Background
Oral cancer is one of the most common malignancies, ranking the top 10 of all malignant tumors, with more than 450,000 new cases worldwide each year, accounting for 2.4% of all cancer deaths. More than 90% of oral cancers are oral squamous cell carcinoma (OSCC)[
1]. The mainstays of OSCC treatment are surgery, chemotherapy, radiotherapy or a combination of these modalities. Unfortunately, despite the various efforts that have been made for OSCC [
2,
3], the survival rate of OSCC has not remarkably increased in the last three decades, which remains below 60%[
4].Recurrence (especially early recurrence) is one of the major causes of low survival rates of OSCC after definitive therapy. It has been reported that the cure rates of salvage surgery for patients with recurrence is 15%~67%, suggesting the role for surgical treatment after recurrence with both palliative and curative intent [
5‐
9]. Therefore, regular and close follow-up examinations are particularly important for early diagnosis of recurrence and improvement of survival of OSCC patients [
10]. At present, “one-size-fits all” follow up program cannot meet the clinical needs. First, this regimen may not be appropriate for some patients who may be at higher risk for a second event than others [
11‐
13]. In addition, personalized follow-up schedule after primary tumor treatment has never been investigated [
13]. While longer follow-up time and shorter follow-up interval may help us diagnose relapse early, they may also place an undue burden on patients, especially those in remote areas or with poor economic conditions.
Many literatures stated that the recurrence rate of OSCC varies from 7–47.4%[
7,
14‐
18]. Despite the enormous impact of recurrence (local, loco-regional and regional) on the prognosis of OSCC, limited understanding is currently available about its frequency, exact location and development pattern [
19‐
22]. Many clinicopathological factors and molecular markers are associated with recurrence of OSCC, including smoking, advanced clinical stage, poorly differentiated tumors, radiotherapy, microvascular invasion, and high positive lymph node ratio, etc. [
7,
17,
19,
23‐
25]. However, the prognostic value of these clinicopathological factors is always uncertain and controversial [
19,
25‐
27]. It is also worth noting that it’s not clear which clinicopathological factors and molecular biomarkers could identify the patients at high risk of early recurrence.
Materials and methods
Data collection methods
The patients’ data was collected and sorted from our hospital’s electronic database, which contains prior medical records and postoperative follow-up information.
Study design and clinical samples
The study was approved by the Ethics Committee of Second Xiangya Hospital of Central South University. A total of 1560 consecutive OSCC patients underwent major surgery with primary reconstruction at our Department of Oral and Maxillofacial Surgery from January 2010 to December 2016. All patients underwent complete surgical resection with curative intent. For OSCC recurrence, 5 years is a cut-off point. If OSCC does not relapse in 5 years, the chance of recurrence will be greatly reduced. Therefore, the 5-year survival rate is always used to indicate the effect of OSCC treatment. Inclusion criteria were defined: (1) the patients were histologically confirmed with primary OSCC, (2) more than 5 years follow-up after surgical treatment without a loss or until death, (3) primary tumor without distant metastasis. Patients who had (1) a history of radiation or chemotherapy for oncological diseases of other origins and (2) inadequate clinical follow-up information were excluded.
Recurrence interval was defined as the duration from the end of initial treatment to the confirmation of recurrence by pathological examination after incisional biopsy. For patients with multiple tumor recurrences, only the time interval from the completion of the first treatment to the first recurrence was calculated. The recurrence interval is significantly associated with the survival rate of OSCC patients, with an interval < 18 months appearing to be a more dismal cut-off point for survival according to some literature reports [
5,
9]. In this study, we used < 18 months as the cut-off point for defining early recurrence (ER).
To distinguish between recurrent tumors and new primary tumors, only patients with OSCC with the same or higher histologic grade as the primary tumor were included [
5]. In this study, local recurrence, loco-regional recurrence and regional recurrence were all defined as recurrence. Recurrence arising only in the primary site was defined as local recurrence. Recurrence arising only in the neck was defined as regional recurrence. Recurrence arising in both the primary site and neck was loco-regional recurrence [
28].
TCGA database
TCGA (
https://tcga-data.nci.nih.gov/tcga/) is a publicly available dataset. We screened 347 OSCC patients from the head and neck squamous cell carcinoma (HNSCC) dataset in TCGA and downloaded the clinical data for this study. Among them, 64 cases had recurrence.
Treatment protocol and follow-up examinations
Treatment regimens for each patient in our department include resection of OSCC within the appropriate safe margin and the use of neck dissection recommended by Chinese guidelines [
10,
29], which are mainly based on NCCN and ASCO, with no difference in treatment methods. Postoperative radiation was advised for cases with histologic evidence of lymph node metastasis, extracapsular extension of cervical lymph node metastasis and extensive local tumor burden. Typically, patients receive a follow-up examination every 1 month during the first year after completion of treatment, and a routine PET-CT is performed at the end of the first year. Follow-up examinations will be performed every 3 months in the second year, and then every 6 months in the third to fifth years. In case of clinical suspicion of local, loco-regional or regional recurrence, further re-staging procedures (CT, positron-emission CT, ultrasound, magnet-resonance imaging) may be performed as appropriate. Salvage surgery was considered the therapy for recurrent OSCC. No recurrence after 5 years of follow-up was considered to be the end of follow-up.
