The Importance of Extranodal Extension Grading in Laryngeal Squamous Cell Carcinoma

Abstract

Objective

The primary objective was to investigate the effect of extra‐nodal extension (ENE) grading on the survival of pN‐positive patients with laryngeal squamous cell carcinoma (LSCC).

Study Design

A retrospective cohort study.

Setting

A tertiary referral center.

Methods

The patients with LSCC were retrospectively reviewed. The histopathological slides of patients were re‐examined, and ENE was graded. Survival analyses were performed.

Results

Seventy‐six patients were enrolled in this study. The average age of patients was 61.29 years. 3‐year overall survival (OS), disease‐specific survival (DSS), and disease‐free survival (DFS) rates were 69.7%, 73.7%, and 73.7%, respectively. ENE grading had no statistically significant impact on survival rates.

Conclusion

While the presence of ENE in more than 4 lymph nodes, lymph node density (LND) greater than 0.2, poor histologic differentiation, and not receiving chemotherapy were identified as independent poor prognosticators in LSCC, the study did not show any effect of ENE grading on survival rates.

Tumor size, pathological neck node metastases, tumor depth, and cartilage or perineural invasion are important prognostic parameters in head and neck cancers. While neck node metastasis remains the most important prognosticator of survival, the presence of extra‐nodal extension (ENE) has also become as a significant poor prognosticator. ¹ , ² , ³ , ⁴ , ⁵ Strongly associated with survival and locoregional relapse, ENE is considered one of the 2 main parameters along with positive surgical margins that require incorporating chemotherapy as an adjuvant treatment. ⁶ , ⁷ , ⁸ As a result of many studies demonstrating the relationship between ENE and a poor prognosis, ENE was introduced into the N category in the eighth edition of the American Joint Committee on Cancer (AJCC) Staging Manual as a parameter modifying the staging system. ² , ⁹ Grading systems have been developed for a more detailed evaluation of ENE. ⁵ , ⁸ , ¹⁰ Studies have been conducted examining the effects of ENE grading and identification of lymph node density (LND) beyond simple ENE positivity on prognosis and survival. Along with studies stratifying ENE into minor ENE when the extension of the tumor is less than 2 mm beyond the lymph node capsule and major ENE when it is over 2 mm, some have found a cutoff of 1.7 mm to be the optimal threshold. ⁵ , ¹¹ Lewis et al have classified ENE in patients with oropharyngeal squamous cell carcinoma (SCC) into 5 grades (grades 0‐4). ¹⁰

Since there are not yet clear, standard macroscopic and histologic grading criteria for ENE, there is also no consensus on how to approach this in treatment management. Our objective was to evaluate the prognostic impact of ENE grading in patients with laryngeal squamous cell carcinoma (LSCC) and demonstrate its potential contribution to tumor staging by determining its relationship with other prognostic factors and tumor histology.

Materials and Methods

The study was planned as a retrospective cohort study following its approval by the local ethics committee (Ethics Committee of Istanbul Faculty of Medicine, protocol number: 2019/663). The medical records of LSCC patients with a positive lymph node (pN) stage treated with primary surgery and appropriate adjuvant therapy at our hospital between January 2010 and December 2017 were reviewed (n = 158). Primary hypopharyngeal cancer patients with laryngeal invasion (n = 20), duplicate records (n = 15), patients with incomplete follow‐up data (n = 33), and patients whose pathological slides were unfit for re‐examination (n = 14) were excluded from the study.

Demographic variables (sex, age at surgery) and cancer‐related clinical variables (type of tumor resection, side and type of neck dissection, type of adjuvant therapy, TNM stage according to the seventh and eighth editions of the AJCC manual, length of clinical follow‐up, follow‐up, and survival data) were sourced from medical records. Cancer‐related histopathological variables (LND [number of positive lymph nodes/number of excised lymph nodes], perineural invasion, cartilage invasion, large caliber vessel invasion, angio‐lymphatic invasion, histologic grade, tumor size, surgical margin, and number of lymph nodes with ENE) were sourced from postoperative histopathological reports. Information on 4 new cancer‐related histopathological variables (number of lymph nodes with ENE, ENE grade, the diameter of the largest lymph node with ENE, and the diameter of the largest metastatic focus) was gathered by re‐examining histopathological slides.

