Impact of WPOI‐5 on Risk of Regional Recurrence in Early Oral Cancer With pN0 Neck

ABSTRACT

Introduction

Regional recurrence (RR) has been reported to occur in 7%–17% of patients with oral squamous cell carcinoma (OSCC) and pathologically negative (pN0) necks. Risk factors for isolated RR, in the absence of simultaneous local recurrence, are poorly defined. Our aim was to study risk factors for isolated RR in early OSCC with clinically and pathologically negative necks.

Methods

Retrospective cohort study of 202 patients with cT1/2N0 OSCC.

Results

Among 130 patients undergoing elective neck dissection, 98 were pN0. Five (5.1%) developed isolated RR, and 7 (7.1%) developed any RR. Risk factors for isolated RR were worst pattern of invasion category 5 (WPOI‐5) (p = 0.02), and absence of postoperative radiotherapy (PORT) (p = 0.04). Among 170 patients with pN0 or cN0/x necks, risk factors for isolated RR were WPOI‐5 and depth of invasion > 10 mm. On multivariate analysis, only WPOI‐5 remained significant.

Conclusion

Isolated RR may occur in over 5% of patients with early OSCC and pathologically negative necks after elective ND. WPOI‐5 would appear to be the most important risk factor for isolated RR in pN0 cases. Further work is required to define the role of PORT in patients with WPOI‐5.

Level of Evidence

3.

In the present retrospective study, worst pattern of invasion category 5 (WPOI‐5) was found to be significantly associated with regional recurrence (RR) among patients with early oral cancer and pathologically negative necks after elective neck dissection. Isolated RR was seen only among patients not receiving postoperative radiotherapy (PORT), however, more work is required to define the role of PORT in patients with WPOI and pN0 necks.

1. Introduction

The presence of lymph node metastases is well established as one of the most important adverse prognostic factors in oral cavity squamous cell carcinoma (OSCC) [1, 2]. Among patients presenting with clinically negative necks (cN0), elective neck dissection (ND) is generally recommended, as this provides not only prognostic information to help stratify patients for adjuvant treatment [3], but may also provide a survival benefit for patients with occult metastases [4, 5]. Besides nodal metastases, other factors that have been described as adverse prognosticators in OSCC include depth of invasion (DOI) [6, 7], perineural invasion (PNI) [8, 9], lymphovascular invasion (LVI) [10, 11], positive surgical margins [12, 13, 14], and histological pattern of invasion (POI) at the invasive front [8, 11, 15, 16, 17, 18]. Regarding the latter, Brandwein‐Gensler described five categories of POI, with worst pattern of invasion‐5 (WPOI‐5) being of most relevance [8]. WPOI‐5 is defined according to the College of American Pathologists as ≥ 1 mm dispersion between tumor satellites, and/or extratumoral LVI or PNI [19]. This differs somewhat from the POI definition utilized by the Royal College of Pathologists (RCPath), which is now a tertiary system in their updated 2023 guidelines, and classifies the invasive front as cohesive, non‐cohesive, or the newly introduced category of widely dispersed. Using this approach, cohesive is equivalent to WPOI patterns 1–3, non‐cohesive to WPOI‐4, and widely dispersed to WPOI‐5 [20].

Among patients with cN0 necks undergoing elective ND without pathological evidence of nodal metastases (pN0), prognosis is considered favorable, and the risk of regional recurrence (RR) is generally considered to be low. However, RR in patients who are pN0 after elective ND (END) is described. In a recent systematic review and meta‐analysis, Chegini et al. reported RR to occur in the pN0 neck in 17.3% of cases staged T1‐4, and in 7.5% of cases staged T1/2 [21]. However, there are little published data regarding risk factors for RR among patients who are pN0 following elective clearance of the nodal basin.

The primary aim of the present study was thus to study the incidence of and risk factors for RR in pN0 neck among patients with early OSCC. A secondary aim was to investigate risk factors for the finding of occult metastases in the cN0 neck.

2. Methods

This was a retrospective cohort study of a prospectively maintained database of oral cavity squamous cell carcinomas in a regional referral center for head and neck cancer between 2000 and 2021. The study was performed according to STROBE reporting guidelines. Ethical approval was granted by the Cork Clinical Research Ethics Committee. Inclusion criteria were new cases of T1 and T2 OSCC (7th edition AJCC staging), with cN0 at presentation, undergoing primary surgical treatment. Exclusion criteria included a previous history of head and neck cancer or radiotherapy (RT) to the neck.

