Effect of adjuvant radiotherapy on the local recurrence of oral squamous cell carcinoma with perineural invasion: A systematic review

Objectives of the Review The decision whether to include postoperative radiotherapy on patients with oral squamous cell carcinoma depends on the risk of local recurrence. The objectives of this study were to systematically review literature on whether perineural invasion in oral squamous cell carcinoma patients is associated with higher local recurrence rates and whether local recurrence is influenced by the administration of postoperative radiotherapy in patients presenting with perineural invasion. Type of Review Systematic review. Search Strategy Embase, PubMed, Web Of Science. Evaluation Method The databases above were searched for studies that analysed: the treatment of oral squamous cell carcinoma patients with perineural invasion, local recurrence and postoperative radiotherapy. The data of seven studies were analysed qualitatively. Results The overall quality of the studies was moderate to low. There was no evidence of the effect of postoperative radiotherapy on local recurrence rates in patients presenting with perineural invasion. Some evidence suggests that local recurrence rates may increase in cases of multifocal perineural invasion, especially if nerves >1 mm are involved but these data should be interpreted with caution due to the low‐quality evidence. Conclusions High‐quality evidence regarding the prognostic value of perineural invasion and the impact of postoperative radiotherapy in patients presenting with perineural invasion is lacking in the literature, making it difficult to select a postoperative strategy for early‐stage tumours.

margins (1-5 mm), poor differentiation, pT3-4, lymphangio invasion and perineural invasion (PNI), as to when PORT should be applied. 5 There is a need for research regarding adjuvant radiotherapy in intermediate risk cases, which was also pointed out by Blackburn et al. 6 The risk of LR in the presence of PNI is unclear. 2,7-10 Some studies found a significant increase in LR rates in PNI cases, [10][11][12] while other studies did not. 13,14 Moreover, it is not clear whether PORT has an additional, positive effect in reducing the local recurrence rate in early OSCC with PNI. Evidence regarding prognosis is needed to justify the role of PORT, because of its side effects such as xerostomia, dysphagia, loss of taste, trismus and osteoradionecrosis. [15][16][17] The objectives of this study were to systematically review the literature to find whether PNI in OSCC patients is associated with a higher LR rate and whether LR rate is influenced by administration of PORT in OSCC patients presenting with PNI.

| Study identification and selection
The study protocol was designed using the PRISMA statement for reporting systematic reviews and meta-analyses. 18 A search protocol was developed prior to the study. Studies were sought in electronic databases namely, PubMed, Embase and Web of Science. The last search date was 18 September 2017. No limitations were applied regarding time of study or study design. Only studies written in English were included. A general search strategy was developed together with an information specialist and adapted appropriately to each database (Appendix 1). Publications were included if they described: the treatment of patients with squamous cell carcinoma of the oral cavity; PNI; local recurrence; whether PORT was given or not; and a sample size of ≥10. Local recurrence was defined as histopathologically proven tumour arising within 10 mm from the primary tumour.
Publications were excluded if brachytherapy was given, if preoperative radiotherapy or other earlier treatments were provided, if recurrent or secondary tumours were included, if pathological T-stage was not reported or if the intent of treatment was not curative.
First, titles were assessed for selection by two observers independently (JV and KS). Abstracts from the included titles were then assessed by the same observers. Titles with insufficient information or causing disagreement between the observers were also included for abstract assessment. If an abstract provided insufficient information or disagreement existed between observers, the text was checked. Thereafter, full-text papers were assessed in a similar way.
Finally, the references of the included studies were also perused for inclusion and if any were selected, the same procedure was followed as described above. Studies in which only a part of the study group met the inclusion criteria were included for further analysis of the relevant group. Interobserver agreement was expressed as Cohen's κ and as a percentage of agreement.
In cases of disagreement about inclusion or exclusion, a decision was made by consensus. A third reviewer (MJHW) was consulted to resolve remaining disagreements.
between the studies. The most used definition of PNI was the presence of tumour cells in any of the three layers of the nerve sheath and/or tumour cells in close proximity to the nerve involving more than one-third of its circumference. 1,8,20 In another study, the  8,[20][21][22]25 One study excluded patients with adverse histopathologic factors such as extracapsular spread and lymphangio invasion. 1 In five studies, patients were treated with PORT, but no specific indication regarding PNI was reported. 1,8,20,21,23 Of those studies, four reported that PORT 1,8,20,23 was applied to 3.6%-43.4% of the patients. One study did not report which patients were treated with PORT. 21 Two other studies excluded the patients treated with PORT. 22,25 All the studies analysed different variables, but all had local recurrence as either a primary or secondary outcome.

| Local recurrence
None of the seven studies reported PNI as a significant prognostic factor for LR; however, one study reported PNI as a significant prognostic factor if it was presenting multifocally, especially if nerves >1 mm were involved (P = 0.049). 8 None of the seven included studies evaluated the impact of PORT in patients presenting with PNI. A summary of study findings can be found in Table 2.

| Summary of main results
We used a systematic review to investigate the impact of PNI on LR in OSCC patients and whether LR is influenced by the administration of PORT. Only seven studies could be included, and these papers only partly answered the research questions; therefore, high-quality evidence regarding the impact of PNI on LR rates in patients with OSCC is lacking. All studies reported that there was no significant difference in LR between patients presenting with and without PNI. However, one study reported that PNI was a significant prognostic factor when it is multifocal, especially if nerves >1 mm are involved. None of the included studies reported the impact of PORT on LR in patients presenting with PNI.

| Comparison with other reviews
An earlier systematic review reported that PNI is not a significant prognostic factor for locoregional recurrence. 26 That systematic review included studies describing squamous cell carcinoma in the complete head and neck area and studies only reporting clinical Tstage were not excluded. Also, local and regional recurrences were listed as locoregional recurrence irrespective of whether these types of recurrences had other aetiologies. Our aim was to evaluate the impact of PNI on LR more precisely by excluding studies only reporting clinical T-stage and those only including OSCC. A partially retrospective and partially prospective study reported PNI as an independent predictor of LR if nerves >1 mm were involved 10 ; however, this study was excluded in our study because pathological T-stage was not described and squamous cell There were also a number of patients in which PNI was not reported. c PNI and lymphovascular invasion were combined as a high risk group in the calculations of local recurrence.

| Implications for research
In order to determine the indication for PORT, prospective studies need to be performed to investigate the effect of PORT in patients presenting with PNI on LR. It would be important to introduce a standardised definition of PNI to obtain exact incidence rates. Also, the extent of PNI should be described more specifically by means of the location and the size of the involved nerves. Finally, accurate data registration including precise description of the cohort, pathological T-and N-stages, resection margins, local recurrence and other pathological tumour characteristics (such as depth of invasion, pattern of growth and lymphovascular invasion) would be essential in order to evaluate prognosis.

| CONCLUSION
Based on the available evidence, it is not clear whether there is an indication for PORT in unifocal PNI; moreover, high-quality evidence is lacking on the impact of PORT in OSCC patients presenting with PNI.

CONFLI CT OF INTEREST
There are no conflicts of interest to declare.

CASE CONTROL S TUD IES
Note: A study can be awarded a maximum of one star for each numbered item within the Selection and Exposure categories. A maximum of two stars can be given for Comparability.