Cutting-edge insights: near-infrared imaging for surgical margin assessment in head and neck tumor resection: a systematic review and meta-analysis

Introduction

Surgical excision is the primary treatment modality for head and neck cancers (HNC). The primary objective of surgical intervention in the management of HNC is the complete eradication of the neoplastic lesion while simultaneously preserving the integrity of the adjacent healthy tissue within this intricate anatomical and functional domain. Ultimately, a pathologically confirmed negative margin status is inextricably linked to a reduced risk of local recurrence, improved progression-free interval, and enhanced overall survival, establishing it as the most pivotal predictive factor for HNC (1,2). This is particularly noteworthy when residual disease is identified at or in close proximity to the surgical margins, and the incidence of localized recurrence substantially escalates (3,4). Failure to achieve pristine surgical margins may necessitate supplementary therapeutic modalities, including chemotherapy, radiotherapy, and/or further surgical intervention (5). Given the intricate nature and critical role of structures in the head and neck region, along with the imperative need for optimal functional and aesthetic outcomes, the standards for margin status in HNC resection remain rigorously stringent.

Nevertheless, the positivity rate of surgical margins in the treatment of HNC remains considerably high. In national and institutional cohort studies, 11% to 26% of HNC patients exhibit positive surgical margins after resection (6,7). Conventionally, the assessment of resection margins has relied predominantly on subjective visual, tactile, and experiential evaluations conducted by surgeons. Nevertheless, these assessments, reliant upon subjective discernment of subtle variations in tissue density and surface morphology, are not sufficient for the detection of residual disease.

Presently, the established standard of care for detecting residual tumors involves sampling the primary specimen and its margins, followed by frozen section analysis. However, this technique has some limitations. The first constraint lies in the inherent subjectivity and potential for error in sampling tumor margins, resulting in a high probability of false negative assessments. The second constraint manifests as a requirement for the tumor specimen to exit the surgical suite after resection. In doing so, the specimen often loses its orientation relative to the surgical bed, and frozen section analysis incurs exorbitant costs and is a time-intensive process (8).

Near-infrared fluorescence (NIRF) imaging has recently emerged as a prospective method for real-time tumor visualization and localization during HNC surgery. Compared to existing intraoperative guidance methods, such as radioactive seed localization, its significance lies in its capability to offer non-invasive, real-time, high-resolution images of biological structures and processes, thereby minimizing radiation exposure (9,10). The real-time imaging capabilities of this technology allow for the superimposition of fluorescent images atop conventional bright-field imagery, significantly simplifying the correlation between signals and anatomical structures. Although there are many preoperative imaging methods for tumors, such as computed tomography (CT) and positron emission tomography-CT (PET-CT), there is currently no approved real-time method for intraoperative tumor imaging. However, NIRF imaging can aid physicians in swiftly evaluating surgical outcomes, enabling immediate adjustments. NIRF not only improves surgical outcomes but also minimizes time costs by considering changes in tumor positions through real-time tracking, thereby enhancing post-operative comfort for patients (11,12). Compared to other methods used for tumor margin imaging, such as narrow-band imaging, NIRF imaging is more precise in delineating the tumor border (13). Employing near-infrared (NIR) light with wavelengths of approximately 650–900 nm to excite tumor-targeted fluorescent probes enhances intraoperative margin delineation and detection of residual tumors. These fluorescent probes enable the visualization of tumor margins and the surgical cavity, consequently increasing the rate of successful resection. This technology has utility both intraoperatively and ex vivo, as tracers can selectively bind to tumor tissues, a binding that persists even after tumor excision. Moreover, ex vivo imaging allows the meticulous control of imaging parameters (14,15).

However, despite the existence of a systematic review on this subject, a meta-analysis has not yet been conducted, and there may be instances of underreporting (16). We conducted a systematic review of the literature adhering to the Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies (PRISMA-DTA) reporting checklist (17,18) (available at https://qims.amegroups.com/article/view/10.21037/qims-24-564/rc), aiming to analyze the efficacy and diagnostic role of NIRF imaging methods in margin assessment after HNC resection.

