Skip to main content
NCBI home page
Journal of Cancer Research and Clinical Oncology logoLink to Journal of Cancer Research and Clinical Oncology
. 2019 Dec 19;146(4):925–933. doi: 10.1007/s00432-019-03111-x

Predictors of survival and recurrence after primary surgery for cervical metastasis of unknown primary

Won Ki Cho 1, Jong-Lyel Roh 2,✉, Kyung-Ja Cho 3, Seung-Ho Choi 1, Soon Yuhl Nam 1, Sang Yoon Kim 1
PMCID: PMC11804353  PMID: 31858242

Abstract

Purpose

Cervical metastasis from unknown primary (CUP) is commonly classified as an advanced overall stage. P16 or human papillomavirus (HPV) positivity in metastatic lymph nodes (LN) might be associated with a favorable survival outcome of CUP. Therefore, we evaluated the prognostic values of p16 immuno-positivity in LN and other clinicopathological factors in patients with squamous cell carcinoma CUP (SCCUP).

Methods

This study involved 83 patients who underwent therapeutic neck dissection and panendoscopic examination and biopsy for suspected CUP. P16 immunostaining and HPV typing in LN were performed in 56 patients. Cox proportional hazard regression analyses were used to identify factors associated with overall survival (OS) and disease-free survival (DFS).

Results

Postoperatively, primary tumors (PT) were found in 32 (38.6%) patients, mainly (90.6%) in the oropharynx, and not found in 51 (61.4%) patients. The clinicopathological data (except for histological grade) and 5-year OS and DFS rates did not significantly differ between patients with and without PT identification (all P > 0.05). P16 positivity was associated with favorable OS and DFS outcomes in the patients with PT (P < 0.05) but not in those without PT (P > 0.1). Multivariate analyses showed that age (> 60 years) and LN ratio (≥ 0.1) were the independent predictors of OS and DFS outcomes (all P < 0.05). P16 positivity or other factors were not independent factors.

Conclusion

Age and LN ratio are significant risk factors of survival and recurrence after primary surgery for SCCUP. Prognostic significance of LN p16 positivity should be further studied.

Keywords: Cervical metastasis of unknown primary, p16, Survival, Recurrence, Risk factors

Introduction

Cervical metastasis from unknown primary (CUP) is a rare subset of head and neck cancer that accounts for 3–7% of the cases (Miller et al. 2008; Galloway and Ridge 2015). CUP is defined as the initial development of metastatic neck mass with no evidence of a primary tumor (PT) despite thorough evaluation of the suspected primary origin site. Owing to the absence of identified PT, initial staging, treatment planning, and prognosis prediction are challenging. Subclinical PT in patients with initial presentation of apparent metastatic neck lymph node (LN) might be identified by complete physical and office-based endoscopic examinations, contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET). Additionally, pathological assessment of LN with human papillomavirus (HPV) detection and p16 immunostaining help to localize primary site (Begum et al. 2003; Zhang et al. 2008). In line with development of workup modalities, the incidence of CUP declined over the past several years (Mintzer et al. 2004). However, 9–64% of patients have unidentified PT sites after thorough endoscopic and imaging examinations (Johansen et al. 2002; Roh et al. 2009).

A substantial proportion of patients with HPV and/or p16 positivity are found among those with squamous cell carcinoma (SCC) CUP (SCCUP) (Keller et al. 2014; Boscolo-Rizzo et al. 2015). The overexpression of p16INK4a as a surrogate marker for HPV-induced transformation is the most important prognostic feature correlated with survival outcomes in patients with SCCUP and oropharyngeal or other head and neck cancer (Keller et al. 2014; Schroeder et al. 2017, 2018). This might explain the extraordinarily low recurrence rates with the current treatment of SCCUP by neck dissection combined with radiotherapy, chemoradiotherapy, or primary chemoradiotherapy (Galloway and Ridge 2015). Additionally, deintensification of radiation is recommended in patients with HPV-driven SCCUP and oropharyngeal SCC (Schroeder et al. 2017).

Further, the recent tumor-node-metastasis (TNM) staging proposed by the American Joint Committee on Cancer (AJCC) (8th ed.) first included the staging system for CUP as T0N1M0, T0N3M0, or T0 any N M1 that are allocated to advanced diseases of overall stage III to IVC (Patel et al. 2017). In contrast, HPV- or Epstein-Barr virus-driven SCCUP, considered as metastatic oropharynx or nasopharynx cancer with unidentified PT, is staged as T0N1 or T0N2 that is allocated to the early overall stage I or II when distant metastasis is not found (M0) (Patel et al. 2017; Lydiatt et al. 2017). Additionally, a recent study reported similar survival outcomes in 48 HPV-induced SCCUP patients and 298 early oropharyngeal cancer patients (Ross et al. 2018). However, the association between LN p16 positivity and survival outcomes as the independent predictor of overall survival (OS) is still controversial (Dixon et al. 2016). In addition, other clinical or pathological nodal findings such as extranodal extension (ENE) are important prognostic factors for SCCUP (Keller et al. 2014). Therefore, we evaluated the prognostic values of p16 immuno-positivity in LNs and other clinicopathological factors in patients with SCCUP.