Patient files were reviewed and data collected including demographics and clinical characteristics, type of surgery performed, and administration of adjuvant treatment. Clinical nodal status of the neck (cN+ versus cN0) was determined by review of notes and original radiology reports. Pathological data recorded included tumor size, DOI, POI, PNI, LVI, and margin status. Although inclusion criteria for the present study were based on the 7th edition AJCC staging, all patients were re‐staged according to the AJCC 8th edition. Of note, DOI measurements were previously re‐evaluated for all patients in our study cohort who underwent surgery prior to the issuance by the AJCC in 2017 of the clarified definition for this parameter. Slide review was also performed among cases reported as having a non‐cohesive POI, PNI, or LVI, to ascertain if satellite nodules and/or extratumoral PNI or LVI, fulfilling the definition for WPOI‐5, were present. Of note, cases reported as having non‐cohesive growth on initial reports incorporated both WPOI‐4 and WPOI‐5 cases, as, at that time, the RCPath utilized a two‐tier system of cohesive versus non‐cohesive only for assessing POI. Slide review was undertaken by two pathologists at a multiheaded microscope and consensus results recorded. Extratumoral PNI and extratumoral LVI were defined as PNI or LVI present ≥ 1 mm beyond the invasive tumor front, respectively. WPOI‐5 was defined as dispersed satellite nodule(s) ≥ 1 mm from the main tumor or presence of extratumoral PNI or LVI. Positive margins were defined as invasive SCC or carcinoma in situ at the margin.

At our institution, ND is generally performed as default among patients with early OSCC, excepting cases of very thin (< 3 mm) primary tumors or with significant comorbidity. For tumors approaching midline, bilateral ND is performed. Decisions regarding postoperative RT (PORT) are made at the Head & Neck multidisciplinary meeting. General indications for PORT include positive margins or the presence of nodal metastases. Relative indications include the presence of very close margins, PNI, or large primary tumors (pT3/4).

3. Outcomes Measures

The primary outcome measure was the incidence of isolated RR among patients with pN0 necks. Secondary outcome measures included the incidence of and risk factors for occult metastases detected by pathological examination of the cN0 neck among patients undergoing elective ND. Due to the small number of events of isolated RR among patients with pN0 necks, we also performed an analysis of the incidence of any RR in this cohort, as well as an analysis of the risk of isolated or any RR among patients with pN0/x necks. Isolated RR was defined as RR occurring in the absence of preceding or concomitant local recurrence (LR) or secondary primary tumor (SPT). RR‐free survival was defined as the time period between the date of surgery and the date of diagnosis with RR or last follow‐up in clinic. Patients developing head and neck SPTs were censored at the time of presentation with the SPT.

4. Data Analysis

Statistical analysis was performed using IBM SPSS version 26. Comparisons of categorical data were made using crosstabs with either chi‐squared test or Fisher's exact test if cell numbers were below 10. Post hoc analysis between groups was performed using Bonferroni correction, where applicable. 5‐year overall survival (OS) and disease‐specific survival (DSS) were computed using the Kaplan–Meier method and compared using log‐rank test. Univariate and multivariate Cox regression analyses were performed to determine factors associated with DSS and OS. Variables deemed clinically relevant or statistically significant on univariate Cox regression analysis were included in multivariate Cox regression analysis. A backwards stepwise entry method with p > 0.1 as an exclusion criterion was used for multivariate analysis.

5. Results

202 (132 male) patients with cT1/2N0 oral cavity SCC were identified. Mean age was 62.9 ± 13.3 years. Primary tumor subsites were tongue (101, 50%), floor of mouth (52, 25.7%), buccal mucosa (11, 5.4%), lower alveolus (16, 7.9%), retromolar trigone (12, 5.9%), lip (6, 3%), upper alveolus (3, 1.5%), and hard palate (1, 0.5%). 130 patients underwent unilateral (105) or bilateral (25) ND. The mean lymph node yield per hemineck was 24.9 ± 11.8.

5.1. Recurrences in pN0 Patients

Ninety‐eight patients were pN0 after ND. Demographic and clinicopathological data of this cohort are shown in Table 1. Mean (median) follow‐up was 72 (60) months. 7 (7.1%) patients developed RR, of whom two had RR simultaneous with local recurrence (LR), giving 5 (5.1%) patients with isolated RR. Three patients with isolated RR had recurrence in the contralateral undissected neck. Of the two patients with simultaneous LR and RR, one had RR in the contralateral undissected neck, and the other had bilateral neck recurrence. Of note, none of the patients with isolated RR had received PORT.

