONE-D-19-15950R1

Dear Dr. Farnebo,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Jessica D McDermott, MD, MSCS

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: I think that the authors answered all quesitons and the revised manuscript is well done and can be accepted for publication

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

24 Sep 2019

PONE-D-19-15950R1

Early mortality after diagnosis of cancer of the head and neck – A population-based nationwide study

Dear Dr. Farnebo:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Jessica D McDermott

Academic Editor

PLOS ONE