Materials and methods

Study patients

We conducted a retrospective observational study by reviewing the electronic medical records of patients diagnosed with CUP at our tertiary referral hospital from January 2006 to December 2016. The patients initially underwent physical and endoscopic examinations and ultrasonography-guided fine-needle aspiration biopsy at outpatient clinics followed by the imaging workups of contrast-enhanced CT and/or MRI of the head and neck and 18F-FDG PET/CT scanning. Any primary lesion suspected on endoscopy or imaging studies were biopsied, and cases in which PT was identified were excluded. Subsequently, the patients who had no identified PT from imaging and biopsies underwent panendoscopic examinations and multiple biopsies at the mucosal sites of the nasopharynx, lingual tonsils, hypopharynx, and other suspicious sites under anesthesia. Moreover, the patients underwent bilateral palatine tonsillectomy (Koch et al. 2001).

We have conducted HPV/p16 detection at the LNs in patients with an initial diagnosis of SCCUP since 2011. P16 immunohistochemistry was performed using monoclonal antibody p16INK4 (1:10; BD Pharmingen, Franklin Lakes, NJ) in tissues obtained from fine-needle aspiration or core needle biopsy. P16 positivity was defined when ≥ 80% of tumor cells with strong nuclear and cytoplasmic staining were present (Park et al. 2012). Additionally, HPV typing was performed using GG HPV CHIP (Goodgene Inc., Seoul, Republic of Korea), which is based on PCR-based identification of the HPV subtype including 22 high-risk HPVs and 19 low-risk HPVs.

Inclusion criteria were patients with previously untreated, biopsy-proven neck SCC LN, regardless of HPV and/or p16 status information and those with SCCUP with unidentified PT sites prior to panendoscopic biopsy under anesthesia. The patients subsequently underwent primary curative neck dissection combined with panendoscopic biopsy. Exclusion criteria were patients who received previous treatments and those with initial presentation of inoperable advanced diseases, distant metastasis, other non-SCC pathologies, and inadequate follow-up information from early loss < 2 years after treatment. Overall, 83 patients were included in our study. The nodal staging was based on the AJCC staging manual. This study was reviewed and approved by the institutional review board of our hospital and informed consent from each patient was waived.

All patients underwent surgery combined with or without radiotherapy or chemoradiotherapy. Therapeutic neck dissection was performed at cervical levels I to V in the involved neck sides. Postoperative radiotherapy was conducted in patients using intensity-modulated radiotherapy with a median 65 Gy (range 56–74 Gy) and single daily fraction of 1.8–2.0 Gy for 5 days per week for 6 to 8 weeks. Concurrent chemoradiotherapy was performed using cisplatin (80–100 mg/m2) infused on days 1, 22, and 43. The patients were followed at the outpatient clinic every 1–3 months in the first year, 2–4 months in the second year, 6 months in years 3–5, and thereafter, annually (Denaro et al. 2016). Any loco-regional recurrences suspicious on examinations and imaging were confirmed by biopsy. Distant site recurrences were confirmed by biopsy or serial follow-up imaging when a biopsy was unavailable. The patients with relapse underwent salvage surgery or palliative treatments.

Variables

Clinical and pathological information were included as variables. The following demographic and clinical data were included: age (> 60 years), sex, smoking (> 10 pack-years), alcohol (≥ 1 drink/day), Charlson comorbidity index (≥ 1), and Eastern Cooperative Oncology Group (ECOG) performance scale (≥ 1). The following pathological data were included: p16 positivity, histological grade (G1/G2/G3), the maximal size of LNs, number of LNs, number of examined LNs, cervical levels involved, LN ratio (number of LN divided by the number of examined LN) (Zheng et al. 2018), ENE, and pathological nodal staging.

Statistical analysis

Continuous variables using the t-test and categorical variables using the χ2 or Fisher’s exact test were compared between the groups with and without PT identification. Primary and secondary endpoints were OS and disease-free survival (DFS), respectively. The cut-off values for the optimal number of positive and examined LNs or the LN ratio were determined using the time-dependent receiver operating characteristics (ROC) curve analyses and the area under the ROC curve (AUC) estimates (Heagerty et al. 2000). Univariate and multivariate Cox proportional hazard regression models were used to define the significant predictors of OS and DFS, and they were analyzed separately in the groups with and without PT identification. Variables with P < 0.1 in the univariate analyses were selected for the multivariate Cox proportional hazard regression analyses using the backward elimination method. Variables with multi-collinearity (number of positive LNs and LN ratio) were separately fitted in the multivariable model (Vatcheva et al. 2016). Additionally, multivariate analyses for missing values were treated with multiple imputation using the Markov chain Monte Carlo method. Hazard ratio (HR) and 95% confidence interval (CI) were estimated. The Kaplan–Meier curves were used to define the significant variables, and log-rank tests were used to define the statistical significance. All statistical tests were two-tailed, and a P-value < 0.05 was considered statistically significant. The statistical analyses were performed using IBM® SPSS® Statistics version 24.0 for Windows (IBM Corp., Armonk, NY).

Results

The characteristics of patients with or without PT identification

This study included 83 patients including 66 (79.5%) men and 17 (20.5%) women with a median age of 56 years (range 31–84 years). PT was identified in 32 (38.6%) patients and not identified in 51 (61.4%) patients. Clinical and pathological data are summarized in Table 1 with a comparison between patients with and without PT identification; Charlson comorbidity index ≥ 1 and ECOG performance scale ≥ 1 were found in 7 (8.4%) and 19 (22.9%) patients, respectively. P16 immunohistochemistry from metastatic LNs was performed in 56 (67.5%) patients, of which most were diagnosed since 2011 and 40 (71.4%) showed positivity. Moreover, 41 of 56 (73.2%) patients tested positive for HPV. Thirty-six of 41 (87.8%) HPV-positive patients had p16 positivity, showing the consistence between HPV infection and p16 upregulation (P < 0.001). Of 32 patients with PT identification, PT was located in the oropharynx (n = 29, 90.6%), hypopharynx (n = 2, 6.3%), and nasopharynx (n = 1, 3.1%), and the largest dimension of tumors was ≤ 2 cm. The total number of LNs harvested from neck dissection was median 53 (range 11–138), and the number of positive LNs was 2 (range 1–48).