TABLE 1.

Clinicopathological characteristics of patients who were pN0 or pNx.

		pN0 (n = 98)	pNx (n = 72)
Age (years)		59.1	68.2	< 0.0001
Sex	Male	57	52	0.08
	Female	41	20
Site	Tongue	53	31	0.17 (tongue v non‐tongue)
	Floor of mouth	26	17
	Lower alveolus	6	9
	Retromolar	5	6
	Upper alveolus/hard palate	2	1
	Buccal	6	4
	Lip	0	4
Neck dissection	None	0	72
	Unilateral	80	0
	Bilateral	18	0
DOI (mm)		7.42	4.76	0.0007
T‐classification (AJCC 8th edition)	pT1	26	46	< 0.0001 (pT1 v pT2/3)
	pT2	59	21
	pT3	13	5
Invasive front	Cohesive	55	62	< 0.0001
	Non‐cohesive	43	10
PNI		25	6	0.005
LVI		8	2	0.19
Positive margin		13	7	0.63
WPOI‐5	Satellite nodules	10	5	0.59
	Extratumoral PNI	1	2	0.57
	Extratumoral LVI	4	1	0.40
	Total	12	5	0.31
Adjuvant treatment	None	59	68	< 0.0001
	RT	36	3
	ChemoRT	3	1

Note: Bold indicates statistical significance.

Table 2 shows factors associated with RR among pN0 patients. For isolated RR, only WPOI‐5 was associated with increased risk (p = 0.02). PORT was associated with reduced risk of isolated RR on the log‐rank test (p = 0.04); however, because none of the patients with isolated RR received PORT, we were not able to calculate the odds ratio. For any RR, non‐cohesive POI and WPOI‐5 were associated with increased risk, while PORT was associated with reduced risk.

TABLE 2.

Risk factors for RR among pN0 patients.

		Isolated RR (n = 5)		Any RR (n = 7)		Log rank test
		Odds ratio	Log rank test	Odds ratio	p
Subsite	Tongue	2.07 (0.32, 13.40)	0.44	1.55 (0.40, 6.05)	0.53	0.52
LN yield ≤ 18		5.14 (0.57, 46.24)	0.10	2.67 (0.67, 10.73)	0.17	0.15
Non‐cohesive invasive front		5.63 (0.63, 50.40)	0.08	4.92 (1.01, 23.74)	0.047	0.03
Depth of invasion	≤ 5 mm	Ref	0.37	Ref	Ref	0.51
	6–10 mm	a		3.06 (0.37, 25.55)	0.30
	> 10 mm	1.35 (0.15, 12.19)		3.41 (0.31, 37.78)	0.32
PNI		3.33 (0.47, 23.72)	0.20	3.45 (0.86, 13.90)	0.08	0.06
LVI		a	0.52	1.28 (0.16, 10.48)	0.82	0.82
Tumor at margin		3.72 (0.61, 22.47)	0.13	1.62 (0.33, 7.86)	0.55	0.55
WPOI‐5		6.43 (1.06, 39.04)	0.02	8.00 (2.10, 30.44)	0.002	0.0003
PORT		a	0.04	6.67 (0.83, 53.78)	0.08	0.04

Note: Bold indicates statistical significance.

^a Unable to calculate odds ratio.

Kaplan–Meier survival curves for any RR and isolated RR according to WPOI‐5 are shown in Figure 1a,b.

FIGURE 1.

Kaplan–Meier survival curves for any RR and isolated RR according to WPOI among patients with pN0 necks (Figure 1a and b), and pN0/x necks (Figure 1c and d). [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

5.2. Recurrences in pN0/x Patients

Because of the small number of events of isolated RR among pN0 patients, we repeated the analysis including both pN0 and pNx patients, that is, cN0 patients who did not undergo ND as part of their primary treatment. Clinicopathological features of pNx patients are shown alongside those of pN0 patients in Table 1. Compared to pN0 patients, pNx patients were older, were more likely to have pT1 primary tumors, with smaller DOI, and with cohesive POI, and were less likely to have PNI or receive adjuvant treatment.

