The prognostic value of lymph node ratio in survival of head-and-neck squamous cell carcinoma.

BACKGROUND: Head-and-neck squamous cell carcinoma (HNSCC) is the sixth most prevalent type of cancers in the world. Due to its relatively high rate of recurrence, the prognosis of patients is poor and the survival rate is low; therefore, identifying the prognostic factors is considered necessary for better treatment.
MATERIALS AND METHODS: This historical cohort study was conducted on 201 patients diagnosed with aerodigestive SCC who underwent surgery and lymph node dissection. We determined the prognostic value of lymph node ratio (LNR) on overall survival (OS), disease-free survival (DFS), and locoregional failure-free survival (LFFS). We noticed an association between LNR and survival by Kaplan-Meier analysis. Hazard ratio (HR) of LNR was determined by Cox's regression model.
RESULTS: Two hundred and one patients entered this study after their medical histories were evaluated. The mean of lymph node count and LNR was 14.30 (±9.50) and 0.12 (±0.23), respectively. Eighty patients (39.80%) experienced recurrence of SCC. Five-year OS, DFS, and LFFS were 32%, 21%, and 64%, respectively. The median of OS was 40.70 months and 30.11 months in patients with LNR of ≤0.06 and >0.06, respectively (P < 0.01). The LNR >0.06 was found to be a significant prognostic factor for lower OS of patients with HNSCC (HR = 2.11 [1.10, 4.40]; P = 0.04). DFS was not significantly different among patients with LNR ≤0.06 and patients with LNR >0.06 (P = 0.9). However, LFFS was slightly different among two groups (HR = 2.04 [0.90-4.80]; P < 0.1).
CONCLUSION: We recommend more intensive adjuvant therapies such as chemotherapy with radiotherapy and short interval follow-up for patients with LNR >0.06. Further investigations with larger sample sizes are recommended.

INTRODUCTION

Head-and-neck cancers comprise approximately 4% of all cancers in the United States.[1] These cancers which diagnosed over 50 years, are twice common in men rather than women.[2] The most common type of head-and-neck cancer is squamous cell carcinoma (SCC) and they arise more commonly from oral cavity, oropharynx, larynx, and hypopharynx.[3]

Treatment strategies of head-and-neck squamous cell carcinoma (HNSCC) depends on tumor site and stage and patient-specific factors such as comorbidities and preference.[4] Multimodality treatments of HNSCC can include surgery, radiation therapy (RT), chemotherapy, targeted therapy, or a combination of treatments.[5] Patients with localized (Stage I and II) HNSCC are usually managed with either surgery or RT alone. However, patients with more advanced (Stage III and IVa/b) disease generally require various combinations of RT, surgery, and chemotherapy or cetuximab.[6]

Risk stratification is crucial in choosing the appropriate treatment and the need for adjuvant therapy.[7] Currently, tumor, node, and metastasis (TNM) staging system is used for the determination of risk and prognosis of HNSCC. Nodal staging in this system is based on the following parameters: number, site, and size of positive lymph nodes.[89] Recent evidence suggests that besides these factors, the number of positive nodes, lymph node ratio (LNR), and lymph node yield are critical prognostic factors that have not included in the TNM staging system.[10]

Several studies have revealed that LNR has great prognostic value in different cancers such as head-and-neck cancer,[11] bladder cancer,[12] esophageal cancer,[13] and cervical cancer.[1415] LNR is defined as the ratio of the number of affected lymph nodes to the total number of excised lymph nodes.[16] The aim of the present study was to evaluate the prognostic value of LNR in survival of HNSCC patients.

MATERIALS AND METHODS

Patients

All patients diagnosed with pathologically approved aerodigestive SCC and lymph node metastasis in neck dissection according to the pathologic report who were admitted to Tehran Cancer Institute (Tehran University of Medical Sciences, Tehran, Iran) between 2005 and 2011 were the candidates for participating in this historical cohort study. Patients with previous surgical intervention in head-and-neck region, noncomplete preoperative staging, distant metastasis, nonradical tumor resection, margin involvement, inappropriate neck lymph node dissection, and preoperative chemo- and radio-therapy were excluded from the study.

Data collection

All medical records of participants were surveyed; demographic data, histologic grade and stage of tumor, place of involvement, lymph node count, and LNR were collected from the medical records. Moreover, information surrounding the recurrence, site of recurrence, and survival was obtained by contacting the patients and/or their family member(s) and arranging a tele-interview.