Table 1.

Patients characteristics (N = 83)

Variable Total (n = 83) PT (n = 32) No PT (n = 51) Pa
Age (years), median (range) 56 (31–84) 53 (38–77) 58 (31–84) 0.452
Sex, male/female 66/17 (79.5/20.5) 26/6 (81.3/18.8) 40/11 (78.4/21.6) 0.757
Smoking, > 10 pack-years 45 (54.2) 16 (50.0) 29 (56.9) 0.652
Alcohol, ≥ 1 drink/day 58 (69.9) 25 (78.1) 33 (64.7) 0.195
Charlson comorbidity index, ≥ 1 7 (8.4) 2 (6.3) 5 (9.8) 0.571
ECOG performance scale, ≥ 1 19 (22.9) 7 (21.9) 12 (23.5) 0.861
P16, positive/examinedb 40/56 (71.4) 25/28 (89.3) 15/28 (53.6) 0.007*
Histologic grade, G1/G2/G3 24/25/34 (28.9/30.1/41.0) 3/16/13 (9.3/50.0/40.6) 21/9/21 (41.2/17.6/41.2) 0.001*
Primary site identified, OP/HP/NP 29/2/1 29/2/1 0  < 0.001*
Treatment
 Surgery alone 4 (4.8) 2 (6.3) 2 (3.9) 0.458
 Surgery plus RT 56 (67.5) 19 (59.4) 37 (72.5)
 Surgery plus CRT 23 (27.7) 11 (34.4) 12 (23.5)
Size of LN (cm), median (range) 2.8 (0.8–8.0) 2.7 (1.0–8.0) 2.8 (0.8–7.0) 0.196
No. of LNs harvested, median (range) 53 (11–138) 54 (19–138) 51 (11–129) 0.759
No. of LNs involved, median (range) 2 (1–48) 2 (1–30) 2 (1–48) 0.348
Extranodal extension 30 (36.1) 9 (28.1) 21 (41.2) 0.251
Follow-up duration, median (range) 53 (24–151) 54 (24–97) 51 (24–151) 0.293
Last status
 NED 62 (74.7) 25 (78.1) 37 (72.5) 0.727
 DOD/DOC 14/5 (16.9/6.0) 3/3 (9.4/9.4) 11/2 (21.6/3.9)
 AD 2 (2.4) 1 (3.1) 1 (2.0)
Recurrence, any site 16 (19.3) 4 (12.5) 12 (23.5) 0.263
Open in a new tab

Variables are expressed as the number of patients (percentage) unless indicated otherwise

AD alive with disease, CRT concurrent chemoradiotherapy, DOC died of other cause, DOD died of disease, ECOG Eastern Cooperative Oncology Group, G1/G2/G3 well/moderate/high grade, HP hypopharynx, LN lymph node, NED no evidence of disease, NP nasopharynx, OP oropharynx, PT primary tumor identified by panendoscopic examination and biopsy under anesthesia, RT radiotherapy

aχ2 or Fisher’s exact test, except for age, no. of LNs harvested and involved, and follow-up duration (t-test), *P < 0.05

bExamined from metastatic LNs

The clinicopathological data were comparable between patients with and without PT identification (P > 0.1) except for the histological grade (P = 0.001). The patients were followed for a median of 53 months (range 24–151 months). At last follow-up, 62 (74.7%) patients were alive without disease, 14 (16.9%) patients died of disease, 5 (6.0%) patients died of other causes, and 2 (2.4%) patients were alive with disease. The 5-year OS rate of all patients was 77.1% (95% CI 68.3–79.9%) and was not significantly different between the patients with and without PT identification (82.0% vs 73.6%) (HR = 1.30 [95% CI 0.48–3.51], P = 0.608). The 5-year DFS rate of all patients was 77.8% (72.9–82.7%) and was not significantly different between the patients with and without PT identification (87.2% vs 71.8%) (HR = 2.17 [95% CI 0.71–6.65], P = 0.176).

The 5-year OS rates were significantly different between the p16-positive patients (87.8%, 95% CI: 81.8–93.8%) and p16-negative patients (37.7%, 95% CI: 19.8–55.6%), (HR = 8.46 [95% CI 2.44–92.28], P = 0.001). The 5-year DFS rates were significantly different between the p16-positive patients (92.3%, 95% CI: 88.0–96.6%) and p16-negative patients (41.7%, 95% CI: 22.9–60.5%), (HR = 7.37 [95% CI 1.88–28.89], P = 0.004). The association between p16 positivity and OS and DFS outcomes was statistically significance in the patients with PT identification (P < 0.05) but not in those without PT identification (P > 0.05). The Kaplan–Meier curve estimates are shown in the Fig. 1.

Fig 1.