Re‐examination of Histopathological Slides

Hematoxylin and eosin‐stained sections of laryngectomy and neck dissection specimens were taken from the archive at the pathology department and re‐evaluated. Every single metastatic lymph node was examined under a light microscope for the presence of ENE. Millimetric measurements were conducted in ENE‐positive patients using the eyepiece ocular of the microscope. ENE was graded according to the grading system delineated in the literature. ¹⁰ The diameter of the largest lymph node with ENE and the diameter of the largest metastatic focus were also added to the measurements.

Statistical Analysis

The statistical analysis was done with version 26 of the Statistical Package for the Social Sciences software (IBM Corp.). A descriptive analysis was performed for demographic data. Categorical variables were expressed as numbers and percentages. All continuous variables were checked for normality using the Shapiro–Wilk test and Q–Q plots. Normally distributed variables were presented as average (standard deviation) (range), while the others were presented as median (first to‐third quartile).

Fisher's exact probability test assessed the associations between categorical variables. Spearman's rank‐order correlation was performed to determine whether there were correlations between 2 continuous variables or 1 categorical and 1 continuous variable. In comparisons between groups, one‐way ANOVA was used if variables were normally distributed, and the Kruskal–Wallis H test if they were non‐normally distributed.

Three different survival definitions were set for survival analysis. These definitions were as follows:

• 1‐Overall survival (OS): The interval between the date of the primary treatment and the date of death‐ or the date of the latest follow‐up in patients still alive.
• 2‐Disease‐specific survival (DSS): The interval between the date of the primary treatment and the date of death from primary cancer‐ or the date of the latest follow‐up in patients still alive.
• 3‐Disease‐free survival (DFS): The interval between the end of adjuvant treatment to the date of any recurrence‐ or the date of the latest follow‐up in patients having no recurrence.

Survival data were reported in the form of 3‐year survival rates: 3‐year OS, 3‐year DSS, and 3‐year DFS.

X‐tile cut point finder ¹² software was used to determine optimal cut points for continuous variables (age, LND, number of lymph nodes with ENE, diameter of the largest lymph node with ENE, diameter of the largest metastatic focus, and tumor size). ¹² The Kaplan–Meier estimator was used for the univariate survival analysis of categorical variables and continuous variables converted into categorical variables through the cut function. The log‐rank test was employed to compare survival curves. Variables with a P value below .10 in the log‐rank test were subjected to multivariate survival analysis. The Cox proportional hazards model was used to determine independent prognostic factors impacting survival. Time‐dependent covariates were generated and included in the model to test for proportionality. None of the time‐dependent covariates showed statistical significance (P > .05). The assumption of proportionality was therefore satisfied by each variable.

Results

Demographic Findings and Clinical Data

Seventy‐six patients were enrolled in the study. The average age of patients was 61.29 (SD: 10.33) (range: 37‐82) years. Most patients were male (73%‐96.1%). The median (first to third quartile) cohort follow‐up was 50.62 (28.74‐90.3) months. The type of surgery, side and type of neck dissection, treatment failure data, and other clinical data, such as the number of patients receiving radiotherapy and chemotherapy, are provided in Table 1. TNM staging features are listed in Table 2.

Table 1.

Demographic and Clinical Data

Agea	61.29 + 10.33 (37‐82)
Type of surgery
TL	46
SGL	15
SCL	9
TOLS	2
VHL	2
Near‐TL	1
TLF	1
Neck dissection (ND)
Bilateral	66
Unilateral	10
Type of ND
Selective	98
Modified radical/radical	42
Not specified	2
Follow‐up timeb	50.62 (28.74‐90.3)
Exitus	34
Because of primary cancer	24
Other disease	10
Recurrence	25
Site of recurrence
Locoregional	7
Distant metastasis	11
Locoregional+distant metastatis	6
Not specified	1
Site of distant metastasis
Pulmonary and mediastinum	16
Vertebra and other bone	3
Brain	1
Liver	3
Not specified	1
Radiotherapy (no/yes)	4/72
Chemotherapy (no/yes)	30/44

Other data were presented as number.

