NBI for Early Diagnosis of OPMD/OSCC

Introduction Oral squamous cell carcinoma (OSCC) is a malignant condition arising from the mucosal lining of the mouth. It is the most frequent malignancy that affects the oral cavity, with more than 350,000 new incident cases estimated worldwide yearly, causing over 150,000 deaths in 2020.

The 5-year overall survival rate of OSCC - around 60%- has not significantly improved during the last decades, despite general clinical and therapeutic advances. Considering that patients with initial stages (I-II) of OSCC show survival rates of 80-90%, while advanced-stages (III-IV) survival stands lower than 50%, it is apparent that the overall survival low value reflects that most cases are diagnosed at advanced stages. Early detection - possibly at its pre-malignant stages (Oral Potentially Malignant Disorders, OPMD) - with periodic surveillance is thus fundamental for limiting disease burden, hopefully reducing the incidence of advanced stages OSCC and increasing survival.

Narrow Band Imaging (NBI) is a non-invasive imaging fiberoptic technique which allows the analysis of the thin sub-epithelial vascularisation through visual magnification. There is high scientific evidence that NBI fibroscopy facilitate early diagnosis of squamous cancers of the upper aerodigestive tract. In the oral cavity, it has proved itself as promising tool for helping clinician both for diagnosis and therapy.

Yet, there is no definitive scientific evidence that NBI is superior to common oral examination with white light (WLOE) for diagnosing OSCC/OPMD.

Materials \& Methods The primary outcome of this study is a comparison of the detection rates of OSCC or an OPMD using oral examination with white light (WLOE) and Narrow Band Imaging (NBI).

The secondary outcome is to estimate the overall diagnostic accuracy (sensitivity, specificity, NPV, PPV) of the two methods, according to the presenting condition:

• patients with any oral mucosal lesions necessitating an initial diagnosis;
• patients with known OPMDs or had their OPMDs excised.
• patients with history of Head \& Neck (HN)SCC
• high risk patients (heavy smokers, drinkers, individuals with Fanconi anaemia, dyskeratosis congenita, xeroderma pigmentosum, Li-Fraumeni syndrome, Bloom syndrome, ataxia-telangiectasia, and Cowden syndrome) with no oral lesion/disease at the time of examination.

To evaluate diagnostic accuracy, we propose to use the histologic diagnosis from a biopsy specimen as the gold standard diagnosis.

Study Design This clinical trial, once approved by Ethical Committee, will be registered on ClinicalTrial.gov online platform. Patients will be randomly assigned to receive primary WLOE or primary NBI. To improve the quality of the reporting in the diagnostic accuracy study, we complied with the Standards for Reporting of Diagnostic Accuracy (STARD) initiative. We set WLOE as reference standard and NBI as index test. Random assignment will be performed for each case by an investigator using National Insitute of Health (NIH) - National Cancer Institute Clinical Trial Randomization Tool (https://ctrandomization.cancer.gov/tool). This Web site is available only to the study participants. Using a minimization algorithm, the selection of the primary examination is balanced with respect to five stratification variables: institution, age, sex, alcohol consumption, and smoking habit. In order to give patients the highest possible standard of care, we will perform both imaging methods in a back-to-back fashion so that primary WLOE is followed by NBI and primary NBI is followed by WLOE. To avoid any bias, the report of the first examination is completed before the second imaging is started.

Study population and study design are summarised in the flowchart. Study Populations

The protocol and consent form for this study has been approved by the Ethical committee of the A.O.R.N. "ANTONIO CARDARELLI" Hospital (n°06/24), Naples, Italy; written informed consent is obtained from all patients. The inclusion criteria are:

• Patients with soft tissue, mucosal lesions of the oral cavity, who arrive for an initial first diagnosis (group 1);
• Patients with history OPMD (clinically evident lesion/s group 2a, lesion/s excised group 2b), who are in follow-up (OPMDs included are eg. Leukoplakia, erythroleukoplakia, erythroplakia, oral lichen planus, oral lichenoid lesion);
• Patients with history of HNSCC (OSCC group 3a, sinonasal, nasopharynx, oropharynx, larynx, oesophagus group 3b);
• High risk (see above) individuals with no known oral mucosal disease (group 4);

Exclusion criteria will be:

• patients who did not need a biopsy nor after WLOE (e.g. normal mucosa, anatomical variation) nor after NBI (pattern I)
• patients who, despite indication, were not suitable to undergo biopsy given his / her systemic conditions The pathologists will be blinded to the endoscopic information. In case biopsy showed OSCC, patients would be directed for subsequent diagnostic/therapeutic pathway. OPMD cases will be managed in the unit under routine care.

Calculation of the Sample Size For the purposes of this study, we set the probability for error (alfa) to .05 with a power of 0.80 (reflecting a beta error of .2). We estimated that the NBI system would increase the detection yield for superficial cancer by at least threefold compared with conventional WLOE. This resulted in a calculated sample size of 125 patients per category WLOE/NBI, rounded up in 60 (30 WLOE + 30 NBI) patients per group 1/2/3/4.

Endoscopic Examination All NBI observations will include the whole oral cavity including mucosal aspect of lips. NBI system consist in a flexible fiberscope, producing magnified images to a fullHD monitor. Angle of the fiberscope, emitting light (WL, NBI) and recording is controlled throug a joystick at fiberscope's grip.

NBI imaging for Intrapapillary capillary loops (IPCL) patterns of each lesion will be determined according to modifications made by Farah of the system proposed by Takano, as it proved to be the most effective system for oral lesions:

• type 0 (IPCL not detectable)
• type I (physiological arborisation of IPCL)
• type II (meandering or dilated IPCL)
• type III (convoluted/winding and/or elongated IPCL)
• type IV (complete loss of organisation/annihilation of IPCL). The biopsy sample will be taken from the area of highest NBI pattern detected during NBI examination.

To maintain the quality of the NBI inspection and to reduce risk of operator-dependent bias, before the study is started, all the participating operators will be trained by an expert. Each NBI fibroscopy is recorded and reviewed by an expert (AG), blinded to the result of an eventual biopsy and to the other evaluation result. Expert review is then sent back to the initial operator. Differences in determination of IPCL pattern will be resolved through discussion, obtaining a consensus pattern. In case the consensus pattern showed necessity for biopsy (pattern III, IV), if not already performed, the patient is re-called to undergo biopsy.

Pathologic Evaluation Biopsy specimens are taken from each lesion after the completion of both types of imaging and then reviewed by an experienced pathologist according to the WHO Blue book (2024) classification .

Statistical Analysis The absolute and relative frequencies for qualitative variables were calculated for each group. Statistical analysis was performed using SPSS version 17 software (SPSS, Chicago, IL). The continuous variables are expressed as medians and ranges. Continuous data were compared using the MannWhitney U test. Pearson's 2 test or Fisher's exact test was used to analyze categoric data to compare proportions. All P values were two-tailed, and a P value of .05 was considered significant. Concordance between operators' NBI pattern was evaluated with Fleiss' fixed-marginal Kappa test (with Gwet's variance formula, 95% CI), considering values ≤0 as indicating no agreement, 0.01-0.20 as none to slight, 0.21-0.40 as fair, 0.41-0.60 as moderate, 0.61-0.80 as substantial and 0.81-1.00 as almost perfect agreement.