Methods

Search strategy

The search methodology, study selection, and reporting in this study were conducted in accordance with the PRISMA-DTA guidelines, using the Participants, Interventions, Comparisons, Outcomes, and Study Design (PICOS) criteria for systematic reviews as follows:

• Population: patients undergoing interventions for HNC.
• Intervention: surgical resection guided by NIRF.
• Comparator: histopathological analysis of tissues, which is the gold standard for diagnosis.
• Outcomes: sensitivity, specificity, and area under the curve.
• Studies: prospective and retrospective studies.

PubMed (https://pubmed.ncbi.nlm.nih.gov), Embase (http://www.embase.com), China National Knowledge Infrastructure (CNKI; https://www.cnki.net), and Scopus (https://www.scopus.com) databases were searched from inception until August 1^st, 2023 for studies involving NIRF-guided HNC resection. Peer-reviewed clinical articles were searched using terms such as “fluorescence-guided surgery”, “head and neck tumors”, and their synonyms. Each search query for a specific database was optimized (Appendix 1).

Eligibility criteria

The systematic review, adhering to PRISMA-DTA guidelines and PICOS criteria, included studies utilizing NIRF imaging technology to assess surgical resection margins in HNC.

Studies meeting at least one of the following criteria were excluded from this systematic review: (I) inappropriate study design for this review (case series/reports, experimental or laboratory studies, literature reviews, letters to editors, personal opinions of authors, books/chapters, and conference abstracts); (II) repetitive studies and/ or studies with results unrelated to the topic; (III) studies not yet peer-reviewed or formally accepted; and (IV) studies on other imaging modalities (narrowband imaging, high-resolution microscopic endoscopy imaging, and others) without involving NIRF. This study aimed to evaluate the effectiveness of NIRF imaging in diagnosing the post-operative margins of HNC resections.

Study selection

Studies were selected in two distinct phases. In the initial phase, two independent reviewers (K.L. and J.Y.) meticulously evaluated the eligibility of all titles and abstracts retrieved from the search databases, excluding articles that did not meet the predetermined criteria. Subsequently, in the second phase, the remaining articles were independently scrutinized by the same two reviewers to identify the articles contributing to this systematic review. A third author (H.L.) critically assessed this process and resolved potential discrepancies through consensus discussions. Additionally, manual searches were conducted for references to the selected articles and prominent scientific journals pertinent to the field of oral and maxillofacial surgery.

Data items and collection

The data collection process included gathering the following information: author names, publication year, affiliated hospital or medical institution, research objectives, number of included patients, tumor type, tumor staging information, tumor location, diagnostic criteria for the target disease, fluorophore type, administration of fluorophores, imaging protocols, outcomes, general demographic characteristics of the study population, and study conclusions. One reviewer (K.L.) extracted the general study features and results, which were independently verified by another reviewer (J.Y.). Any discrepancies were resolved through consensus discussions with additional reviewers (H.L. and Z.T.).

Study quality assessment

Two researchers (K.L. and J.Y.) independently analyzed methodological quality. The Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool was used. This tool includes four domains: patient selection, index test, reference standard, and flow and timing. Each domain was evaluated for bias risk, and the first three domains for applicability, leading to a clearer assessment of bias and applicability in primary studies on diagnostic accuracy. Each domain was categorized as low, high, or unclear risk (19).

For this analysis, methodological quality assessment was performed using Review Manager version 5.4.