Demographics and clinical data
Data obtained from medical records included demographic features, tumor sites, histopathology, grade, stage, total number of resected lymph nodes, total number of positive lymph nodes, total number of extracapsular infiltrations of lymph nodes, type of surgical procedure, TNM stage, type of salvage treatment, outcome, symptoms of patients with recurrence, time of symptom appearance, recurrence interval, diagnosis method of recurrence and location of recurrence, etc.
For comparison, all patients with recurrence were divided into 2 groups: early recurrence (ER) (n = 219) group and late recurrence (LR) (n = 15) group.
Statistical analysis
Statistical analysis was performed using SPSS 25.0 software (SPSS Inc., USA), and R, version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria). The clinical prognostic risk factors associated with early recurrence were screened out by Log-rank test, and then these screened variables were further incorporated into the Multivariate Cox Regressive Analysis model for further analysis. The Kaplan–Meier method was used to plot survival curves for each recurrence interval. P value < 0.05 was considered statistically significant.
Discussion
Tumor recurrence is associated with deterioration in patients outcomes and is therefore one of the major issues for OSCC treatment [
30]. Survival rates of early relapse were reported to be much lower than late relapse [
31]. Meanwhile, some literatures showed that the survival rate of patients with OSCC recurrence is related to the site of recurrence [
21,
23,
25,
32,
33]. However, current medical literatures contain few studies on temporal and spatial patterns of OSCC recurrence.
Due to the global imbalance of medical care and the continuous improvement of OSCC diagnosis and treatment, previous scholars’ reports on RR were not consistent, which was ranging from 7–47.4%[
5,
7,
14‐
18]. In this study, the 5-year follow-up data of 234 OSCC patients form our department and 64 OSCC patients from TCGA database were analyzed, the results indicated that the 5-year recurrence rates of OSCC patients in clinical and TCGA were 15% (234/1560) and 18.44% (64/347), respectively. The research of Rogers SN [
34] and Vázquez-Mahía I [
35] have shown that up to 86% of patients with OSCC relapsed within 2 years after the first treatment, and that those who relapsed early had a worse prognosis than those who relapsed afterwards [
18,
34‐
38]. In our hospital, 93.6% of the recurrences occurred within 18 months (Fig. 1A), which was significantly higher than 84.4% in TCGA database (Fig. 1B) or 60% reported by Annelies Weckxa [
18]. We speculated that this difference may be related to the betel nut chewing habits of East Asian. In this regard, we performed Log-rank test on the clinicopathological factors of early and late recurrence of OSCC patients and found that chewing betel nut was associated with the timing of recurrence (Table
2). After further subdividing the recurrence interval, we found that the recurrent patients at 3, 6, 12 and 18 months after surgery accounted for 20.1%, 55.1%, 85.9%, and 93.6%, respectively. Only 6.4% of patients relapsed within 18 ~ 60 months after surgery (Fig. 1A). Similar results were obtained from TCGA database. 79.7% of relapses occurred within 12 months after the first treatment, and 84.4% of relapsed patients relapsed early (Fig. 1B). Therefore, given that nearly 90% of recurrence occurred within 18 months after surgery (approximately 80% of which occurred within 1 year after surgery), and that close follow-up and early diagnosis could improve the success rate of rescue for relapsed patients.
We believe that regular examinations should be performed within 24 months after surgery, while continuing follow up medical history and physical examinations for about 5 years based on individual risk assessment. In our hospital, it is recommended to pay regular examination every month for 1 year after the operation and keep checkups every 3 months from 1 to 2 years after the operation. Superficial recurrence can be diagnosed by pathological biopsy, while deep recurrence can only be diagnosed based on the patient’s symptoms (such as local pain, etc.), signs and imaging examinations [
18,
39]. Through this study, we found that tongue cancer is more prone to regional metastasis, while buccal cancer is more likely to recur locally. At the same time, age, betel nut chewing, previous radiotherapy history, lymph node metastasis and postoperative infection are risk factors for early recurrence of OSCC. Combining with the studies of previous scholars [
40‐
42], we believe that MRI/CT/PET-CT examination, especially for patients with deep neck pain, MRI/CT examination or even PET-CT should be performed immediately to confirm the diagnosis. Considering that buccal cancer tends to recur locally [
39,
42], we believe that for buccal cancer patients, the definite diagnosis can be confirmed by combining the patient’s signs, visual inspection, palpation and pathological biopsy. Additionally, NCCN guidelines suggest that there may be little proven benefit in further imaging if the initial 3-month FDG-PET/CT scan is negative. If FDG-PET/CT at 3 months post-surgery is negative, then there are no data to support a substantial benefit for further imaging in an asymptomatic patient [
43]. However, similar to the study results of Chonji Fukumoto [
40], considering that the probability of recurrence in patients still exists after 18 months after surgery, although the proportion is already very low, we routinely recommend patients undergo PET-CT examination at 12 months after surgery to rule out early microscopic lesions.