Fig 1

Open in a new tab

Kaplan–Meier curves estimating overall survival (OS) in all patients (a) and patients with (b) and without (c) primary tumor identified by panendoscopic examination and biopsy under anesthesia according to the p16 positivity on immunohistochemical staining of metastatic lymph nodes. P values were calculated using the log-rank test

Association between clinicopathological factors and survival outcomes

The association of clinical and pathological factors with OS and DFS outcomes was analyzed separately in the patients with and without PT identification because of potentially different biological behavior and significant factors. The time-dependent ROC curve analyses showed that the cut-off values for the optimal number of examined and positive LNs and LN ratio were 50, 4, and 0.1, respectively. In the PT group, univariate analyses showed that age (> 60 years), Charlson comorbidity index (≥ 1), ECOG performance scale (≥ 1), histological grade (well-differentiated), p16 (negative), number of LN examined (< 50), ENE, and nodal classification (N3) were the significant factors associated with poor OS outcomes (all P < 0.05) (Table 2). Additionally, P16 (negative), LN ratio (≥ 0.1), and ENE were significantly associated with poor DFS outcomes (all P < 0.05). In the no PT group, univariate analyses showed that age, Charlson comorbidity index, ECOG performance scale, number of LN, LN ratio, ENE, and N classification were the significant factors associated with poor OS and DFS outcomes (all P < 0.005). In the PT group, multivariate analyses showed that p16 negativity was the sole independent predictor of OS and DFS outcomes (P < 0.05) (Table 3). In the no PT group, multivariate analyses showed that age and LN ratio were the independent predictors of OS and DFS outcomes (all P < 0.05). Patients > 60 years at presentation had 9.6-fold higher mortality and 6.5-fold higher recurrence than their counterparts. Patients with LN ratio ≥ 0.1 had 7.1-fold higher mortality and recurrence than their counterparts. The Kaplan–Meier curves estimating OS and DFS in terms of age and LN ratio are presented in Fig. 2.

Table 2.

Univariate analyses of factors on overall survival and disease-free survival according to primary tumor identified or not

Variable PT (n = 32) No PT (n = 51)
OS DFS OS DFS
HR (95% CI) Pa HR (95% CI) Pa HR (95% CI) Pa HR (95% CI) Pa
Age, > 60 years 8.34 (1.49–36.64) 0.016* 9.04 (0.94–87.28) 0.057 6.87 (1.87–25.18) 0.004* 11.98 (2.63–54.56) 0.001*
Sex, male 2.76 (0.30–24.34) 0.476 1.39 (0.15–13.40) 0.775 1.53 (0.33–7.07) 0.588 3.97 (0.52–30.55) 0.186
Smoking, > 10 pack-years 5.40 (0.63–46.37) 0.124 3.39 (0.35–32.64) 0.291 2.00 (0.62–6.53) 0.249 1.91 (0.59–6.21) 0.282
Alcohol, ≥ 1 drink/day 3.96 (0.28–55.74) 0.320 1.96 (0.20–18.89) 0.559 1.61 (0.54–4.85) 0.396 1.26 (0.42–3.75) 0.679
Charlson comorbidity index, ≥ 1 17.29 (1.86–160.37) 0.012* 2.41 (0.34–17.12) 0.380 7.72 (2.04–29.22) 0.003* 3.26 (1.09–9.75) 0.034*
ECOG PS, ≥ 1 2.67 (1.22–24.49) 0.013* 4.50 (0.63–32.27) 0.134 4.71 (1.43–15.46) 0.011* 2.45 (0.80–7.49) 0.117
LN size, ≤ 3 cm 1 1 1 1
 3.1–6 cm 5.45 (0.87–34.33) 0.071 4.04 (0.70–18.44) 0.956 1.00 (0.30–3.38) 0.994 1.56 (0.50–4.94) 0.447
  > 6 cm 14.96 (1.11–202.76) 0.042* 7.65 (0.92–36.91) 0.954 0.65 (0.08–5.30) 0.691 0.64 (0.08–5.21) 0.676
Histological grade, G1 1 1 1
 G2 0.11 (0.02–0.66) 0.018* 0.09 (0.08–0.94) 0.044* 1.11 (0.21–5.87) 0.906 0.49 (0.06–4.25) 0.521
 G3 0.06 (0.01–0.56) 0.014* 0.10 (0.09–1.06) 0.056 1.36 (0.41–4.48) 0.612 1.57 (0.50–4.95) 0.446
P16, negative 17.64 (2.92–106.57) 0.002* 18.67 (1.68–2017.75) 0.017* 3.32 (0.64–17.34) 0.156 3.45 (0.66–18.02) 0.142
Neck levels involved, ≥ 2/1 5.55 (0.99–130.94) 0.051 2.56 (0.11–62.29) 0.281 2.55 (0.78–8.33) 0.122 2.66 (0.82–8.68) 0.104
Neck level IV or V involved 2.36 (0.42–13.13) 0.328 3.99 (0.56–28.34) 0.167 2.00 (0.64–6.24) 0.233 3.65 (1.22–10.89) 0.020*
No. of positive LNs
 1–4 1 1 1 1
  ≥ 5 3.9 (0.20–92.30) 0.528 1.76 (0.18–16.92) 0.626 3.63 (1.16–11.35) 0.027* 3.16 (1.05–9.47) 0.040*
No. of LNs examined
  < 50 9.34 (1.09–80.01) 0.041* 1.84 (0.26–13.06) 0.543 2.39 (0.73–7.80) 0.150 2.58 (0.79–8.42) 0.116
LN ratio
 ≥ 0.1 5.90 (0.96–36.38) 0.056 2.79 (1.18–27.09) 0.004* 5.23 (1.65–16.59) 0.005* 4.64 (1.52–14.14) 0.007*
Extranodal extension, yes/no 27.65 (3.12–244.97) 0.003* 14.99 (1.39–102.83) 0.017* 18.35 (2.34–143.87) 0.006* 21.67 (2.81–167.11) 0.003*
N classification
 N1–N2 1
 N3 31.6 (3.6–277.66) 0.002* NC 16.17 (2.07–126.62) 0.008* NC
Open in a new tab