Statistical analysis

We employed random- and mixed-effects models for the standard diagnostic test accuracy (DTA) meta-analysis (20,21). Sensitivity and specificity were summarized, and summary receiver operating characteristic (SROC) curves were constructed to estimate the diagnostic accuracy (according to the area under the curve). Statistical heterogeneity was determined using the I-square (I²) statistic: ≤25% indicated low heterogeneity (acceptable), >25% and ≤50% indicated moderate heterogeneity (moderate), >50% and ≤75% indicated high heterogeneity (substantial), and >75% indicated very high heterogeneity (significant) (22). The possible threshold effect was assessed using Spearman’s correlation coefficient. Statistical analysis was performed using Stata software (Statistics and Data 18^th edition, RRID: SCR_012763) for the meta-analysis, with a confidence interval (CI) of 95%. A P value of ≤0.05 was considered statistically significant.

Studies lacking the complete data to construct a 2×2 table were excluded from the meta-analysis of diagnostic test features, although they may have been included in the overall synthesis of results in the systematic review to facilitate analysis.

Results

Search results and characteristics

A total of 1,607 articles were retrieved from the four databases. After removing duplicates (n=395), 1,212 articles were screened for titles and abstracts, of which 998 were excluded. The primary reasons for exclusion were topics unrelated to the research objectives and an inappropriate study design. After a full-text review of the remaining 214 studies, 207 were excluded. The reasons for exclusion included unrelated results (n=88), inappropriate study design (n=73), other imaging modalities (n=37), and repetitive studies (n=9). For articles with overlapping populations, we comprehensively evaluated the quality of experimental design, sample size, and publication date to select one article from the same clinical trial for inclusion (Appendix 2). Finally, seven articles were included in the systematic review (Figure 1) (23-29).

Figure 1 Identification of studies from databases and registers. CNKI, China National Knowledge Infrastructure.

The studies included in this systematic review were published between 2017 and 2023. Among them, 5 were prospective studies (23-25,28,29) and 2 were retrospective studies (26,27). Additionally, 5 studies were single-arm studies (24,26-29), while 2 were cohort studies (23,25). The primary outcomes measured in these studies were the sensitivity, specificity, and area under the curve for detecting HNC surgical margins. These studies included 103 patients who underwent surgery to assess the diagnostic accuracy of surgical margins using NIRF. The margin status was assessed by evaluating the edges of excised specimens (25,28) in two studies, through the surgical cavity in one study (29), and by both approaches simultaneously in four studies (23,24,26,27). The reference standard for all seven included studies was pathological examination. Disease characteristics and imaging protocols were reported in the seven included studies, encompassing a total of 103 HNC patients, with the majority having squamous cell carcinoma (94 cases, 91%). Overall, among the studies providing tumor staging information, there were a total of 30 T1–2 tumors (51%) and 29 T3–4 tumors (49%). Four articles employed nonspecific fluorophores, three of which utilized free monomeric indocyanine green (ICG) (24,26,29), while the other investigated ONM-100 (OncoNano Medicine-100) (27), a pH-sensitive amphiphilic polymer that can assemble into micelles with ICG chelated in the micelle core. As for the specific targeted fluorescent agents used, one study employed panitumumab-IRDye800CW (28), and two studies used cetuximab-IRDye800CW (23,25). One study administered the fluorescent agent over 30 min before surgery (26), whereas the remaining six studies administered it 24 hours or more prior to surgery (23-25,27-29). These findings’ data are summarized in Table 1.

Risk of bias and quality assessment

The quality of the selected studies was assessed using a customized QUADAS-2 tool. When using QUADAS-2, we evaluated the risk of bias for the four domains: patient selection, index test, reference standard, and flow and timing. Additionally, we assessed applicability for the three domains: patient selection, index test, and reference standard. It is worth noting that the “flow and timing” domain of QUADAS-2 does not involve an applicability assessment. Instead, this domain primarily focuses on evaluating the internal validity concerning the timing and sequence of the diagnostic process within the studies (19,30). The total score for each study was calculated by adding the number of criteria met, with a higher score indicating a higher quality of evidence (Figure 2). Overall, the number of implemented quality assurance measures implemented in each study ranged from 8 to 12 (out of a total of 13). Across the four QUADAS-2 tool domains mentioned earlier, each study had an average of 10.1 quality assurance measures. The study by Richard et al. presents two “unclear” aspects in terms of risk of bias and one “unclear” and one “high risk” aspect in applicability (29). Notably, Stubbs et al.’s study had a “high risk” aspect in terms of risk of bias (24).