Among this combined group of 170 pN0/x patients, mean (median) follow‐up was 66 (58) months. 15 (8.8%) developed RR, of whom 9 (5.3%) had isolated RR. Isolated RR developed only among patients who did not undergo PORT.

Table 3 shows the association between risk factors and RR. For isolated RR, DOI > 10 mm (p = 0.04), WPOI (p = 0.005), and PORT (p = 0.045) were significant on univariate analysis. On multivariate analysis, only WPOI (OR 5.21, 95% CI 1.24, 21.98) was significant. DOI > 10 mm was just outside significance (OR 4.19, 95% CI 0.96, 18.29). For any RR, non‐cohesive invasive front, DOI > 10 mm, PNI, and WPOI were significantly associated with any RR, while PORT was just outside significance. On multivariate analysis, WPOI (OR 8.53, 95% CI 1.76, 41.41), DOI > 10 mm (OR 9.45, 95% CI 2.63, 33.94), and PORT (OR 0.04, 95% CI 0.004, 0.36) were significant.

TABLE 3.

Risk factors for RR among pN0/x patients.

		Isolated RR (n = 9) a			Any RR (n = 15)
		Odds ratio	p	Log rank test	Odds ratio (95% CI)	p	Log rank test
Subsite	Tongue	1.97 (0.45, 8.70)	0.37	0.36	2.02 (0.66, 6.20)	0.22	0.21
Neck dissection	None	1.24 (0.31, 5.04)	0.76	0.76	1.92 (0.69, 5.36)	0.21	0.20
PORT	None	a	a	0.05	0.17 (0.02, 1.29)	0.09	0.05
Non‐cohesive invasive front		3.45 (0.92, 12.97)	0.07	0.05	3.00 (1.08, 8.33)	0.04	0.03
Depth of invasion	≤ 5 mm	Ref	Ref	0.04	Ref	Ref	0.002
	6‐10 mm	2.54 (0.46, 14.04)	0.29		1.58 (0.42, 5.95)	0.50
	> 10 mm	8.19 (1.28, 52.32)	0.03		7.05 (1.92, 25.92)	0.003
PNI		2.21 (0.43, 11.43)	0.34	0.33	3.40 (1.11, 10.45)	0.03	0.02
LVI		a	a	0.53	3.45 (0.64, 13.11)	0.17	0.15
Tumor at margin		1.82 (0.37, 8.89)	0.46	0.46	1.01 (0.22, 4.44)	0.99	0.99
WPOI‐5		5.95 (1.45, 24.51)	0.01	0.005	8.19 (2.84, 23.60)	< 0.0001	< 0.0001

Note: Bold indicates statistical significance.

^a Unable to calculate odds ratio.

Kaplan–Meier survival curves for any RR and isolated RR according to WPOI among patients with pN0/x necks are shown in Figure 1c,d.

5.3. Survival Among Patients With RR

Among a total of nine patients with isolated RR, seven (including three who had undergone previous ND) underwent further neck surgery. Four were rendered disease‐free, of whom two died from other causes, and two were alive at the time of last follow‐up. Mean survival among patients undergoing successful salvage of isolated RR was 55.3 (30–117) months, versus 4.6 (1–8) months among patients not undergoing successful salvage surgery. Among patients developing simultaneous LR and RR, four (including two who had undergone previous ND) underwent salvage surgery. However, all developed further recurrence and died from cancer. Mean survival after simultaneous LR and RR was 14.1 (1–27) months.

5.4. Risk Factors for Occult Metastases in cN0 Necks

Among 130 cN0 patients undergoing elective ND, occult metastases were detected pathologically in 32 (24.6%). Risk factors for occult metastatic disease on univariate analysis were DOI > 10 mm (p = 0.02), non‐cohesive POI (p = 0.02), PNI (p = 0.03), and LVI (p = 0.008) (Table 4). On multivariate analysis, only LVI (p = 0.03) remained significant.

TABLE 4.

Risk factors for occult metastases in cN0 necks undergoing elective ND.

	pN0 (n = 98)	pN+ (n = 32)	Fisher exact test	Multivariate analysis	OR (95% CI)
Depth > 5 mm	64	26	0.12
Depth > 10 mm	19	13	0.02	0.32	1.6 (0.6–4.4)
Non‐cohesive Invasive front	44	22	0.02	0.19	1.9 (0.7–4.8)
PNI	25	16	0.03	0.39	1.5 (0.6–4.0)
LVI	8	11	0.0008	0.026	3.6 (1.2–10.9)
Positive margin	13	2	0.36
WPOI‐5	12	6	0.38
Satellite nodules	10	5	0.52
Extratumoral LVI	1	2	0.15
Extratumoral PNI	4	0	0.57

Abbreviations: LVI: lymphovascular invasion, OR: odds ratio, pN+: pathologically node positive, pN0: pathologically node negative, PNI: perineural invasion, WPOI: worst pattern of invasion.