Abbreviations: LND, lymph node density; SCL, supracricoid laryngectomy; SGL, supraglottic laryngectomy; TL, total laryngectomy; TLF, total laryngofaryngectomy; TOLS, transoral laser surgery; VHL, vertical hemilaryngectomy.

^a Data were presented as mean (standard deviation) (range).

^b Data were presented as median (first to third quartiles).

Table 2.

TNM Stage and ENE Grade

T stage
T2	16
T3	23
T4a	37
pN stage (AJCC 7th)
N1	22
N2a	1
N2b	25
N2c	27
N3	1
pN stage (AJCC 8th)a
N1	13
N2a	9
N2b	5
N2c	6
N3b	43
Stage (AJCC 7th)
3	4
4a	71
4b	1
Stage (AJCC 8th)b
4a	33
4b	43
ENE Grade
0	19
1	16
2	17
3	15
4	9

Abbreviations: AJCC, American Joint Committee of Cancer; ENE, extranodal extension; pN, pathologic nodal stage; T, tumor stage.

^a The number of patients whose pN stage was upstaged was 51; of these, 43 were upstaged to pN3b and 8 to pN2a.

^b The number of patients whose TNM stage was upstaged was 45, of these, 42 were upstaged to stage 4b and 3 to stage 4a.

Review of Postoperative Histopathology Reports

Average tumor size was 3.61 (SD: 1.05) (range: 1.4‐6.0) cm. The median number of metastatic lymph nodes was 2 (1‐ 4.75). Thirty (39.5%) patients presented with perineural invasion, 30 (39.5%) patients with cartilage invasion (thyroid or cricoid), 45 (59.2%) patients with angio‐lymphatic invasion, 16 (21.1%) patients with large caliber vessel invasion, and 11 (14.5%) patients with a positive surgical margin (<1 mm). While 29 (38.2%) were of poorly differentiated histologic grade, 43 patients (56.6%) were of moderate to well‐differentiated grade. The number of patients with ENE‐positive lymph nodes was 49 (64.5%), and the median ENE‐positive lymph node number was 1 (0‐3). Thirty‐one patients had more than 1 ENE‐positive lymph node. The median LND was 0.066 (0.03‐0.13).

Re‐examination of Histopathological Slides

Nine patients showed an increased number of ENE‐positive lymph nodes upon re‐examining slides. In 8 of these patients, 1 extra ENE‐positive lymph node was found. While 7 of these 8 had been classified as ENE‐negative in the first histopathological assessment, 1 had 5 ENE‐positive lymph nodes. One patient, classified as ENE‐negative in the first assessment, had 2 ENE‐positive lymph nodes. Consequently, the number of patients with ENE‐positive lymph nodes rose to 57 (68.4%). More than 1 ENE‐positive lymph node was present in 32 patients. The median number of ENE‐positive lymph nodes was 1 (0.25‐3). The median largest ENE‐positive lymph node diameter was 15 (1.75‐27.5) mm, and the median largest metastatic focus diameter was 15 (6‐21) mm. Data on ENE grading is provided in Table 2.

The re‐examination of slides resulted in an alteration of the pN‐stage of 8 patients. Three patients were upstaged from N1 to N2a and 5 patients from N2 (1 N2a, 3 N2b, and 1 N2c) to N3b. Finally, 4 patients advanced from stage 4a to stage 4b.

Correlations Between LND/Number of Lymph Node With ENE and Other Parameters

A positive correlation was found between LND and the number of lymph nodes with ENE (r: 0.672, P < .001), the diameter of the largest lymph node with ENE (r: 0.36, P: .001), the diameter of the largest metastatic focus (r: 0.48, P < .001), and angio‐lymphatic invasion (r: 0.37, P < .001); while a negative correlation was found between LND and cartilage invasion (r: −0.25, P < .05). The relationship between the number of lymph nodes with ENE and other parameters followed the same trend as LND (Table 3).

Table 3.