Figure 2 Display of risk of bias and applicability concerns. (A) Risk of bias and applicability concerns graph presents the assessment results of the study’s risk of bias. (B) Risk of bias and applicability concerns summary presents the specific content of the risk of bias and applicability concerns associated with the studies, included in the review.

Heterogeneity analysis showed I² values of 26% (95% CI: 0–100%) and 78% (95% CI: 52–100%) for NIRF specimen imaging with positive or negative close margins.

The diagnostic accuracy of NIRF imaging in tumor specimens

When collecting textual data, we observed that certain articles pertaining to the classification of tumor margin status deviated from the binary categorization. Instead, they adopted a trinary classification encompassing the category close. However, during distinct stages of tumor excision, it is imperative to apply divergent surgical margin criteria (31,32). Consequently, we incorporated the ‘close’ margin into either positive or negative margins, facilitating separate analyses and discussions. Our study included different sites and types of HNC with a relatively small sample size. The original studies used varying guidelines to define margins; therefore, we adopted fluorescence standards from the original studies.

We analyzed the diagnostic accuracy of NIRF imaging in determining the surgical margin status. Five studies with 80 participants included close margins as positive margins and used NIRF imaging to determine the margin status of the tumor specimens. The pooled results reported a sensitivity of 0.84 (95% CI: 0.39–0.98; Figure 3), specificity of 0.96 (95% CI: 0.80–0.99; Figure 3), and an area under SROC curve of 0.97 (95% CI: 0.95–0.98; Figure 4). Four studies involving 66 participants included a close margin as a negative margin, and their pooled results reported a sensitivity of 0.98 (95% CI: 0.10–1.00; Figure 5), specificity of 0.96 (95% CI: 0.45–1.00; Figure 5), and an area under the SROC curve of 0.99 (95% CI: 0.98–1.00; Figure 6).

Figure 3 Accuracy of individual studies on NIRF imaging of HNC specimens for surgical margin status. The close margin is included in positive margins. TP, true positive; FP, false positive; FN, false negative; TN, true negative; CI, confidence interval; NIRF, near-infrared fluorescence; HNC, head and neck cancer.

Figure 4 SROC curve of NIRF imaging of HNC specimens for surgical margin status. The close margin is included as a positive margin. SROC, summary receiver operating characteristic; SENS, sensitivity; SPEC, specificity; AUC, area under the curve; NIRF, near-infrared fluorescence; HNC, head and neck cancer.

Figure 5 Accuracy of individual studies on NIRF imaging of HNC specimens for surgical margin status. The close margin is included as a negative margin. TP, true positive; FP, false positive; FN, false negative; TN, true negative; CI, confidence interval; NIRF, near-infrared fluorescence; HNC, head and neck cancer.

Figure 6 SROC curve of NIRF imaging of HNC specimens for surgical margin status. The close margin is included as a negative margin. SROC, summary receiver operating characteristic; SENS, sensitivity; SPEC, specificity; AUC, area under the curve; NIRF, near-infrared fluorescence; HNC, head and neck cancer.

No statistically significant differences were observed when threshold effects were explored. The Spearman correlation coefficients were Spearman’s rho =−0.3441, P=0.5563, and Spearman’s rho =−0.5443 and P=0.6310 for the two classifications of including close margins as positive margins and including close margins as negative margins, respectively.

The diagnostic accuracy of NIRF imaging in the surgical cavity

Five studies evaluated the utility of NIRF in assessing residual cancer in the surgical cavity/tumor bed post-resection (23,24,26,27,29). However, owing to the limited data available, a meta-analysis was not feasible.