6. Discussion

Theoretically, the risk of RR among patients with early OSCC and pN0 necks should be very low. However, a systematic review and meta‐analysis performed by Chegini et al. reported a mean isolated RR among patients with pN0 necks after elective ND of 13% among patients with OSCC of any T‐classification and 7.5% among patients with T1/2 primary tumors. However, many of the studies included in this review had different endpoints, and most did not specifically analyze risk factors for RR.

Among studies reporting on early OSCC, Ganly et al. reported an isolated RR rate of 14% among a series of 164 patients with pT1/2N0 OSCC treated with surgery alone, without PORT. The only factor predictive of RR was thickness > 4 mm; however, this association was for all RR, including those occurring simultaneous with LR [22]. Liao at al reported on overall RR of 8% among 387 patients with pT1/2 N0 OSCC treated with surgery alone and also reported DOI ≥ 4 mm to be an independent predictor of RR [23]. Low et al. reported on 121 patients with T1N0/x OSCC treated at the Sydney Head and Neck Cancer Institute, who also did not require PORT, and reported thickness > 5 mm, PNI, infiltrative margins, and poor differentiation to be associated with increased risk of RR. However, most patients in this series did not undergo ND, so the findings of this study may not be generalizable to cases with pathologically proven negative necks [24]. Feng et al. reported a 9.5% isolated RR among 116 pT1/2N0 OSCC patients [25]. Lin et al. reported that patients with pN0 necks had lower RR rates (9%) compared with patients with pN+ necks (50%) among patients with pT1/2 tongue SCC; however, in this series, only a minority of patients (29/72) underwent END, and it is not clear whether RR includes cases occurring simultaneous with LR [26]. Thus, risk factors for isolated RR among early OSCC patients with pN0 necks remain unclear. This represents an important gap, as there is increasing interest in avoiding PORT to the neck among patients with indications for PORT based on the primary tumor but who have pN0 necks after elective ND [27, 28].

In the present study, we report an overall incidence of RR of 7.1%, and an incidence of isolated RR of 5.1% among patients with pT1/2N0 OSCC. The most important risk factor for predicting RR was WPOI‐5. WPOI has previously been reported to predict poor outcomes in OSCC, with WPOI‐5 versus patterns 1–4 representing the key significant cut‐point [18, 29, 30, 31]. Li et al. reported WPOI‐5 to significantly correlate with both locoregional recurrence and disease‐specific survival on multivariate analysis in a stage I/II cohort [18], while in a previous study, we found WPOI‐5 to be associated with worse survival in early oral cancer [16]. In the present study, we report WPOI‐5 to be a significant predictor specifically of RR among early OSCC patients with pN0 necks, and an independent predictor of isolated RR among patients with pN0/x necks. In contrast, WPOI‐5 was not significantly associated with occult nodal disease among cN0 patients undergoing elective ND. However, due to the low frequency of WPOI‐5, it is likely we were underpowered to detect such an association. This hypothesis is supported by the finding that the non‐cohesive invasive front, PNI, and LVI, all of which are components of WPOI‐5, were significant for occult nodal deposits.

In the present study, a substantial proportion of RR occurred in the contralateral undissected neck. This is consistent with the findings of Ganly, Low, and Feng [22, 24, 25]. Contralateral nodal recurrence in well‐lateralized OSCC has been reported in up to 6%–16% of patients, but most commonly in node‐positive patients, with particular risk factors being pN2+ disease or extracapsular extension (ECE) [32, 33, 34]. Interestingly, PNI [32, 33] and LVI [32] have been reported as risk factors for contralateral nodal recurrence, but without distinction between intratumoral and extratumoral.