Correlations Between LND/NLNwE and Other Variables

		LND	NLNwE
Age	r	−0.054	−0.121
	P	.646	.298
LND	r		0.672*
	P		.000
NLNwE	r	0.672*
	P	.000
MDoLLNwE (mm)	r	0.360*	0.743*
	P	.001	.000
MDoLMF (mm)	r	0.480*	0.729*
	P	.000	.000
Tumor Size	r	0.125	0.168
	P	.300	.162
PNI	r	0.025	0.074
	P	.832	.529
Cartilage invasion	r	−0.247**	−0.278**
	P	.034	.016
Angio‐lymphatic invasion	r	0.369*	0.340*
	P	.001	.003
Large caliber vessel invasion	r	0.086	0.184
	P	.467	.120
Histologic grade	r	−0.052	−0.043
	P	.666	.722

Abbreviations: LND, lymph node density; MDoLLNwE, maximum diameter of the largest lymph node with extranodal extension; MDoLMF, maximum diameter of the largest metastatic focus; NLNwE, number of lymph nodes with extranodal extension; PNI, perineural invasion.

^*Correlation is significant at the 0.01 level.

^**Correlation is significant at the 0.05 level.

Comparison of Parameters According to ENE Grades

No significant difference was observed between patients grouped according to ENE grades regarding perineural invasion, cartilage invasion, angio‐lymphatic, and large caliber vessel invasion rates (P > .05). No significant difference in tumor size was observed between ENE grades either, χ ²(4) = 0.238, P = .993.

Significant differences were identified between ENE grades in terms of the variables of age (F (4.71) = 3.734, P = .008), the number of lymph nodes with ENE (χ ²(3) = 12.647, P = .005), the diameter of the largest lymph node with ENE (χ ²(3) = 13.353, P = .004), and the diameter of the largest metastatic focus (χ ²(4) = 41.667, P < .001). Paired comparisons showed that ENE grade 4 patients were significantly older than ENE grade 1 (adj. P = .02) and ENE grade 3 (adj. P = .01) patients (Figure 1A). The number of lymph nodes with ENE was lower in ENE grade 1 patients in comparison to ENE grade 3 (adj. P = .01) and ENE grade 4 (adj. P = .046) patients (Figure 1B). ENE grade 3 patients exhibited a higher LND than ENE‐negative patients (adj. P = .019) (Figure 1C). The diameter of the largest lymph node with ENE was smaller in ENE grade 2 patients compared to ENE grade 3 (adj. P = .035) and ENE grade 4 (adj. P = .007) patients (Figure 1D). The diameter of the largest metastatic focus was smaller in ENE‐negative (ENE degree 0) patients compared to ENE‐positive (grades 1‐2‐3‐4 [adj. P < .001] patients) (Figure 1E).

Figure 1.

Comparison of parameters according to ENE grades. (A) Comparison of ages between ENE grades. (B) Comparison of the number of lymph nodes with ENE between ENE grades. (C) Comparison of lymph node density between ENE grades. (D) Comparison of maximum diameter of lymph node with ENE between ENE grades. (E) Comparison of maximum diameter of the largest metastatic focus between ENE grades. ENE, extranodal extension.

Survival Analysis

The 3‐year OS, DSS, and DFS rates were 69.7%, 73.7%, and 73.7%, respectively. Table 4 presents a univariate analysis of survival. Variables with P < .05 in the log‐rank test were considered significantly associated with survival. Figure 2 provides Kaplan‐Meier survival curves for the variables of ENE grade, the number of lymph nodes with ENE, and LND. Multivariate analysis results demonstrate that poor histologic differentiation and not receiving chemotherapy are independent poor prognosticators for 3‐year OS. Poor histologic differentiation and the presence of ENE in more than 4 lymph nodes were determined to be independent poor prognosticators for 3‐year DSS. Not receiving chemotherapy was the only poor prognosticator for 3‐year DFS. Other data from the multivariate analysis (P values, hazard ratios, etc.) are provided in Table 5.

Table 4.