CI confidence interval, DFS disease-free survival, ECOG PS Eastern Cooperative Oncology Group performance scale, HR hazard ratio, LN lymph node, NC not calculated, OS overall survival, PT primary tumor identified

aUnivariate Cox proportional hazard regression models, *P < 0.05

Table 3.

Multivariate analyses of factors on overall survival and disease-free survival

Variable OS DFS
HR (95% CI) Pa HR (95% CI) Pa
PT patients (n = 32)
 P16, negative 12.75 (1.78–90.90) 0.011* 13.49 (1.41–112.69) 0.028*
No PT patients (n = 51)
 Age, > 60 years 9.63 (2.25–41.30) 0.002* 6.50 (1.16–36.31) 0.033*
 LN ratio, ≥ 0.1 7.11 (1.91–26.47) 0.003* 7.13 (1.81–28.04) 0.005*
Open in a new tab

CI confidence interval, DFS disease-free survival, HR hazard ratio, LN lymph node, OS overall survival, PT primary tumor identified

aThe multivariate models were performed using a backward stepwise selection procedure with all the variables of P < 0.1 on univariate results (Table 2), *P < 0.05

Fig. 2.

Fig. 2

Open in a new tab

Kaplan–Meier curves estimating overall survival (OS) and disease-free survival (DFS) according to the cut-off values of patient age (a, b) and lymph node ratio (LNR, c and d). Log-rank test, P < 0.005

Discussion

The prognostic values of p16 immuno-positivity in LNs and other clinicopathological factors in patients with SCCUP were evaluated. There are different definitions and inclusion criteria for SCCUP from negative endoscopic examination only to panendoscopic examination and biopsy. Our cohort included patients based on the strict definition of SCCUP using endoscopic examination and imaging studies including 18F-FDG PET/CT followed by panendoscopic examination, multiple site biopsies, and bilateral tonsillectomy under anesthesia. The patients with PT identification were included as controls for those without PT identification. P16 negativity was the sole independent predictor for survival and recurrence in the PT group but not in the no PT group. Age and LN ratio were the independent predictors of OS and DFS in the no PT group. Our results might help guide risk stratification and prognosis prediction in patients with SCCUP, a rare disease (Mintzer et al. 2004; Schroeder et al. 2017).

The prognostic significance of p16 in head and neck cancer is well-established owing to the identification of a causal link between HPV and oropharyngeal SCC (Mork et al. 2001; D'Souza et al. 2007). HPV or p16 positivity is one of the most important prognostic determinants in oropharyngeal SCC and other head and neck SCC (Ang et al. 2010; O'Rorke et al. 2012; Sedghizadeh et al. 2016). Accordingly, the recent AJCC 8th edition TNM staging system separated HPV-mediated (p16 positive) and non HPV-mediated (p16 negative) oropharyngeal SCC for improved prognostic grouping by adopting the importance of p16 (Lydiatt et al. 2017). HPV was recognized as a significant prognostic marker in neck SCCUP (Tribius et al. 2012; Schroeder et al. 2017). A previous study showed that in 23/63 (37%) SCCUP patients with positive HPV/p16 and < 10 pack-years smoking history had superior OS outcomes (Tribius et al. 2012). A retrospective European multicenter study with 180 SCCUP patients showed that a substantial proportion of cases were HPV-driven, which was associated with improved OS and progression-free survival (Schroeder et al. 2017). Additionally, a recent study showed that 48 patients with HPV-positive SCCUP had survival outcomes similar to that of 48 matched control patients with HPV-positive T1-2N1-3M0 oropharyngeal SCC suggesting similar risk stratification (Ross et al. 2018). However, the relationship between LN HPV/p16 positivity and survival outcomes in SCCUP is still controversial (Dixon et al. 2016). A study including 73 SCCUP patients showed an association of p16-positive status with DFS but not OS (Dixon et al. 2016). In the present study, p16 status was not a significant factor associated with OS and DFS outcomes. The differences might be attributable to the potential differences in inclusion criteria and definition of SCCUP, p16 immunostaining sensitivity, and treatment modalities. Some studies also include a small group of subclinical, p16-positive, PT unidentified patients and show the prognostic significance of p16 positivity, as in this study.

Other clinical and pathological findings might be associated with survival and recurrence after definitive treatments for SCCUP. In a previous study with 39 SCCUP cases, ENE (≥ 2 mm) and p16 negativity (26%) were significant indicators of poor OS (Keller et al. 2014). ENE is associated with an increased risk of post-treatment recurrence and death in patients with head and neck cancer (Dunne et al. 2006; Kwon et al. 2015; Joo et al. 2019). Accordingly, the recent version (8th edition) of the AJCC TNM staging system revised the N classification by including ENE to increase the robustness of the staging system in predicting and discriminating survival rates of patients with head and neck cancer (Patel et al. 2017; Lydiatt et al. 2017). The presence of ENE negatively impacts loco-regional recurrence and distant metastasis but there was no association with HPV-positive oropharyngeal SCC (Mermod et al. 2016). Macroscopic ENE with transcapsular spread ≥ 2 mm but not microscopic ENE might impact survival after treatment for head and neck SCC (Yoo et al. 2015). The same findings were observed in SCCUP cases without PT identification that showed only gross ENE but not microscopic ENE in association with OS and cause-specific survival (Keller et al. 2014). In the present study, the presence of microscopic or macroscopic ENE was associated with OS and DFS outcomes in the univariate analysis, which was not supported by multivariate analysis. This might result from the different inclusion criteria for SCCUP or the small number of patients included.