Further, we analyzed the clinicopathological factors related to early and late recurrence. Abundant evidence exists for association between the occurrence interval and grade of primary tumor [
18,
19,
25,
37], lymph node metastasis [
18,
21,
23] and previous radiotherapy [
44,
45]. Tumor pathologic features such as DOI (depth of invasion), PNI(perineural invasion), LVI (lymphovascular invasion), and ENE (extranodal extension) ,etc. have been shown to be associated with lymph node metastasis in OSCC [
46]. In this study, we have not done an in-depth analysis of these factors, which would require more data and a more detailed analysis. Previously clinical studies have shown that young age [
47,
48], postoperative infection [
49,
50] are related to the recurrence in many tumors. Similarly, chewing betel nut has been shown to be related to the progression of OSCC in multiple studies [
51,
52]. In this study, a significant correlation was found by Log-rank test between the recurrence interval and age (P = 0.045), chewing betel nut (P = 0.048), previous radiotherapy(P = 0.038), lymph node metastasis (P = 0.032), and postoperative infection (P = 0.043) (Table
2). The study-is to our best knowledge-the first to show a significant association between the recurrence interval and age, chewing betel nut and postoperative infection.
In addition to the recurrence interval, some other clinical factors are found to be associated with the recurrence site. As described in previously studies, OSCC patients with primary lymph node metastases to the neck have an increased recurrence rate [
28,
53‐
55]. Late pN status, extracapsular spread, perineural infiltration, vascular/lymphatic embolism, diffuse infiltration and neck dissection were found to be associated with regional (neck) recurrence which was an important prognostic factor for overall survival (OS)[
56]. While poor differentiation, location (hard palate and retromolar trigone), bone invasion, lymphatic invasion, surgical margins and invasion depths were significantly associated with local-regional recurrence [
57]. All of these risk factors lead to poorer tumor control.
Although there have been numerous studies on the incidence of local, regional, and local-regional recurrence in patients with OSCC, fewer have been reported on the proportion of these three in the same group of relapsed patients or in different OSCC subtypes. Studies by Yasmine Ghantous [
58] and Thomas Mücke [
59] found that the probability of local recurrence in OSCC was about 64.9%[
58,
59], which was more common than loco-regional recurrence, and the ration was 2.58:1(64.9%:25.1%) [
59]. While the study of Troeltzsch showed that regional recurrence is usually combined with local recurrence within 2 years of the initial diagnosis of OSCC [
60].Through our study, we found that among all the OSCC patients with postoperative recurrence in clinical (234 patients) and TCGA (64 patients), the ratios of local recurrence to loco-regional recurrence were about 3.89:1 (105:27) and 1.67:1(25:15), respectively. These results were consistent with the study of Thomas Mücke [
59]. Further, we found that the ratio of local recurrence to regional recurrence was about 1.03:1 (105:102) in clinical, which was similar to the ratio of 1.04:1(25:24) in TCGA database. Furthermore, based on the location of the primary tumor, we analyzed the spatial pattern of recurrence of OSCC in different locations. Through cross tabulation with chi-squared testing, we were surprised to find that buccal cancer was more prone to local recurrence (Table
4), while tongue cancer was more prone to regional recurrence (Table
5). The reasons may be due to the local anatomy and the location of lymphatic drainage [
10,
61]. Unlike tongue cancer, buccal cancer has been shown to develop in an orderly progression, skip metastasis was rare [
62], however, its specific biological mechanism is still unclear and needs further study. It is worth noting that our results of clinical samples showed that buccal cancer was more prone to regional recurrence (P = 0.035) but results of TCGA database showed no statistical difference in the probability of regional recurrence between the two groups (P > 0.05) (Table
6). Differences in ethnicity, number of cases, and dietary habits (for example, most Chinese OSCC patients chew betel nut while Americans do not) could contribute to the above difference.
More and more studies have shown that the recurrence interval and recurrence location were related to the prognosis of OSCC patients [
5,
9,
18,
20,
37]. However, there are currently few studies on the relationship between the recurrence interval and the recurrence site. In this study, we found that the earlier the recurrence, the more likely it is to be loco-regional recurrence (Fig. 3A and B). Further, a survival analysis of patients who underwent secondary surgery after recurrence showed that the earlier the recurrence, the lower the survival rate (Fig. 3C, 3D). These results indicated that patients with local-regional recurrence have a wide range of second tumors, rapid tumor progression, and poor prognosis. The cause has nothing to do with the complete resection of the primary tumor, may be related to tumor heterogeneity [
63,
64]. Considering the postoperative quality of life and survival rate, we should be cautious about reoperation for such patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.