Statistical analysis

The data were analyzed using SPSS v. 20 software for Windows (SPSS Inc., Chicago, IL, USA). Quantitative data were presented as mean ± standard deviation (or median as appropriate) while qualitative data as frequency (percentage). Continuous data were compared among groups using independent samples t-test and categorical ones using Chi-square test. Kaplan–Meier method along with log-rank test were used to compare the overall loco-regional failure and disease-free survival (DFS) among two groups of patients with LNR of ≤ 0.06 and > 0.06 for different stages of cancer. Crude and multivariable adjusted hazard ratio (HR) were calculated using Cox proportional hazard regression to evaluate the potential impact of LNR on overall survival (OS), DFS, and loco-regional failure-free survival (LFFS).

We considered cut point of 0.06 for LNR, similar to the study conducted by Gil et al. in 2009. All reported P values were two-sided.

Ethics

Our study was conducted according to the Helsinki Declaration. Furthermore, the study protocol was approved by the Research Ethics Committee of Tehran University of Medical Sciences (IR. Tums.vcr.rec. 1396.2831).

RESULTS

Two hundred and one patients entered this study after their medical histories were evaluated. Eighty-nine patients (44.28%) were female and the remaining individuals were male. Thirty-eight patients (18.91%) were in the age group of <50 years, 74 (36.82%) in 50–70 years, 75 (37.31%) in >70 years, and 14 (6.97%) with unknown age due to the incompetency of patients' records. Sites of involvement were lips in 21 (10.44%), oral cavity in 153 (76.11%), oropharynx in 6 (2.98%), larynx in 5 (2.48%), and other sites in 16 (7. 96%) patients. Forty patients were in Stage 1 (19.90%), 39 (19.40%) in Stage 2, 56 (27.86%) in Stage 3, and 66 (32.84%) in Stage 4a and 4b. Grade of histology of 99 patients (49.25%), 70 (34.83%), 4 (1.99%), 1 (0.50%), and 27 (13.43%) were well differentiated, moderately differentiated, poorly undifferentiated, and unclear, respectively. The mean of lymph node count and LNR was 14.30 (±9.50) and 0.12 (±0.23), respectively [Table 1].

Table 1

Basic and clinical characteristics of study participants

The total recurrence count was 80 (39.80%) in participants of this study, 7 (8.75%), 65 (81.25%) and 8 (10%) of which were local, regional, and distant metastasis, respectively. Five-year OS, DFS, and LFFS were 32%, 21%, and 64%, with a median of 30.11, 69.42, and 22.44 months, respectively [Table 2].

Table 2

Different types of survival rates among all study participants

Table 3 presents the 1-year, 3-year and 5-year OSs in two groups of patients with LNR of ≤0.06 and >0.06. Two groups of patients were significantly different in terms of types of survival rate (P < 0.01). The 5-year OS according to LNR was 44% and 14% with a median of 40.70 and 30.11 in patients with LNR of ≤0.06 and >0.06, respectively.

Table 3

Overall survival based on lymph node ratio

The results of Kaplan–Meier and log-rank test for comparing the OS among patients with LNR ≤0.06 and LNR >0.06 indicated a significant difference (P ≤ 0.001). In addition, a significant difference was observed among patients in different stages of disease (P < 0.05). The observed differences were related to patients in Stage 4 compared to those in Stage 1 (P < 0.001), patients in Stage 4 compared to those in Stages 2 and 3 (P < 0.01), and patients in Stage 1 compared to those in Stages 2 and 3 (P < 0.05). However, no significant difference was found between patients in Stages 2 and 3. [Figures 1 and 2].

Figure 1

Overall survival estimates based on lymph node ratio

Figure 2

Overall survival estimates based on stage

Table 4 presents the crude and multivariable adjusted HR of potential impact of LNR on all studied types of survival. The LNR not only in crude model was a significant prognostic factor (3.20 [2.05, 4.95]) but also after adjustment for the impacts of potential confounding factor for OS of patients with HNSCC (HR = 2.11, 95% confidence interval [CI] (1.10, 4.40)).

