Intraoperative frozen section as a reliable ancillary technique in salivary gland surgery: A cross sectional study [version 1; referees: 1 approved, 1 not approved]

Salivary glands tumours are uncommon, frequently benign Background. lesions, prevalently located in the parotid gland (80%). Surgical decision making is based on the patient’s history, examination findings, imaging and fine needle aspiration (FNA). FNA is a pre-operative method with good ability in detecting malignancy. During surgery, therefore, Frozen section (FS) can differentiate benign lesions from malignant tumours, to reduce incorrect treatments, to increase the chances of conservative surgery and to better evaluate surgical margins. The aim of our study is to demonstrate the accuracy of the FS procedure in surgery of the salivary glands and to stress the need for dedicated pathology units specialized in lesions of the oral cavity. The study included 499 patients who underwent surgery from May Methods . 2005 and October 2014. An intra-operative frozen section procedure was done for 288 of them. All frozen sections were compared with the final results. The cases were classified by site, nature of the lesion and histotype, according to the WHO classification. Comparison was made between the intra-operative and the definitive diagnosis. Of the 288 FS procedures, 259 were for neoplastic lesions, 199 of Results. neoplastic false types. Of the 29 non-neoplastic FS results, 4 were false negatives. Our results showed that the accuracy of frozen section Conclusions. procedure is 98% for salivary glands tumors. The highest concordance between frozen section and the definitive diagnosis was for inflammatory processes (99%), pleomorphic adenoma (98%), Warthin’s tumor (97%) and malignant neoplasms (96%). In conclusion, based on these findings, frozen section of the salivary glands may be proposed as a routine procedure and should be used in decision-making. This paper is recommending frozen section as a primary evaluation of salivary gland tumors, and it is considered out of date and unethical. Many salivary gland tumors share histologic features and have dintinct expression or molecular characteristics. Ancillary tests are occasionally necessary. Accurate diagnosis before operation, if possible, is needed for proper patient management. Core needle biopsy of the salivary gland tumors is recently substituting aspiration cytology. Authors need to comment the usefulness of such preoperative diagnostic procedures. Three exclusion criteria from selection seem irrelevant with diagnosis of salivary gland tumors. The distribution of tumors appear very biased even after exclusion. Mucoepidermoid carcinoma is the most common malignancy worldwide, while squamous cell carcinoma of salivary gland is rare. It is difficult to be convinced that most of lymphomas, and all small cell carcinomas and carcinoma ex pleomorphic adenomas were accurately diagnosed by frozen section. Illustrations or other back-up data are needed. Interesting paper about the importance of frozen section in the diagnosis of Salivary glands lesions. The casistic is very wide and the results are innovative also if many guidelines report the need of performing Fine Needle Aspiration in the preoperative setting of a salivary gland lesion and I do not think that we can actually derogate from this guideline. Anyway this paper is worthy of indexing but some points need to be reviewed:


Introduction
Salivary gland tumors are rare lesions that occur mainly in the major salivary glands; 80% of tumors occur in the parotid gland, and these are prevalently benign.
The parotid gland histology is complex: there are i) abundant intralobular and extralobular adipose tissues, which increase in relative volume with age, ii) randomly distributed lymphoid aggregates, and iii) lymph nodes that occasionally contain ducts or salivary acini. Therefore, it is often difficult to distinguish a neoplastic lesion from a non-neoplastic lesion, as well as a benign lesion from a malignant lesion, especially in view of the morphological variability of salivary tumors. However, a correct differential diagnosis, safely and promptly executed, is crucial for the entire patient's management, both clinical and surgical, including possible tissue regeneration [1][2][3][4][5] .
The clinical approach to salivary lesions is supported by imaging techniques, such as ultrasound, computed tomography or magnetic resonance imaging; these can provide greater definition of the lesion, but are not always sufficient to formulate a definitive diagnosis. Therefore, it is necessary to resort to pre-surgical techniques to better define a salivary lesion.
Fine needle aspiration (FNA), which can be performed at the time of the initial clinical consultation, can be used both as a diagnostic test and as a guideline in selecting the patient's management: surgical vs follow-up without surgery. The FNA technique demonstrates high sensitivity and specificity (80% and 97%) for benign tumors, but is not very sensitive for malignant neoplasms (sensitivity ranges from 54% to 92%; specificity 87% to 98%). False-negative rates range from 2% to 31% and false positive rates from 0% to 7% 6-11 .
FNA is a simple, safe procedure that does not require the use of local anesthesia, and can be performed either blinded or under ultrasound guidance. It is important to remember that the methodological approach, clinical needle aspiration skills and experience of the pathologist are the elements that affect the definitive diagnosis. Frozen section (FS) is a less rapid, more invasive intraoperative diagnostic procedure, but precisely because of this, it is a guarantee of better histological results. It allows surgery to be performed in a targeted treatment continuum 1,[9][10] .
Once again, it should be emphasized that salivary tumors are a heterogeneous group of lesions that necessitate the examination of many sections. Early diagnosis is essential, to establish the correct histological type of salivary glands lesion, in order to achieve proper planning of the surgical treatment, which may also involve the regional lymph nodes, and adjacent tissues 11 . Variable percentages of effectiveness of the FS method applied to the diagnosis of salivary glands lesions are reported in the literature. The reliability rate ranges from 40% to 100%; this variability is often attributed to the pathologist's experience or to the technician responsible for slide preparation 12-16 .