The present study identified the prognostic significance of LN ratio in SCCUP patients. Metastatic LN burden indicators, such as number of LNs or LN ratio, is one of most important prognosticators in human solid cancers including head and neck cancer (Ho et al. 2017, 2018; Kim et al. 2018). Recent studies showed that overall mortality escalated as the number of LNs increased (Ho et al. 2018) and was associated with a decreasing number of LNs examined in neck dissection of head and neck cancer (Divi et al. 2016). In these studies, LNs > 4 and LNs ≤ 35 were strongly associated with poor patient survival (Divi et al. 2016; Ho et al. 2018). Of several nodal factors, LN ratio was recognized as a useful marker in stratifying the likelihood of recurrence and survival in the head and neck SCC patients (Kim et al. 2011; Talmi et al. 2018). However, the cut-offs of LN ratio were largely variable (range 0.02–0.20, an average of 0.09) in different reports, and thus, needed further definition (Talmi et al. 2018). A previous study showed the prognostic value of LN volume (> 30 ml) and ratio (> 0.14) in 39 SCCUP patients (Park et al. 2014). In the present study, the LN ratio ≥ 0.1 was associated with a 7.1-fold increased mortality and recurrence. However, the previous or recent AJCC N classification does not include the number of LNs or LN ratio owing to the lack of convincing evidence (Talmi et al. 2018). Therefore, this needs to be further validated before inclusion into the staging system.

This study has limitations owing to its retrospective design. The number of patients might not allow sufficient statistical power. The p16 status was not examined in all study patients because we conducted the LN HPV/p16 detection since 2011. This might have affected the results, particularly in the no PT group. Nonetheless, the patients were selected using a strict definition of SCCUP after panendoscopic examination and biopsy under anesthesia and underwent uniform treatment including neck dissection. The results should be further analyzed using multi-institutional data of SCCUP.

In conclusion, the present study suggests that age and LN ratio are significant predictors of survival and recurrence after primary surgery for SCCUP. However, p16 immuno-positivity in metastatic LNs was unrelated to the treatment outcomes in SCCUP. Therefore, the prognostic significance of LN p16 positivity should be further evaluated. Our results might help define the at-risk patients who need intense, postoperative adjuvant therapy and surveillance.

Funding

This study was supported by the National Research Foundation of Korea (NRF) grant, funded by the Ministry of Science and ICT (MSIT), the Government of Korea (No. 2019R1A2C2002259) (J.-L.R.).