Univariate Analysis of Variables for DSS, OS, and DFS

		DSS			OS			DFS
	N	Deaths (in 3 years)a	3‐year DSS	Log rank test	Deaths (in 3 years)a	3‐year OS	Log rank test	Deaths (in 3 years)a	3‐year DFS	Log rank test
Total patients	76		73.7%			69.7%			73.7%
Patient characteristics
Age				0.051			0.039*			0.066
≤63 years	39	7	82.1%		8	79.5%		7	82.1%
>63 years	37	13	64.9%		15	59.5%		13	64.9%
Cancer characteristics
LND				0.041*			0.017*			0.032*
<0.10	51	11	78.4%		12	76.5%		11	78.4%
0.10‐0.20	16	4	75.0%		5	68.8%		4	75.0%
>0.20	9	5	44.4%		6	33.3%		5	44.4%
NLNwE				0.021*			0.002*			0.022*
0‐4	63	14	77.8%		15	76.2%		14	77.8%
>5	13	6	53.8%		8	38.5%		6	53.8%
ENE grade				0.308			0.124			0.646
0	19	7	63.2%		7	63.2%		6	68.4%
1	16	3	81.3%		3	81.3%		3	81.3%
2	17	2	88.2%		2	88.2%		3	82.4%
3	15	5	66.7%		7	53.3%		5	66.7%
4	9	3	66.7%		4	55.6%		3	66.7%
MDoLLNwE (mm)				0.798			0.694			0.626
0‐12	29	7	75.9%		7	75.9%		6	79.3%
13‐24	21	5	76.2%		7	66.7%		6	71.4%
>24	26	8	69.2%		9	65.4%		8	69.2%
MDoLMF (mm)				0.686			0.677			0.621
0.2‐14	38	10	73.7%		10	73.7%		9	76.3%
15‐22	21	4	81.0%		7	66.7%		5	76.2%
>23	17	6	64.7%		6	64.7%		6	64.7%
Tumor size (cm)				0.045*			0.052			0.16
≤3.1	28	4	85.7%		5	82.1%		5	82.1%
>3.1	43	15	65.1%		17	60.5%		14	67.4%
PNI				0.312			0.4			0.29
No	44	10	77.3%		12	72.7%		10	77.3%
Yes	30	10	66.7%		11	63.3%		10	66.7%
Cartilage invasion				0.861			0.931			0.889
No	44	11	75.0%		13	70.5%		11	33
Yes	30	8	73.3%		9	70.0%		8	22
Angio‐lymphatic invasion				0.15			0.154			0.155
No	29	5	82.8%		6	79.3%		5	82.8%
Yes	45	15	66.7%		17	62.2%		15	66.7%
Large caliber vessel invasion				0.268			0.22			0.263
No	57	14	75.4%		16	71.9%		14	75.4%
Yes	16	6	62.5%		7	56.3%		6	62.5%
Histologic grade				0.016*			0.009*			0.056
Good‐moderate	43	6	86.0%		7	83.7%		7	83.7%
Poor	29	11	62.1%		13	55.2%		10	65.5%
Surgical margin				0.811			0.557			0.372
Negative	65	17	73.8%		19	70.8%		16	75.4%
Positive	11	3	72.7%		4	63.6%		4	63.6%
pT stage				0.676			0.421			0.727
T2	16	3	81.3%		3	81.3%		3	81.3%
T3	23	7	69.6%		9	60.9%		7	69.6%
T4a	37	10	73.0%		11	70.3%		10	73.0%
pN stage (AJCC 7th)				0.949			x			0.89
N1	22	6	72.7%		7	68.2%		6	72.7%
N2b	25	6	76.0%		6	76.0%		7	72.0%
N2c	27	7	74.1%		9	66.7%		6	77.8%
pN stage (AJCC 8th)				0.062			0.127			0.236
N1	13	3	76.9%		3	76.9%		3	76.9%
N2a	9	3	66.7%		4	55.6%		3	66.7%
N2b	5	2	60.0%		2	60.0%		3	40.0%
N2c	6	4	33.3%		4	33.3%		3	50.0%
N3b	43	8	81.4%		10	76.7%		8	81.4%
Stage (AJCC 7th)				0.993			0.881			0.986
3	4	1	75.0%		1	75.0%		1	75.0%
4a	71	18	74.6%		21	70.4%		18	74.6%
Stage (AJCC 8th)				0.267			0.362			0.29
4a	33	11	66.7%		12	63.6%		11	66.7%
4b	43	9	79.1%		11	74.4%		9	79.1%
Treatment characteristics
Chemotherapy				0.002*			<0.001*			0.008*
No	30	13	56.7%		16	46.7%		12	60.0%
Yes	44	7	84.1%		7	84.1%		7	84.1%

Abbreviations: AJCC, American Joint Committee of Cancer; CI, confidence interval; ENE, extranodal extension; LND, lymph node density; MDoLLNwE, maximum diameter of the largest lymph node with extranodal extension; MDoLMF, maximum diameter of the largest metastatic focus; NLNwE, number of lymph nodes with extranodal extension; NR, not reached; OS, overall survival; pN, pathologic nodal stage; PNI, perineural invasion; T, Tumor stage.