Table 4

Crude and adjusted hazard ratio and 95% confidence interval for hazard ratio for the association lymph node ratio with survival of head-and-neck squamous cell carcinoma patients

Although there was a significant relation between LNR, DFS, and LFFS in crude models, the DFS was not significantly different between patients with LNR ≤0.06 and patients with LNR >0.06 in adjusted model (HR = 1.01, 95% CI: 0.6–1.72, P = 0.9), while LFFS was slightly different between two groups (HR = 2.04, 95% CI: 0.90–4.80, P = 0.09).

DISCUSSION

The present study was designed to evaluate the prognostic value of LNR in survival of HNSCC. The findings suggest that the OS among patients with LNR ≤ 0.06 and LNR > 0.06 has a significant difference (P ≤ 0.001). Furthermore, a significant difference was observed in LNR among patients in different stages of disease (P < 0.05).

Due to high mortality and recurrence rate in HNSCC and also intensive treatments including surgery, chemotherapy, and radiotherapy, determining a major prognostic factor for patients seems to be necessary.[1718] Many factors have been considered as prognostic factor for patients with this type of cancer in studies conducted before such as TNM staging, age, gender, place of involvement, lymphovascular invasion, perineural invasion, lymph node involvement, and extracapsular invasion of metastatic lymph node.[1819] In 2015, Chen et al. emphasized the importance of LNR as a prognostic factor in head-and-neck cancer patients that affect OS and LFFS.[11]

In previous studies, the presence of node involvement or count of involved lymph node was considered as a prognostic factor, but because of operator dependency of this factor that seems to be not reliable.[2021] In this way, more dissection resulted in more involved lymph nodes and that makes poor prognosis. To solve this problem, the number of involved lymph nodes to all dissected lymph nodes, named as LNR or density, has been discussed as a prognostic factor in recent studies.[2223] In this provided method, more dissection resulted in more lymph node finding, but because of using the ratio, the operator dependency was excluded.[2425] LNR has also been discussed as a prognostic factor in other cancers such as gastric, breast, and colon cancers.[262728]

Gil et al. in 2009 conducted a study on 386 patients with oral cavity cancer who underwent surgical treatment with or without adjuvant therapy and concluded that LNR (cut point of 0.06) is an independent prognostic factor in survival of patients.[29]

Reinisch et al. in 2014 revealed that LNR (cut point of 0.06) was a significant predictor of OS and recurrence-free survival.[30] Lanzer et al. similarly mentioned that LNR and capsule penetration were independent prognostic factors in HNSCC.[31] In contrast, the study by Roberts et al. indicated that the number of positive nodes is a better prognostic factor than LNR and American Joint Committee on Cancer N staging.[21] In addition, de Ridder et al. concluded that due to changes in specimen processing, the number of positive lymph node is more reliable factor than LNR.[32] However, it was later shown by Feng et al. that the combination of LNR and positive lymph node is superior predictor than traditional TNM staging for the benefits of adjuvant concurrent chemo-radiotherapy.[7]

Yong-Hong et al., who conducted a study on patients with SCC of hypopharynx, represented the LNR as a prognostic factor with cut point of 0.1.[33] However, Feng et al. choose 0.075 as cutoff value, so different cut points for LNR were reported in the studies carried out in this field.[711] We perused the effect of LNR on prognosis of patients with cut point of 0.06 and the LNR was a significant prognostic factor for OS.

According to our study, the LNR is one of the significant prognostic factors for OS of patients, but not for DFS and LFFS in the period of follow-up (albeit the association because of HR >1).

Shrime et al. allocated patients with oral cavity SCC in three groups by survival and count of involved lymph node. In this study, the authors concluded that the LNR had a significant negative effect on survival.[34]

In several studies, the number of positive lymph node or LNR was surveyed for determining its effect on survival of patients and each study reported various cut points.[2129] There were many confounding factors in these studies for determination of cut points, for example, the presence of previous chemo-radiotherapy. Hence, in this study to neutralize the effect of previous chemo-radiotherapy, we excluded all patients with history of chemo-radiotherapy.

Limitation

One of the problems we encountered was the incomplete medical records. From the other limitations, we can mention poor cooperation of some participants for follow-up.

CONCLUSION

We conclude that the LNR is a significant prognostic factor in OS of patients diagnosed with HNSCC. We recommend more intensive adjuvant therapy such as chemotherapy with radiotherapy and short interval follow-up for patients with LNR >0.06.

Financial support and sponsorship

There was no financial support for present study

Conflicts of interest

There are no conflicts of interest.