An important difference between our study and those of Ganly [22], Liao [23] and Low [24] was the exclusion in those studies of patients undergoing PORT. It is likely this led to the exclusion also of patients with PNI and LVI, which are component features of WPOI‐5 [22]. In our series, isolated RR occurred only among patients who did not receive PORT. Of note, among patients with OSCC who require PORT based on primary tumor indications, current protocols are for irradiation to both the primary and the neck, even when the neck is pathologically negative [35]. Recently, a multicentre clinical trial (NCT03997643) randomizing patients with pN0 necks to the standard of care arm with PORT to all dissected areas, or to an experimental arm with omission of PORT to the pN0 neck has completed accrual [27]. Although our study does not provide evidence that PORT may reduce the risk of RR among patients with WPOI‐5, our findings might suggest that patients with WPOI‐5 may not be good candidates for adjuvant therapy de‐escalation, even if they have pN0 necks.

Reasons for the development of isolated RR after pN0 ND are unclear. Chegini suggested that isolated RR in pN0 neck represented a true failure of END to detect nodal disease, which could be due to inadequate surgical nodal clearance, skip metastases to undissected levels, or deficient pathological analysis [21]. Regarding skip metastases, recurrence outside dissected levels in the dissected neck is reported to be rare [22, 36], even among cases with cN+ necks [37]. However, it is possible that inadequate surgical clearance may risk leaving occult metastases in situ, which may give rise to RR. In support of this hypothesis, a number of authors have reported that among patients with pN0 necks, a lymph node yield ≤ 18 is associated with worse outcome [38, 39], although these associations were with locoregional recurrence (LRR) [39] and survival [38, 39] rather than RR specifically. In the present study, we did see separation of the Kaplan–Meier survival curves for RR between patients with lymph node yield ≤ 18 and > 18; however, the difference was not significant, which may be due to underpowering. The other possibility is pathological failure to detect metastases. Amit et al. demonstrated that among pN0 patients where the radiological review was suspicious of cervical nodal metastases, pathological re‐review demonstrated that the original pathological report was falsely negative in 15% [40], while Ganly et al. reported that 7/52 neck dissection specimens reported as negative contained micrometastases on review [22]. Of note, a large proportion of true isolated RR in this study, as well as the studies of Ganly [22] and Low [24], occurred on the contralateral neck, and the reasons for isolated contralateral RR in these cases remain unclear.

Recently, sentinel node biopsy (SNB) as an alternative to elective ND has become increasingly popular in OSCC management. Notably, late neck failure rates after SNB appear to be higher (9.3%–14%) [41, 42] than false negative rates when patients undergo simultaneous SNB and ND (3%–5%) [43, 44]. Further work is required to ascertain whether WPOI‐5 may be a risk factor for late neck failure after negative SNB. One advantage of SNB is the ability to detect contralateral neck drainage [41], but as yet, it is unknown whether this has the potential to obviate contralateral neck RR seen among patients with ipsilateral pN0 necks.

Weaknesses of the paper include the retrospective nature and the small number of events of isolated RR, which likely made our paper underpowered to detect significant associations between variables besides WPOI and isolated RR. We tried to minimize this by repeating the analysis on a larger cohort including pNx patients. This latter group should have a lower risk of cervical metastases than patients undergoing ND based on the lower incidence of risk factors for metastases in this group; however, we cannot discount the possibility that some of these cases may have had occult metastases at the time of surgery that were missed by not having performed ND. A further consideration regarding the generalizability of our results is the possibility of differences in patient cohorts due to institutional differences in treatment protocol. For example, at our institution, END is generally performed as default among medically fit patients excepting those with very thin tumors. Thus, our findings may not be applicable to series where END is performed more selectively in early OSCC [10, 26]. Other institutional differences may include indicators for PORT and whether this is administered to unilateral or bilateral necks. On the other hand, strengths of the study include the systematic slide review for determination of DOI and WPOI‐5, as well as the distinction between isolated RR and locoregional control as outcome measures.

7. Conclusion

In excess of 5% of patients with early OSCC and pathologically negative necks after elective ND may develop RR, even in the absence of demonstrable LR or SPT. WPOI‐5 is emerging as an important adverse prognosticator in early OSCC, and was found in the present study to be the main risk factor predictive of isolated RR. Further work is required to define the role of cervical PORT among patients with WPOI‐5 without other adverse prognosticators.

Conflicts of Interest

The authors declare no conflicts of interest.

Hintze J., van den Berg N., Callanan D., Mohamed S., Feeley L., and Sheahan P., “Impact of WPOI‐5 on Risk of Regional Recurrence in Early Oral Cancer With pN0 Neck,” The Laryngoscope 135, no. 9 (2025): 3213–3221, 10.1002/lary.32182.