^a Data were presented as number (mortality rate).

^*Variables were significantly associated with survival (P < .05).

Figure 2.

Kaplan–Meier survival curves. (A) Kaplan–Meier survival curves for ENE grade. (B) Kaplan–Meier survival curves for number of lymph nodes with ENE. (C) Kaplan–Meier survival curves for lymph node density.

Table 5.

Multivariate Analysis of Variables for 3‐Year OS, DSS, and DFS

	Hazard ratio	95% CI	P value
3‐year OS
Age
≤63 years	Reference
>63 years	2.454	0.79‐7.61	.120
LND
<0.10	Reference
0.10‐0.20	0.941	0.21‐4.21	.936
>0.20	4.267	0.77‐23.61	.096
NLNwE
0‐4	Reference
>4	4.308	0.81‐23.05	.088
ENE Grade
0	Reference
1	0.251	0.04‐1.44	.121
2	0.453	0.08‐2.66	.380
3	1.266	0.26‐6.27	.772
4	0.601	0.14‐2.57	.492
Tumor Size (cm)
≤3.1	Reference
>3.1	1.978	0.53‐7.42	.312
Histologic grade
Good‐Moderate	Reference
Poor	4.892	1.64‐14.59	.004*
Chemotherapy
No	Reference
Yes	0.112	0.03‐0.41	.001*
3‐year DSS
Age
≤63 years	Reference
>63 years	2.431	0.69‐8.51	.165
LND
<0.10	Reference
0.10‐0.20	0.646	0.01‐4.24	.649
>0.20	3.181	0.37‐27.35	.292
NLNwE
0‐4	Reference
>4	9.435	1.11‐80.44	.040*
Tumor Size (cm)
≤3.1	Reference
>3.1	3.766	0.92‐15.37	.065
Histologic grade
Good‐Moderate	Reference
Poor	3.320	1.08‐10.2	.036*
pN (AJCC 8th)
N1	Reference
N2a	0.885	0.16‐4.94	.890
N2b	2.016	0.18‐22.73	.570
N2c	2.200	0.39‐12.33	.370
N3b	0.346	0.03‐3.42	.364
Chemotherapy
No	Reference
Yes	0.294	0.07‐1.29	.104
3‐year DFS
Age
≤63 years	Reference
>63 years	2.497	0.86‐7.29	.094
LND
<0.10	Reference
0.10‐0.20	0.988	0.23‐4.29	.987
>0.20	2.519	0.54‐11.65	.237
NLNwE
0‐4	Reference
>4	4.040	0.93‐17.47	.062
Histologic grade
Good‐Moderate	Reference
Poor	2.682	0.91‐7.91	.074
Chemotherapy
No	Reference
Yes	0.310	0.12‐0.83	.019*

Abbreviations: AJCC, American Joint Committee of Cancer; CI, confidence interval; ENE, extranodal extension; LND, lymph node density; NLNwE, number of lymph nodes with extranodal extension; pN, pathologic nodal stage.

^*Variables that were determined as an independent predictor (P < .05).