Compliance with ethical standards

Conflict of interest

The authors have no conflicts of interest to declare.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research board and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors. Informed consent from all individual participants was waved because of the retrospective nature of this study.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan PF, Westra WH, Chung CH, Jordan RC, Lu C, Kim H, Axelrod R, Silverman CC, Redmond KP, Gillison ML (2010) Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 363(1):24–35 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Begum S, Gillison ML, Ansari-Lari MA, Shah K, Westra WH (2003) Detection of human papillomavirus in cervical lymph nodes: a highly effective strategy for localizing site of tumor origin. Clin Cancer Res 9(17):6469–6475 [PubMed] [Google Scholar]
  3. Boscolo-Rizzo P, Schroeder L, Romeo S, Pawlita M (2015) The prevalence of human papillomavirus in squamous cell carcinoma of unknown primary site metastatic to neck lymph nodes: a systematic review. Clin Exp Metastasis 32(8):835–845 [DOI] [PubMed] [Google Scholar]
  4. Denaro N, Merlano MC, Russi EG (2016) Follow-up in head and neck cancer: do more does it mean do better? A systematic review and our proposal based on our experience. Clin Exp Otorhinolaryngol 9(4):287–297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Divi V, Chen MM, Nussenbaum B, Rhoads KF, Sirjani DB, Holsinger FC, Shah JL, Hara W (2016) Lymph node count From neck dissection predicts mortality in head and neck cancer. J Clin Oncol 34(32):3892–3897 [DOI] [PubMed] [Google Scholar]
  6. Dixon PR, Au M, Hosni A, Perez-Ordonez B, Weinreb I, Xu W, Song Y, Huang SH, O'Sullivan B, Goldstein DP, de Almeida JR (2016) Impact of p16 expression, nodal status, and smoking on oncologic outcomes of patients with head and neck unknown primary squamous cell carcinoma. Head Neck 38(9):1347–1353 [DOI] [PubMed] [Google Scholar]
  7. D'Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch WM, Westra WH, Gillison ML (2007) Case-control study of human papillomavirus and oropharyngeal cancer. N Engl J Med 356(19):1944–1956 [DOI] [PubMed] [Google Scholar]
  8. Dunne AA, Muller HH, Eisele DW, Kessel K, Moll R, Werner JA (2006) Meta-analysis of the prognostic significance of perinodal spread in head and neck squamous cell carcinomas (HNSCC) patients. Eur J Cancer 42(12):1863–1868 [DOI] [PubMed] [Google Scholar]
  9. Galloway TJ, Ridge JA (2015) Management of squamous cancer metastatic to cervical nodes with an unknown primary site. J Clin Oncol 33(29):3328–3337 [DOI] [PubMed] [Google Scholar]
  10. Heagerty PJ, Lumley T, Pepe MS (2000) Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics 56(2):337–344 [DOI] [PubMed] [Google Scholar]
  11. Ho AS, Kim S, Tighiouart M, Gudino C, Mita A, Scher KS, Laury A, Prasad R, Shiao SL, Van Eyk JE, Zumsteg ZS (2017) Metastatic lymph node burden and survival in oral cavity cancer. J Clin Oncol 35(31):3601–3609 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ho AS, Kim S, Tighiouart M, Gudino C, Mita A, Scher KS, Laury A, Prasad R, Shiao SL, Ali N, Patio C, Mallen-St Clair J, Van Eyk JE, Zumsteg ZS (2018) Association of quantitative metastatic lymph node burden with survival in hypopharyngeal and laryngeal cancer. JAMA Oncol 4(7):985–989 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Johansen J, Eigtved A, Buchwald C, Theilgaard SA, Hansen HS (2002) Implication of 18F-fluoro-2-deoxy-D-glucose positron emission tomography on management of carcinoma of unknown primary in the head and neck: a Danish cohort study. Laryngoscope 112(11):2009–2014 [DOI] [PubMed] [Google Scholar]
  14. Joo YH, Cho JK, Koo BS, Kwon M, Kwon SK, Kwon SY, Kim MS, Kim JK, Kim H, Nam I, Roh JL, Park YM, Park IS, Park JJ, Shin SC, Ahn SH, Won S, Ryu CH, Yoon TM, Lee G, Lee DY, Lee MC, Lee JK, Lee JC, Lim JY, Chang JW, Jang JY, Chung MK, Jung YS, Cho JG, Choi YS, Choi JS, Lee GH, Chung PS (2019) Guidelines for the surgical management of oral cancer: Korean society of thyroid-head and neck surgery. Clin Exp Otorhinolaryngol 12(2):107–144 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Keller LM, Galloway TJ, Holdbrook T, Ruth K, Yang D, Dubyk C, Flieder D, Lango MN, Mehra R, Burtness B, Ridge JA (2014) p16 status, pathologic and clinical characteristics, biomolecular signature, and long-term outcomes in head and neck squamous cell carcinomas of unknown primary. Head Neck 36(12):1677–1684 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kim SY, Nam SY, Choi SH, Cho KJ, Roh JL (2011) Prognostic value of lymph node density in node-positive patients with oral squamous cell carcinoma. Ann Surg Oncol 18(8):2310–2317 [DOI] [PubMed] [Google Scholar]
  17. Kim BH, Park SJ, Jeong WJ, Ahn SH (2018) Comparison of treatment outcomes for T3 glottic squamous cell carcinoma: a meta-analysis. Clin Exp Otorhinolaryngol 11(1):1–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Koch WM, Bhatti N, Williams MF, Eisele DW (2001) Oncologic rationale for bilateral tonsillectomy in head and neck squamous cell carcinoma of unknown primary source. Otolaryngol Head Neck Surg 124(3):331–333 [DOI] [PubMed] [Google Scholar]
  19. Kwon M, Roh JL, Lee J, Cho KJ, Choi SH, Nam SY, Kim SY (2015) Extranodal extension and thickness of metastatic lymph node as a significant prognostic marker of recurrence and survival in head and neck squamous cell carcinoma. J Craniomaxillofac Surg 43(6):769–778 [DOI] [PubMed] [Google Scholar]
