Line-Field Confocal Optical Coherence Tomography as a Follow-Up Tool for 3D-Printed HDR Surface Brachytherapy of Cutaneous Squamous Cell Carcinoma: A Proof-of-Concept Study at Short-Term (Six-Month) Follow-Up

Background/Objectives: Cutaneous squamous cell carcinoma (cSCC) treated with high-dose-rate (HDR) surface brachytherapy using patient-specific 3D-printed applicators lacks a validated non-invasive follow-up modality capable of resolving microscopic residual disease. Line-field confocal optical coherence tomography (LC-OCT) provides real-time, cellular resolution, three-dimensional imaging of the epidermis and superficial dermis and has been proposed as a candidate tool for post-radiotherapy monitoring. We evaluated the feasibility of LC-OCT for pre- and post-treatment imaging of cSCC and assessed whether an LC-OCT-defined remission phenotype, operationally defined a priori as the absence of all malignancy-associated LC-OCT criteria, can be documented after 3D-printed HDR surface brachytherapy. Methods: Eight consecutive patients with histologically confirmed cSCC, unsuitable for surgery, were treated with HDR surface brachytherapy delivered through patient-specific 3D-printed applicators. LC-OCT imaging was performed immediately before brachytherapy and at short-term follow-up (six months). Nineteen pre-treatment LC-OCT criteria derived from the Cinotti 2021 descriptive criterion set for cSCC were scored: seventeen malignancy-associated criteria, plus elastosis (scored as a background photodamage covariate, not counted towards remission) and fibrosis (scored post-treatment only). Presence/absence of each criterion was rated by two blinded dermatologists; inter-assessor agreement was quantified using percentage agreement and Cohen’s kappa. Results: Baseline LC-OCT showed a rich malignant signature: hyperkeratosis, parakeratosis, disarranged epidermal architecture, dyskeratotic keratinocytes, atypical and crowded nuclei, and dermoepidermal junction alterations were present in eight out of eight lesions (100%) under an inclusive two-reader rule. At short-term follow-up, six months after the final brachytherapy fraction, all seventeen malignancy-associated criteria constituting the pre-specified remission set were absent in all eight patients under the decision rule described in the article—which classifies isolated dilated linear vessels occurring in an otherwise criterion-negative, fibrotic dermis as post-radiation telangiectasia rather than residual tumour, yielding an LC-OCT-defined remission phenotype rate of 100% (eight out of eight; 95% exact binomial [Clopper–Pearson] CI 63.1–100%). Post-radiation fibrosis was observed in 8/8 patients (100%), and dilated linear vessels were observed in six out of eight patients (75%, evaluated under the same decision rule (Observer 1: 4/8; Observer 2: 6/8). Elastosis, a marker of background actinic photodamage present in eight out of eight lesions at baseline and excluded from the malignancy-associated criterion count, is distinct from post-treatment fibrosis, a radiation-induced dermal remodelling feature scored only at follow-up; the two should not be conflated. Conclusions: In this proof-of-concept cohort, LC-OCT was feasible for pre- and post-treatment imaging of cSCC after 3D-printed HDR surface brachytherapy, produced highly reproducible dual-observer readings (κ = 0.903), and documented, at a single short-term time point six months after the final fraction, an LC-OCT-defined remission phenotype in every treated lesion, with a consistent fibrosis signature. Persistent dilated linear vessels in a fibrotic, otherwise criterion-negative dermis were classified as a non-specific post-radiation vascular change and did not preclude LC-OCT-defined complete remission. Because this remission phenotype rests on a pre-specified but not externally validated imaging definition and was not anchored to post-treatment histopathology or reflectance confocal microscopy, it constitutes an imaging surrogate rather than a confirmed complete remission, and microscopic residual disease below the resolution limit of LC-OCT cannot be excluded. These hypothesis-generating findings support LC-OCT as a candidate non-invasive monitoring modality, requiring confirmation in larger, biopsy-anchored prospective studies with longer follow-up.