Discussion

The primary outcome variable of our study was to investigate the effect of ENE grading on survival. In the eighth edition of the AJCC Staging Manual, ENE positivity has been subdivided into major and minor ENE. ¹³ Although this subdivision has no bearing on TNM staging, it is recommended for data collection and analysis. Instead of the major and minor subdivisions, we used the grading system proposed by Lewis et al in grading ENE. ¹⁰ Our univariate survival analyses showed that ENE grading had no statistically significant impact on 3‐year OS, DSS, and DFS in patients with LSCC. Although we observed a trend suggesting better cumulative survival for ENE grades 1 (81.3% 3‐year OS) and 2 (88.2%) compared to grades 0 (63.2%), 3 (53.3%), and 4 (55.6%) (Figure 2A), this difference did not reach statistical significance. The relatively small sample size may have limited our ability to detect such differences. Furthermore, subdividing ENE into 5 grades could further reduce statistical power due to the smaller sample sizes in each group. The lower cumulative survival rates observed in ENE grade 0 patients compared to grades 1 to 2 may be partially explained by the fact that ENE grade 0 patients were, on average, 4 to 5 years older (Figure 1A) and were less likely to receive chemotherapy, which could have independently influenced survival outcomes. In a similar study on the impact of further stratifying ENE into major and minor on survival rates in patients with head and neck cancers, authors reported no significant difference in 3‐year survival rates. ⁵ Greenberg et al also found no significant difference in survival between major and minor ENE in patients with oral SCC of the tongue while associating the presence of ENE in more than 1 lymph node with a poor prognosis. ¹⁴ On the other hand, some studies demonstrated the prognostic value of ENE grading. In their original study delineating the ENE grading system we used, Lewis et al argued that only ENE grade 4 might be associated with a poorer outcome in patients with oropharyngeal SCC. However, since ENE grade 4 correlated with higher T‐stage, it was not considered an independent variable. Also, as p16 positivity was approximately 90% in this study, it would be inaccurate to compare it with other studies differentiating between the grade of ENE presence. ¹⁰ Agarwal et al have similarly investigated the effect of ENE grading on prognosis in patients with SCC of the oral cavity using the same classification system. ¹⁵ This study can demonstrate the impact of ENE grading more independently as it has been conducted on non‐HPV‐related cancer patients. ENE grades 3 and 4 have been associated with poorer survival. In their study using the same ENE grading system to determine its impact on OS in patients with non‐oropharyngeal head and neck cancers, Prabhu et al have also found ENE grade 4 to be associated with poorer survival despite chemoradiotherapy, and have reported no significant difference between ENE grade 0 and ENE grades 1 to 3 in survival. As a result, they have indicated ENE grade 4 patients as candidates for research on adjuvant therapy intensification. ¹⁶ Our study differs from others in that it investigates the prognostic value of ENE grading in only pN‐positive patients with LSCC and is—to our knowledge—the first study in the literature to homogeneously analyze ENE grading in patients with LSCC.

A set of important secondary findings from our study was that LND greater than 0.2, the presence of ENE in more than 4 lymph nodes, and not receiving chemotherapy were identified as poor prognostic factors across all 3 survival categories in univariate analysis. Therefore, these variables may serve as valuable predictors for both disease recurrence (DFS) and survival outcomes such as DSS and OS. Additionally, tumor size greater than 3.1 cm and poor histologic differentiation were statistically significant predictors of 3‐year DSS. On multivariate analysis, ENE in more than 4 lymph nodes and poor histologic differentiation were identified as independent risk factors for 3‐year DSS, while only not receiving chemotherapy was found to be an independent risk factor for 3‐year DFS. Similarly, Agarwal et al have reported LND greater than 0.12 and ENE in more than 2 lymph nodes to be associated with worse survival in patients with SCC of the oral cavity, emphasizing the value of including LND, the number of nodes with ENE, and grade of ENE along with TNM stage as significant prognosticators in patients with ENE. ¹⁵ In their 4710 case series of patients with lymph node metastasis, investigating the prognostic impact of quantitative lymph node burden in cancers of the larynx and hypopharynx, Ho et al found that mortality risk escalated continuously without plateau as the number of metastatic lymph nodes increased. Their study highlights the number of metastatic lymph nodes as a predominant independent factor associated with mortality in hypopharyngeal and laryngeal cancers. ¹⁷ Petrarolha et al have investigated the prognostic value of LND in patients with LSCC. ¹⁸ They have reported that patients with LND ≥ 0.06 had a higher mortality rate and earlier disease recurrence than patients with LND < 0.06. ¹⁸ In their study evaluating the prognostic impact of LND according to the existing classification system, Reinisch et al found that patients with positive lymph nodes but a lymph node ratio of 0‐6% had no survival difference when compared to patients without lymph node metastasis. Therefore, they suggested using LND as a predictor of survival in patients with metastasis on the ipsilateral side of the neck. ¹⁹ Studies also demonstrate the value of LND as a marker in predicting the benefit of postoperative radiotherapy. ²⁰ While size and laterality of metastatic lymph nodes are used in the contemporary staging of head and neck cancers, the number of lymph nodes with ENE and LND do not feature as determinants of the stage. The fact that LND is identified as a predictor of both prognosis and recurrence and that it was positively correlated with such an essential prognosticator (the number of lymph nodes with ENE) in our study supports the existing literature regarding the prognostic impact of LND.