  20. Lydiatt WM, Patel SG, O'Sullivan B, Brandwein MS, Ridge JA, Migliacci JC, Loomis AM, Shah JP (2017) Head and neck cancers-major changes in the American Joint Committee on cancer eighth edition cancer staging manual. CA Cancer J Clin 67(2):122–137 [DOI] [PubMed] [Google Scholar]
  21. Mermod M, Tolstonog G, Simon C, Monnier Y (2016) Extracapsular spread in head and neck squamous cell carcinoma: a systematic review and meta-analysis. Oral Oncol 62:60–71 [DOI] [PubMed] [Google Scholar]
  22. Miller FR, Karnad AB, Eng T, Hussey DH, Stan McGuff H, Otto RA (2008) Management of the unknown primary carcinoma: long-term follow-up on a negative PET scan and negative panendoscopy. Head Neck 30(1):28–34 [DOI] [PubMed] [Google Scholar]
  23. Mintzer DM, Warhol M, Martin AM, Greene G (2004) Cancer of unknown primary: changing approaches. A multidisciplinary case presentation from the Joan Karnell cancer center of pennsylvania hospital. Oncologist 9(3):330–338 [DOI] [PubMed] [Google Scholar]
  24. Mork J, Lie AK, Glattre E, Hallmans G, Jellum E, Koskela P, Moller B, Pukkala E, Schiller JT, Youngman L, Lehtinen M, Dillner J (2001) Human papillomavirus infection as a risk factor for squamous-cell carcinoma of the head and neck. N Engl J Med 344(15):1125–1131 [DOI] [PubMed] [Google Scholar]
  25. O'Rorke MA, Ellison MV, Murray LJ, Moran M, James J, Anderson LA (2012) Human papillomavirus related head and neck cancer survival: a systematic review and meta-analysis. Oral Oncol 48(12):1191–1201 [DOI] [PubMed] [Google Scholar]
  26. Park GC, Lee M, Roh JL, Yu MS, Choi SH, Nam SY, Kim SY, Cho KJ (2012) Human papillomavirus and p16 detection in cervical lymph node metastases from an unknown primary tumor. Oral Oncol 48(12):1250–1256 [DOI] [PubMed] [Google Scholar]
  27. Park GC, Jung JH, Roh JL, Lee JH, Cho KJ, Choi SH, Nam SY, Kim SY (2014) Prognostic value of metastatic nodal volume and lymph node ratio in patients with cervical lymph node metastases from an unknown primary tumor. Oncology 86(3):170–176 [DOI] [PubMed] [Google Scholar]
  28. Patel SG, L WM, Ridge CM, Mukherji SK, Ghossein RA, Brandwein-Gensler M (2017) In: Amin MB, Greene FL, Byrd DR, Brookland RK, Washington MK (eds) AJCC cancer staging manual, 8th edn. Springer, New York, pp 67–78 [Google Scholar]
  29. Roh JL, Kim JS, Lee JH, Cho KJ, Choi SH, Nam SY, Kim SY (2009) Utility of combined (18)F-fluorodeoxyglucose-positron emission tomography and computed tomography in patients with cervical metastases from unknown primary tumors. Oral Oncol 45(3):218–224 [DOI] [PubMed] [Google Scholar]
  30. Ross RB, Koyfman SA, Reddy CA, Houston N, Geiger JL, Woody NM, Joshi NP, Greskovich JF Jr, Burkey BB, Scharpf J, Lamarre ED, Prendes B, Lorenz RR, Adelstein DJ, Ward MC (2018) A matched comparison of human papillomavirus-induced squamous cancer of unknown primary with early oropharynx cancer. Laryngoscope 128(6):1379–1385 [DOI] [PubMed] [Google Scholar]
  31. Schroeder L, Boscolo-Rizzo P, Dal Cin E, Romeo S, Baboci L, Dyckhoff G, Hess J, Lucena-Porcel C, Byl A, Becker N, Alemany L, Castellsague X, Quer M, Leon X, Wiesenfarth M, Pawlita M, Holzinger D (2017) Human papillomavirus as prognostic marker with rising prevalence in neck squamous cell carcinoma of unknown primary: a retrospective multicentre study. Eur J Cancer 74:73–81 [DOI] [PubMed] [Google Scholar]
  32. Schroeder L, Wichmann G, Willner M, Michel A, Wiesenfarth M, Flechtenmacher C, Gradistanac T, Pawlita M, Dietz A, Waterboer T, Holzinger D (2018) Antibodies against human papillomaviruses as diagnostic and prognostic biomarker in patients with neck squamous cell carcinoma from unknown primary tumor. Int J Cancer 142(7):1361–1368 [DOI] [PubMed] [Google Scholar]
  33. Sedghizadeh PP, Billington WD, Paxton D, Ebeed R, Mahabady S, Clark GT, Enciso R (2016) Is p16-positive oropharyngeal squamous cell carcinoma associated with favorable prognosis? A systematic review and meta-analysis. Oral Oncol 54:15–27 [DOI] [PubMed] [Google Scholar]
  34. Talmi YP, Takes RP, Alon EE, Nixon IJ, Lopez F, de Bree R, Rodrigo JP, Shaha AR, Halmos GB, Rinaldo A, Ferlito A (2018) Prognostic value of lymph node ratio in head and neck squamous cell carcinoma. Head Neck 40(5):1082–1090 [DOI] [PubMed] [Google Scholar]
  35. Tribius S, Hoffmann AS, Bastrop S, Gorogh T, Haag J, Rocken C, Clauditz T, Grob T, Wilczak W, Tennstedt P, Borcherding A, Petersen C, Hoffmann M (2012) HPV status in patients with head and neck of carcinoma of unknown primary site: HPV, tobacco smoking, and outcome. Oral Oncol 48(11):1178–1184 [DOI] [PubMed] [Google Scholar]
  36. Vatcheva KP, Lee M, McCormick JB, Rahbar MH (2016) Multicollinearity in regression analyses conducted in epidemiologic studies. Epidemiology (Sunnyvale) 6(2):227 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yoo SH, Roh JL, Kim SO, Cho KJ, Choi SH, Nam SY, Kim SY (2015) Patterns and treatment of neck metastases in patients with salivary gland cancers. J Surg Oncol 111(8):1000–1006 [DOI] [PubMed] [Google Scholar]
  38. Zhang MQ, El-Mofty SK, Davila RM (2008) Detection of human papillomavirus-related squamous cell carcinoma cytologically and by in situ hybridization in fine-needle aspiration biopsies of cervical metastasis: a tool for identifying the site of an occult head and neck primary. Cancer 114(2):118–123 [DOI] [PubMed] [Google Scholar]
  39. Zheng CM, Ji YB, Song CM, Ge MH, Tae K (2018) Number of metastatic lymph nodes and ratio of metastatic lymph nodes to total number of retrieved lymph nodes are risk factors for recurrence in patients with clinically node negative papillary thyroid carcinoma. Clin Exp Otorhinolaryngol 11(1):58–64 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Cancer Research and Clinical Oncology are provided here courtesy of Springer

RESOURCES