Objectives
In this study we assessed the diagnostic accuracy of FS; specifically, the concordance between FS and the definitive histological diagnosis, as well as verifying the importance of being able to rely on a team of experts, to reduce false positive or negative cases.

Patients
Between May 2005 to October 2014, 499 patients (275 males and 224 females, mean age 54±17.2 years) suffering from localized masses in the salivary glands were recruited at the Complex Unit of Otolaryngology at the University of Bari (Italy).
Inclusion criteria were related to the clinical aspects of the first access diagnosis: only patients with localized masses developed in the major salivary glands regions were included. Exclusion criteria were a previous history of cancer or any suspicion of infectious disease as main noxa of the mass. Also, patients reporting smoking habits were excluded as well ( Figure 1).

Data collection
We selected 288 salivary lesions (out of the total 499) operated on by the same team of surgeons and pre-analyzed with FS.

Bias
To reduce bias, in 90% of cases, the intraoperative examination was performed by the same team of pathologists. In all cases, the radiological examination posed an indication for surgical treatment and suggested a provisional diagnosis (i.e. benign vs malignant). FNA was not considered because if done at all, it was at non-dedicated centers and there was a high number of "nondiagnostic" results.

Quantitative variables
Our study was aimed to assess if and how many were the neoplastic lesions, how many were benign and/or malignant, and if some non-neoplastic lesions were also diagnosed. Our attention was also directed towards the identification of some false positive/negative results. The accuracy of FS procedure was compared, in all the reported cases, with the traditional histological assay. Our null-hypothesis was aimed to assess that FS of the salivary glands may be proposed as a routine procedure and should be used in the decision-making process.

FS process
The FS procedure, also called cryosection, is a commonly used procedure to perform rapid microscopic analysis of a specimen: it is mostly used as a first-look diagnostic tool in intraoperative oncological surgery. The protocol adopted for performing intraoperative FS was a standardized method: i) the biopsy was fixed and cut in a cryostat at -25°-28°C, ii) the samples were cut into 5 micrometers thick slices, iii) the samples were subjected to haematoxylin-eosin staining. (Figure 2).
The medical report was obtained, in all patient cases, within 10-15 minutes. Each sample was stored with labels detailing the biodata and the macroscopic characteristics of the excised lesion.

Participants
We reported 288 patients with salivary lesions operated on by the same team of surgeons and pre-analyzed with FS, in order  to make a comparative analysis between the two different techniques.
Outcome data FS was useful to indicate the correct surgical treatment of 269 nodules (93.4%) of the parotid gland, 14 nodules (4.9%) of the submandibular gland and 5 lesions (1.7%) that involved the minor salivary glands located in the palatal mucosa. Using the FS method, a correct diagnosis was obtained in 280 cases (97.2%) ( Table 1).

Main results
The highest concordance between the FS and the definitive diagnosis was for pleomorphic adenoma (98%), Warthin's tumor (97%), inflammatory processes (99%) and malignant neoplasms (96%). In 8 cases (2.8%), the "provisional diagnosis" with intraoperative FS and the definitive diagnosis were discordant. The false negative results for FS consisted of the following: 1 inflammatory process, actually diagnosed as Warthin's tumor; 1 normal tissue, actually diagnosed as an arteriovenous malformation; 1 reactive lymphoid tissue, actually diagnosed as a non-Hodgkin's lymphoma; and 1 sialometaplasia, actually diagnosed as a squamous cell carcinoma. Two false positive results were obtained: in both these cases squamous metaplasia in pleomorphic adenoma was interpreted as malignant (Table 2). Thus, the sensitivity and specificity for malignancy were assessed to 97%. Patients with a positive FS diagnosis underwent lymphadenectomy. The FS diagnosis of malignancy was not confirmed at histology in two cases: 1 carcinoma, actually a NH-Lymphoma; 1 NH-Lymphoma, actually a small cell carcinoma. In these cases, only one patient received overtreatment.