In four studies with clearly recorded results, sensitivity ranged from 33.33% to 100%, and specificity ranged from 85.71% to 100% in detecting residual cancer cells in the tumor bed post-resection. In a study by Pan et al., 4/20 patients showed ICG fluorescence abnormalities in the tumor bed, and two of them had confirmed positive surgical margins (26). Richard et al. and Steinkamp et al. also successfully utilized this method to determine the surgical margin status in 6/8 and 9/13 patients, respectively (27,29). In a study by Stubbs et al., all 14 patients showed no fluorescent residue in the surgical cavity, and pathological examination confirmed the accuracy of the surgical margin assessment (24).

Strength of NIRF

Four of the seven articles mentioned recorded NIRF intensity using the mean fluorescence intensity (MFI) or tumor-to-background ratio (TBR). TBR assesses the clarity of the target signal relative to background noise, while MFI represents the average fluorescence intensity of all detected particles or cells in the sample. Using both TBR and MFI together provides a comprehensive understanding of the fluorescence signal’s intensity and clarity. However, due to variations in imaging systems, types of fluorophores, and limited data, we did not perform a pooled data analysis (33). Two studies used non-specific fluorophores. In a study by Pan et al., ex vivo specimens were bisected, and the mean fluorescence intensities of tumor tissues, peritumoral tissues, and normal tissues were 380.15±141.24, 268.52±79.12, and 262.12±90.16 arbitrary units, respectively; the signal-to-background ratios (SBR) for tumor versus peritumoral and normal tissues were 1.38±0.22 and 1.43±0.27, respectively (26). In a study by Steinkamp et al., surgical cavity-driven margin assessment identified four fluorescent lesions, with three cases of positive or close margins having a mean TBR of 4.68 (range, 2.2–6.2). Using specimen-driven margin assessment, the median TBR for positive deep resection margins was 3.36±1.62 (27). Two studies used the specific fluorophore cetuximab-IRDye800CW. In Voskuil et al.’s study, the TBR ranged from 1.61±0.93 to 3.10±2.53 across different cohorts, with an increase in TBR in the pre-dosing cohort (25). In Moore et al.’s study, all patients, except for the microdose group, had a peak TBR greater than 2 (range, 2.2–14.1) for primary in situ tumor imaging. The pre-resection fluorescence intensity of the primary in situ tumor (20.6±4.6) was significantly greater than the background fluorescence (6.3±0.5) and the post-resection wound bed fluorescence (2.0±0.8) (23).

Discussion

Ensuring appropriate surgical margins is a crucial aspect of HNC surgical treatment, as it can preserve the integrity of the surrounding structures while increasing the extent of tumor removal (34). This systematic review aimed to summarize the effectiveness of NIRF guidance in HNC surgery. This meta-analysis focused on demonstrating the diagnostic capability of NIRF in determining the status of surgical margins.

Principal findings

We approached Richard et al.’s study with caution as it included children and adolescents. However, the study by Richard et al. only involved using residual fluorescence in the surgical cavity to determine the status of tumor margins. A total of three articles employed this method with complete available data, which resulted in a very limited sample size. The impact of individual studies was too significant, and the combined data from the three studies failed to provide sufficient statistical power, leading to instability and clinically meaningless results. Therefore, we did not include it in the quantitative analysis (29). Despite the high risk in the threshold criteria, we included the results of a study by Stubbs et al. in our quantitative analysis due to the comprehensive study design (24). We excluded Moore et al.’s study because of incomplete data extraction (23). In the heterogeneity analysis, when considering the close margin as either positive or negative for NIRF specimen imaging, the I² test results were 26% (95% CI: 0–100%) and 78% (95% CI: 52–100%). However, this heterogeneity was expected due to the different indicators involved. Considering the potential statistical bias from analyzing small datasets, we combined these data. The results of the heterogeneity analysis prompted us to select either a random or a mixed-effects model (35). When evaluating the status of HNC tumor surgical margins based on the fluorescence difference between post-operative tumor specimens and normal tissues, considering a close margin as positive or negative, both yielded wide CIs, probably due to the small sample size, leading to instability in the effect estimates. Under both criteria, the area under the SROC curve was 0.97 (95% CI: 0.95–0.98) and 0.99 (95% CI: 0.98–1.00), indicating a relatively robust result and a high diagnostic accuracy. Although our results indicate that NIRF imaging shows promise as a real-time method for evaluating HNC surgical margins, the 95% CIs for sensitivity and specificity are quite broad. This wide range suggests instability and unreliability in the results. Therefore, while the preliminary findings are encouraging, these results should be interpreted with caution in clinical applications.

