Demographics
A total of 87 T4 HNSCC patients who received neoadjuvant chemo-immunotherapy between April 2020 and March 2023 were included in this study (Fig 1). Patient and tumor characteristics are presented in Table 1. The median age was 58 years (range from 23 to 75), and the majority of patients were male (92%). Forty-two patients had tumors located in the hypopharynx (48.3%), 22 in the oral cavity (25.3%), 17 in the larynx (19.5%), and 6 in the oropharynx (6.9%). T4a and T4b stages accounted for 86.2% and 13.8% of patients, respectively. Seventy-six patients (87.4%) had cervical lymph node metastasis, with 56 of them classified as clinical N2 stage (64.4%). No involvement of the carotid artery or cervical spine by cervical lymph nodes was observed.
Table 1 Baseline characteristics of all patients (n = 87)
Characteristics
|
Cases
|
Percentage (%)
|
Age, median (range), years
|
58(23,75)
|
|
Gender
|
|
|
Male
|
80
|
92.0
|
Female
|
7
|
8.0
|
Smoking
|
|
|
No
|
46
|
52.9
|
Yes
|
41
|
47.1
|
Alcohol use
|
|
|
Never
|
40
|
46.0
|
Ever
|
47
|
54.0
|
Subtypes
|
|
|
Oral cancer
|
22
|
25.3
|
Oropharyngeal cancer
|
6
|
6.9
|
Laryngeal cancer
|
17
|
19.5
|
Hypopharyngeal cancer
|
42
|
48.3
|
Histology
|
|
|
Well differentiated
|
8
|
9.2
|
Moderately differentiated
|
30
|
34.5
|
Poorly differentiated
|
24
|
27.6
|
Not otherwise specified
|
25
|
28.7
|
Pretreatment Clinical T stage
|
|
|
T4a
|
75
|
86.2
|
T4b
|
12
|
13.8
|
Pretreatment Clinical N stage
|
|
|
N0
|
11
|
12.6
|
N1
|
2
|
2.3
|
N2
|
56
|
64.4
|
N3
|
18
|
20.7
|
Pretreatment AJCC stage
|
|
|
IV A
|
61
|
70.1
|
IV B
|
26
|
29.9
|
Neoadjuvant cycle
|
|
|
2
|
11
|
12.6
|
3
|
61
|
70.1
|
4
|
15
|
17.2
|
Best radiographic response
|
|
|
CR
|
9
|
10.3
|
PR
|
60
|
69.0
|
SD
|
16
|
18.4
|
PD
|
2
|
2.3
|
Therapy after neoadjuvant chemo-immunotherapy
|
|
|
Palliative care
|
3
|
3.4
|
Surgery alone
|
3
|
3.4
|
(Chemo)radiotherapy
|
53
|
60.9
|
Surgery + (Chemo)radiotherapy
|
28
|
32.2
|
Treatment characteristics
Sixty-one patients (70.1%) received 3 cycles of neoadjuvant chemo-immunotherapy, 15 patients (17.2%) received 4 cycles, and 11 patients (12.6%) received 2 cycles. After neoadjuvant chemo-immunotherapy, 3 patients did not complete the intended subsequent treatment, including 2 patients with a disease progression (PD) and 1 patient with an stable disease (SD). Among the 84 patients who received subsequent treatment, 31 (36.9%) underwent surgical resection, while 53 (63.1%) received (chemo)radiotherapy without surgery. Of the 31 patients who underwent surgery, 11 achieved a pCR, and of these, three declined further (chemo)radiotherapy, while the remaining 28 patients received adjuvant (chemo)radiotherapy after surgery.
Response rates and survival outcomes
The overall response rate after neoadjuvant chemo-immunotherapy was 75.8%, comprising 9 patients (10.3%) with a complete response (CR) and 60 patients (69.0%) with a partial response (PR). Additionally, 16 patients (18.4%) had an SD, and 2 patients (2.3%) had a PD (Fig 2).
Among the 87 T4 HNSCC patients enrolled, 25 patients (28.7%) died after a median follow-up of 24.6 months (IQR, 17.1 – 33.2 months), and the median OS had not been reached. The 1-year and 3-year OS rates were 90.8% and 66.7%, respectively (Fig 3A). As shown in Supplementary Table 1, a total of 31 patients (35.6%) experienced an event or death as their first event, including 18 patients with local recurrence, 4 patients with distant metastasis, 4 patients who died, 3 patients with second primary cancer, and 2 patients with a PD that precluded definitive surgery. The 1-year and 3-year EFS rates were 81.3% and 56.0%, and the median EFS was either not reached (Fig 3B). Various salvage treatment approaches were employed after the first event occurred in these 26 patients, with the most common being chemotherapy-based strategies (17 patients, 65.4%). Additionally, 2 patients underwent salvage surgery, 1 patient received combined immunotherapy and targeted therapy, and the remaining 6 patients received only palliative care. Among 59 patients with tumor located in the hypopharynx or larynx, 41 (69.5%) achieved retention of throat and hypopharyngeal function.