The limitations of our study include its retrospective design and relatively small cohort size. Furthermore, subdividing the cohort by ENE grading into multiple categories further diluted the sample size in each group, limiting the statistical power and applicability of the results. Another significant challenge is the potential variability in applying the ENE grading system. While a single experienced pathologist reviewed the histopathological slides in our study, eliminating inter‐rater variability, the lack of a fully validated grading system introduces the possibility of inconsistencies or errors in grading. Variability in tissue quality, slide preparation, or the pathologist's interpretation across different slides could affect the accuracy of the grading. Although we did not assess these factors, future studies should involve multiple pathologists and consider statistical measures of inter‐rater variability to improve the reliability and repeatability of the grading system.

The strengths were the inclusion of only pN‐positive patients with LSCC to form a homogeneous group, being the first study investigating the impact of ENE grading on prognosis in patients with LSCC and collecting specific histopathological data through a prospective re‐evaluation. The literature has identified LND, ENE grading, and the number of lymph nodes with ENE as potentially important prognostic factors in patients with head and neck cancers. Although our study did not demonstrate a statistically significant association with ENE grading, our findings suggest that it may still play an important role in prognosis. Additionally, LND and the number of lymph nodes with ENE were identified as relevant prognostic factors in our study, consistent with previous research. Further studies with larger sample sizes and randomized controlled prospective trials are needed to obtain more precise results.

Conclusions

In univariate analysis, the presence of ENE in more than 4 lymph nodes, LND greater than 0.2, poor histologic differentiation, and not receiving chemotherapy were identified as poor prognostic factors for 3‐year DFS, 3‐year DSS, and 3‐year OS in patients with LSCC. Furthermore, not receiving chemotherapy was found to be an independent risk factor for both 3‐year DFS and 3‐year OS, while ENE in more than 4 lymph nodes was an independent risk factor for 3‐year DSS, and poor histologic differentiation was an independent risk factor for both 3‐year DSS and 3‐year OS. Although ENE grading did not show statistical significance in our study, patients with higher ENE grades tended to show worse cumulative survival. These histopathological parameters, which are not currently incorporated into the TNM staging system, may provide valuable additional insights to help guide adjuvant therapy decisions in the future.

Author Contributions

Hakan Kara, conceptualization (equal), data curation (equal), formal analysis (equal), methodology (equal), validation (equal), visualization (equal), writing—original draft (equal), writing—review and editing (equal); Levent Aydemir, conceptualization (equal), methodology (equal), writing—original draft (equal), writing—review and editing (equal); Melek Büyük, conceptualization (equal), investigation (equal), methodology (equal), writing—original draft (supporting); Erol Bozbora, data curation (equal), methodology (equal), software (equal), writing—original draft (supporting); Kübra Özkaya Toraman, conceptualization (equal), data curation (equal), methodology (equal), writing—original draft (supporting); Saim Pamuk, data curation (equal), methodology (equal), visualization (equal), writing—original draft (supporting); Kağan Avcı, data curation (equal), methodology (equal), validation (equal), writing—original draft (supporting); Comert Sen, data curation (equal), investigation (equal), methodology (equal), writing—original draft (supporting), writing—review and editing (supporting); Said Sonmez, conceptualization (equal), data curation (equal), methodology (equal), writing—original draft (supporting), writing—review and editing (supporting); Murat Ulusan, conceptualization (equal), data curation (equal), formal analysis (equal), methodology (equal), writing—original draft (supporting), writing—review and editing (supporting); Bora Basaran, conceptualization (equal), data curation (equal), methodology (equal), writing—original draft (supporting), writing—review and editing (supporting); Musa Altun, conceptualization (equal), data curation (equal), formal analysis (equal), methodology (equal), visualization (equal), writing—original draft (supporting), writing—review and editing (supporting); Erkan Kıyak, conceptualization (equal), data curation (equal), formal analysis (equal), methodology (equal), writing—original draft (supporting), writing—review and editing (supporting).

Disclosures

Competing interests

None.

Funding sources

None.