When utilizing fluorescence in the surgical cavity to assess the surgical margin status, post-operative re-evaluation of the tumor bed using NIR ICG endoscopy aids in achieving clear margins because the wound bed should be ICG-negative after resection. In four studies with clearly recorded results, the sensitivity and specificity ranged from 33.33% to 100% and 85.71% to 100%, respectively, for detecting residual cancer cells in the tumor bed post-resection. Although this reflects a certain diagnostic capability, the lower sensitivity suggests challenges in identifying positive margins using this method. Low sensitivity may result in false negatives. Post-operative fluorescence in the surgical cavity/tumor bed did not conflict with specimen fluorescence. The operational difficulty and the time taken for surgery were manageable and acceptable. The relatively high specificity is advantageous for excluding non-disease cases and suggests that this technique can be used as an adjunct diagnostic tool after NIRF imaging (20).

Due to the presence of a threshold effect, DTA studies often exhibit greater heterogeneity than treatment/intervention studies (36). However, we did not observe statistically significant differences in the threshold effects. The Spearman correlation coefficient was calculated to evaluate the difference in fluorescence between tumor and normal tissues. The Spearman’s rho for tumor tissues when the close margin was included as positive was −0.3441 (P=0.5563) and −0.5443, 0.6310 when the close margin was included as negative. Although in both situations, there was a negative correlation trend, the lack of statistical significance (large P values) suggests that this correlation might not be significant, likely due to the relatively small sample size. Further research with larger sample sizes, more detailed data collection, and alternative analytical methods is needed to comprehensively evaluate whether there is a threshold effect between sensitivity and specificity.

Studies by De Ravin et al. (16) and Fernandes et al. (37) suggest that NIRF imaging-guided surgery can facilitate real-time delineation of surgical margins during HNC surgery, improving margin clearance rates and enhancing progression-free survival. Our findings, to some extent, support their conclusions and further validate the utility in specific scenarios of HNC surgery. In this study, we strictly defined the use of NIRF to assess surgical margins and delineate tumor edges. While accurate delineation of tumors aids surgeons in precise localization and excision during surgery, it also serves as a foundational guide for determining surgical margins. Diagnostic assessment of the tumor margin status requires more precise quantitative standards and supportive outcomes.

Qualitative research

For clinical translation, an optimal fluorescent probe should have good distribution, high affinity for the target, and rapid clearance from the bloodstream, allowing for its effective accumulation at the tumor site and rapid acquisition of high-contrast images (38). ICG is a pioneering fluorescent dye used in fluorescence imaging-guided surgical procedures. It has demonstrated its capability through heightened permeability and retention effects, realizing the prospect of passive targeting for sentinel lymph node detection (39). Cetuximab and panitumumab, which are labeled with NIRF dyes, are two commonly employed probes that target the epidermal growth factor receptor (EGFR). The EGFR stands as one of the most pivotal targets, overexpressed in 80–90% of cases of head and neck squamous cell carcinoma (40). When employed in conjunction with fluorophores, they retain the capability to delineate HNC and lymph node metastases during fluorescence imaging-guided surgery for HNC. Targeted probes offer real-time intraoperative differentiation of the molecular boundaries between cancerous and adjacent normal tissues (41,42). In the included studies, irrespective of whether non-specific fluorescence, free (monomeric) ICG or its conjugates, or specific fluorescence was employed, panitumumab-IRDye800CW and cetuximab-IRDye800CW exhibited a heightened diagnostic capacity for the assessment of surgical margin status.