Other analyses
Analyzing the definitive diagnosis vs FS, the false positive cases were found to have been diagnosed by a non-dedicated pathologist, as also one case of Warthin's cancer, defined as an inflammatory process.

Discussion
This study was conducted in order to assess the diagnostic accuracy, sensitivity and specificity of FS examination, used as a preoperative method to guide the surgeon in the choice of treatment of salivary lesions. A further aim of the study was to analyse the diagnostic accuracy of FS examination to pose the diagnosis of malignancy. Previous studies have compared the accuracy of FS examination with FNA cytology, observing a greater reliability of the FS technique 13 . In fact, FNA shows too high a number of false negatives due to factors related to: i) triage errors, ii) hypocellularity of the material, iii) interpretation errors 10,11 . Some studies have demonstrated an excellent effectiveness of a FS examination, obtaining a specificity of 99% and a sensitivity of 90%, other authors have even reported maximum specificity and sensitivity, equal to 100% 16,17 .

Key results
In our study, we excluded FNA as a perioperative examination because the number of "inadequate" tissues was too high. In our experience, FS could replace FNA, reducing the risk of inappropriate surgery, of unnecessary adjuvant radiotherapy, as well as reducing National Health System costs 18 .
Based on the results described in this study, the FS examination shows a high reliability in the diagnosis of salivary gland tumors. FS is particularly useful in cases of differential diagnosis between neoplastic and non-neoplastic lesions; it also shows a good reliability in the differential diagnosis between benign and malignant neoplasms.

Limitations
Despite our study limitations, related to the small sample size and related to the single unique center participating to this study, our results are in agreement to those reported by various authors, and they also highlight the need to be able to rely on a dedicated team of clinicians and pathologists 15-20 .

Conclusions
"Misinterpretation" was observed in those cases diagnosed by pathologists with no specific experience of head and neck tumors. Our work group experience emphasizes the importance of intraoperative examination in order to define the histotype and the cancer margins, permitting the performance of effective, predictable surgery of the salivary glands.

Ethical statement
In Italy, directors of clinical research units and those responsible for clinical services may access data records for research purposes if patients have previously signed a consent form that confirming that they allow this use of their data. Therefore, no specific ethical approval is required for this study, and all patients included in the study signed written informed consent allowing their data to be used in future research.

Competing interests
No competing interests were declared.

Grant information
The author(s) declared that no grants were involved in supporting this work. This paper is recommending frozen section as a primary evaluation of salivary gland tumors, and it is considered out of date and unethical. Many salivary gland tumors share histologic features and have dintinct expression or molecular characteristics. Ancillary tests are occasionally necessary. Accurate diagnosis before operation, if possible, is needed for proper patient management.

Open Peer Review
Core needle biopsy of the salivary gland tumors is recently substituting aspiration cytology. Authors need to comment the usefulness of such preoperative diagnostic procedures.
Three exclusion criteria from selection seem irrelevant with diagnosis of salivary gland tumors.
The distribution of tumors appear very biased even after exclusion. Mucoepidermoid carcinoma is the most common malignancy worldwide, while squamous cell carcinoma of salivary gland is rare.
It is difficult to be convinced that most of lymphomas, and all small cell carcinomas and carcinoma ex pleomorphic adenomas were accurately diagnosed by frozen section. Illustrations or other back-up data are needed.

Is the background of the cases' history and progression described in sufficient detail? Partly
Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly Is sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No

Is the conclusion balanced and justified on the basis of the findings? Partly
No competing interests were disclosed. Interesting paper about the importance of frozen section in the diagnosis of Salivary glands lesions. The casistic is very wide and the results are innovative also if many guidelines report the need of performing Fine Needle Aspiration in the preoperative setting of a salivary gland lesion and I do not think that we can actually derogate from this guideline. Anyway this paper is worthy of indexing but some points need to be reviewed: The first sentence of the "FS process" in method section should be moved to the introduction to complete the paragraph about the frozen section.
In the inclusion criteria the authors report that just major salivary glands lesions were analyzed but in the outcomes data they state that 5 cases were minor salivary glands glands in the palatal mucosa. Please clarify this point.
In the discussion the authors state that the FS is used as "pre-operative method". I think that is more correct to define it an intraoperative method.

Is the background of the cases' history and progression described in sufficient detail? Yes
Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes Is sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes

Is the conclusion balanced and justified on the basis of the findings? Yes
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.