Prognostic factors for OS and EFS
Results from univariable Cox regression analysis (Table 2 and Table 3) showed that 3-4 cycles of neoadjuvant treatment, pretreatment clinical N0 stage, and favorable radiographic response (CR and PR) were associated with better OS and EFS, while pretreatment clinical T4b stage had a negative prognostic impact on OS. Even after adjusting for variables found to be significant in univariable analyses, 3-4 cycles of neoadjuvant treatment, pretreatment clinical N0 stage, and favorable radiographic response (CR and PR) remained significantly associated with improved OS and EFS. Kaplan-Meier analyses also revealed significant differences in survival outcomes among patients with varying radiographic responses (Fig 4).
Table 2 Univariable and Multivariable Cox Proportional Hazards Regression for predicting Overall Survival
Variables
|
Univariable analysis
|
Multivariable analysis
|
HR (95% CI)
|
p value
|
HR (95% CI)
|
p value
|
Gender
|
|
0.596
|
|
|
Female
|
1 (referent)
|
|
|
|
Male
|
0.50 (0.04-6.38)
|
|
|
|
Age (years)
|
|
0.416
|
|
|
<60
|
1 (referent)
|
|
|
|
≥60
|
0.66 (0.24-1.82)
|
|
|
|
Smoking
|
|
0.716
|
|
|
No
|
1 (referent)
|
|
|
|
Yes
|
1.19 (0.46-3.09)
|
|
|
|
Alcohol use
|
|
0.146
|
|
|
Never
|
1 (referent)
|
|
|
|
Ever
|
2.34 (0.75-7.35)
|
|
|
|
Subtypes
|
|
0.742
|
|
|
Oral cancer
|
1 (referent)
|
|
|
|
Oropharyngeal cancer
|
0.58 (0.05-6.64)
|
|
|
|
Laryngeal cancer
|
0.32 (0.04-2.60)
|
|
|
|
Hypopharyngeal cancer
|
0.50 (0.09-2.77)
|
|
|
|
Histology
|
|
0.377
|
|
|
Well differentiated
|
1 (referent)
|
|
|
|
Moderately differentiated
|
1.51 (0.38-6.01)
|
|
|
|
Poorly differentiated
|
0.48 (0.08-2.77)
|
|
|
|
Not otherwise specified
|
1.50 (0.29-7.72)
|
|
|
|
Neoadjuvant cycle
|
|
0.038
|
|
0.017
|
2
|
1 (referent)
|
|
1 (referent)
|
|
3
|
0.80 (0.25-2.57)
|
|
0.44 (0.18-1.08)
|
|
4
|
0.05 (0.01-0.49)
|
|
0.05 (0.01-0.45)
|
|
Pretreatment Clinical T stage
|
|
0.032
|
|
0.052
|
T4a
|
1 (referent)
|
|
1 (referent)
|
|
T4b
|
4.72 (1.14-19.51)
|
|
3.23 (0.99-10.56)
|
|
Pretreatment Clinical N stage
|
|
0.030
|
|
0.048
|
N0
|
1 (referent)
|
|
1 (referent)
|
|
N1-3
|
10.98 (1.26-95.53)
|
|
8.10 (1.02-64.39)
|
|
Best radiographic response
|
|
0.000
|
|
0.001
|
CR + PR
|
1 (referent)
|
|
1 (referent)
|
|
SD + PD
|
6.12 (2.21-16.93)
|
|
4.82 (1.92-12.14)
|
|
Subsequent therapy
|
|
0.277
|
|
|
Surgical therapy
|
1 (referent)
|
|
|
|
Non-surgical therapy
|
2.32 (0.51-10.55)
|
|
|
|
Table 3 Univariable and Multivariable Cox Proportional Hazards Regression for predicting Event-free Survival
Variables
|
Univariable analysis
|
Multivariable analysis
|
HR (95% CI)
|
p value
|
HR (95% CI)
|
p value
|
Gender
|
|
0.628
|
|
|
Female
|
1 (referent)
|
|
|
|
Male
|
0.63 (0.06-6.32)
|
|
|
|
Age (years)
|
|
0.390
|
|
|
<60
|
1 (referent)
|
|
|
|
≥60
|
0.68 (0.28-1.64)
|
|
|
|
Smoking
|
|
0.817
|
|
|
No
|
1 (referent)
|
|
|
|
Yes
|
1.10 (0.49-2.46)
|
|
|
|
Alcohol use
|
|
0.107
|
|
|
Never
|
1 (referent)
|
|
|
|
Ever
|
2.26 (0.84-6.09)
|
|
|
|
Subtypes
|
|
0.946
|
|
|
Oral cancer
|
1 (referent)
|
|
|
|
Oropharyngeal cancer
|
0.55 (0.06-5.10)
|
|
|
|
Laryngeal cancer
|
0.69 (0.12-4.17)
|
|
|
|
Hypopharyngeal cancer
|
0.87 (0.21-3.60)
|
|
|
|
Histology
|
|
0.383
|
|
|
Well differentiated
|
1 (referent)
|
|
|
|
Moderately differentiated
|
1.20 (0.33-4.34)
|
|
|
|
Poorly differentiated
|
0.50 (0.11-2.28)
|
|
|
|
Not otherwise specified
|
1.35 (0.30-6.03)
|
|
|
|
Neoadjuvant cycle
|
|
0.043
|
|
0.031
|
2
|
1 (referent)
|
|
1 (referent)
|
|
3
|
0.97 (0.33-2.83)
|
|
0.60 (0.25-1.41)
|
|
4
|
0.12 (0.02-0.71)
|
|
0.11 (0.02-0.57)
|
|
Pretreatment Clinical T stage
|
|
0.089
|
|
0.166
|
T4a
|
1 (referent)
|
|
1 (referent)
|
|
T4b
|
2.95 (0.85-10.27)
|
|
2.07 (0.74-5.81)
|
|
Pretreatment Clinical N stage
|
|
0.018
|
|
0.026
|
N0
|
1 (referent)
|
|
1 (referent)
|
|
N1-3
|
13.28 (1.56-113.13)
|
|
9.94 (1.31-75.46)
|
|
Best radiographic response
|
|
0.001
|
|
0.001
|
CR + PR
|
1 (referent)
|
|
1 (referent)
|
|
SD + PD
|
5.48 (2.05-14.64)