It is noteworthy that De Ravin et al. (16) demonstrated not only the feasibility of NIRF-guided HNC resection surgery but also its applicability to various tumor pathologies, primary sites, and surgical procedures. Our results validate this assertion.

Issues in the current practice

Histopathological assessment of biopsy tissue remains the gold standard for HNC diagnosis, although other imaging techniques can be used to complement lesion detection and staging. NIRF is considered a novel surgical adjunct tool that aids in preoperatively delineating surgical margins, potentially sparing patients from additional morbidity associated with extra surgery or chemoradiotherapy. However, their use in humans is limited by the scarcity of clinically approved fluorescent probes (43). In this study, only ICG received formal approval for clinical use, whereas the fluorescence-conjugated antibodies panitumumab-IRDye800CW and cetuximab-IRDye800CW are still in phase I/II clinical trials (33,44).

Precautionary measures must be taken when NIRF is used for surgical margin delineation. For instance, ICG may accumulate not only in cancerous tissues but also in inflammatory tissues and surgical trauma areas. This could potentially reduce the specificity of the initial stages of enhanced permeability and retention. These effects occur at much lower levels and are due to defects in lymphatic drainage structures. The applicability and safety of fluorescence-conjugated antibodies, such as panitumumab-IRDye800CW and cetuximab-IRDye800CW, in the human body require further optimization (33,38).

Although both the convenience of real-time tumor excision surgery guided by NIRF and the advantage of back-table NIRF imaging as an auxiliary modality over postoperative pathological methods are evident, our results support the high diagnostic capability of NIRF. However, relying solely on visual judgments remains challenging. The detection and identification of residual fluorescence within the surgical cavity solely by the naked eye are difficult (25). NIR imaging technology has proven to be a high-information, highly penetrating, and non-invasive prospective adjunctive treatment modality, and surgery guided by NIRF has demonstrated pathological effectiveness and applicability in preserving surgical margins in liver resection procedures (45-47). However, its application and trials in HNC are lacking. Numerous deficiencies are associated with the current model, and further clinical and preclinical research is required better to integrate this method into HNC management (48,49). Furthermore, different imaging schemes require more standardized criteria and comprehensive optimizations.

Study limitations

This systematic review and meta-analysis has certain limitations. As previously noted in the “Results” section, there is a moderate risk of publication bias, favoring better diagnostic accuracy in small-sample studies. Diminished statistical power may not detect the true presence of effects or lead to erroneous results. In cases where the number of included studies or patients is limited, it is challenging to assess the performance with a larger sample size and increased confidence. Furthermore, owing to data limitations, subgroup analyses for various diagnostic tests, such as different fluorophores, imaging times, or different thresholds, could not be performed. There are theoretical differences in the tissue distribution and NIRF imaging efficacy between specific and non-specific fluorophores. However, owing to the limited number of articles, we could only qualitatively describe the different studies without performing a combined analysis to obtain reliable results (16,50,51). Moreover, the standards for surgical tumor margins vary among different tumor sites and types. Since we could not obtain complete data, we had to use the margin standards of each study for the inclusion and analysis of the results rather than conducting further stratification or subgroup analysis (52,53). Moreover, constrained by different imaging systems and fluorophores, we opted against pooled analysis in the absence of sufficient data for subgroup analysis.