|
|
4.08 (1.81-9.19)
|
|
Subsequent therapy
|
|
0.533
|
|
|
Surgical therapy
|
1 (referent)
|
|
|
|
Non-surgical therapy
|
1.50 (0.42-5.39)
|
|
|
|
We then analyzed the relationship between subsequent therapy after neoadjuvant chemo-immunotherapy and survival outcomes in patients that finished the intended treatment procedure. Among these 84 patients, 31 underwent surgery-based treatment, while the other 53 received nonsurgical treatment. Kaplan-Meier estimates of OS and EFS (Fig 5) showed no significant difference between the surgical group and the nonsurgical group (p = 0.839 and 0.862, respectively). When we focused on the 60 patients with a PR after neoadjuvant treatment, Kaplan-Meier analyses also revealed equivalent OS and EFS between the surgical group and the nonsurgical group.
Thirty-one patients (35.6%) underwent surgical resection, including twenty with oral cancer (all radical surgeries), nine with hypopharyngeal cancer (five total laryngectomies with lymph node dissection and four simple neck lymph node dissections), one with laryngeal cancer (total laryngectomy with lymph node dissection), and one with oropharyngeal cancer (tonsil resection with neck lymph node dissection). Among these 31 patients who underwent surgery after neoadjuvant therapy, 11 patients (35.5%) achieved a pCR. OS was higher in patients with a pCR although it did not reach statistical significance (log-rank 3.24, p = 0.072), while EFS was significantly higher in patients who achieved a pCR (log-rank 5.04, p = 0.025) compared to those without a pCR, as determined by Kaplan-Meier analysis (Fig 6). In the 11 patients who achieved a pCR, 8 continued with subsequent (chemo)radiotherapy, while the other 3 did not. Ten out of these 11 patients (90.9%) were alive and disease-free until the end of follow-up, whereas only one patient who did not undergo subsequent (chemo)radiotherapy developed local recurrence and eventually died.
In our analysis, 12 out of 87 patients (13.8%) were classified as T4b pretreatment. Among these 12 patients, 11 (91.7%) achieved a PR, while one had an SD following neoadjuvant chemo-immunotherapy. Notably, 8 of the 12 patients (66.7%) were successfully converted to resectable status. Of these 8 patients, 5 underwent surgical treatment, and none required free flap reconstruction for the resulting defect. Importantly, 3 of the 5 patients who had surgery demonstrated a pCR and remained alive and disease-free throughout the follow-up period. In addition to these 5 patients, the remaining 7 received (chemo)radiotherapy after neoadjuvant therapy. During the follow-up, 5 of the 12 patients (41.7%) developed local recurrence or distant metastasis, and 4 of these 5 patients died due to tumor progression. The OS rate was 66.7%, and the EFS rate was 58.3%, following a median follow-up period of 24.6 months.
Fig 7 illustrates notable tumor regression in two representative cases from this study. In Fig 7A-F, a patient (P85) with T4aN0M0 HNSCC involving the left larynx had his primary tumor completely disappear after neoadjuvant therapy, followed by (chemo)radiotherapy without surgery. In Fig 7G-H, another patient (P32) with T4bN0M0 HNSCC in the left posterior region of the maxilla experienced significant reduction of the primary tumor bulk, achieving a PR following neoadjuvant therapy. Surgery was subsequently performed on this patient, resulting in a pCR. Both patients remained alive and free of disease during the follow-up periods of 31.0 and 17.8 months, respectively.
Adverse events
The most common treatment-related adverse events (TRAEs) were alopecia (100.0%), nausea/vomiting (63.2%), fatigue (46.0%), and pain (40.2%) of all grades (Supplementary Table 2). 6 patients experienced grade 3 TRAEs, including 3 patients with leukopenia, 2 patients with pneumonitis, and 1 patient with peripheral neuropathy. One patient experienced grade 4 TRAEs (leukopenia). No previously unknown or unexpected TRAEs were observed. No TRAEs leading to death occurred.