Other uses and future studies on NIRF in HNC

NIRF-guided surgery allows surgeons to achieve a visible contrast between normal and cancerous tissues. It provides high-resolution images, enabling the visualization of microscopic tumor nodules and with good tumor specificity owing to the targeting of exogenous drugs. One of the initial applications of NIRF imaging-guided surgery in humans was the injection of unbound ICG around a tumor for sentinel lymph node biopsy in HNC (54,55). The evolution of tumor-specific fluorescence imaging has enhanced the detection of metastases to regional lymph nodes, further improving the ability to detect lymphatic vessels in the area (56,57). Fluorescence imaging is employed to detect distant metastases of squamous cell carcinoma in the lungs. It is, therefore, a crucial tool for treating HNC because distant metastases are contraindications for surgical intervention. Both tumor and lymphatic metastases exhibit bright fluorescence emission, in distinct contrast with normal structures (58).

NIRF is also used to delineate the surgical margin by associating pathological diagnosis with visual impressions and assessing the differentiation between normal and cancerous tissues. There have been promising outcomes with fluorescence demarcation in NIRF-guided HNC surgery. Notably, van Keulen et al. reported the detection of secondary lesions, which are often overlooked by surgeons (59). Real-time fluorescence imaging-guided tumor resection, as demonstrated in their study, presents promising prospects.

In addition, the traditional approach of assessing tumor resection based on the surgeon’s tactile feedback has been eliminated in transoral robotic surgery (TORS). TORS is preferred to open surgery because of its superior functional outcomes and low complication rates (60). In the future, more data-driven results may be needed to support the preference for TORS, which could be another crucial factor in supporting the use of NIRF-guided surgery (61). Furthermore, real-time imaging technology is particularly suited to support minimally invasive surgeries such as TORS, as they rely on video equipment to guide the surgical process. Real-time NIRF imaging greatly supports this need (62). Notably, with advances in artificial intelligence (AI), AI-based image augmentation, quantification, and optimization of video rates and imaging quality could reduce potential errors in assessing tumor surgical margins through visual evaluation using NIRF (63-66).

Conclusions

In this systematic review and meta-analysis, we identified seven studies describing the use of different fluorophores for the assessment of surgical margins in postoperative HNC specimens or surgical cavities. Quantitative research analysis has demonstrated the diagnostic accuracy and robust efficacy of NIRF in evaluating the status of surgical margins. Qualitative analysis further supports this conclusion. These results indicate that NIRF imaging is a promising approach for assessing or assisting in determining the status of surgical margins in HNC surgery. The prospects extend to the use of NIRF to assist various aspects of HNC surgery, as well as aid in TORS. This diagnostic assistance is beneficial and may contribute to positive outcomes in the assessment of HNC surgical margins, thereby enhancing overall surgical effectiveness.

Acknowledgments

We would like to acknowledge Editage (www.editage.cn) for providing professional language editing services, which greatly improved the clarity and readability of this manuscript.

Funding: This research was supported by China Postdoctoral Science Foundation (CPSF) (No. 2024M752238 to Z.G.); Postdoctoral Fellowship Program of CPSF (No. GZC20231787 to Z.G.); and Research and Develop Program, West China Hospital of Stomatology Sichuan University (No. LCYJ2023-DL-1 to C.L. and No. RD-01-202406 to N.G.).

Footnote

Reporting Checklist: The authors have completed the PRISMA-DTA reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-564/rc

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-564/coif). C.L. reports that this research was supported by Research and Develop Program, West China Hospital of Stomatology Sichuan University (No. LCYJ2023-DL-1). N.G. reports that this research was supported by Research and Develop Program, West China Hospital of Stomatology Sichuan University (No. RD-01-202406). Z.G. reports supports from China Postdoctoral Science Foundation (CPSF) (No. 2024M752238) and Postdoctoral Fellowship Program of CPSF (No. GZC20231787). The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Mitchell DA, Kanatas A, Murphy C, Chengot P, Smith AB, Ong TK. Margins and survival in oral cancer. Br J Oral Maxillofac Surg 2018;56:820-9. [Crossref] [PubMed]
2. Buchakjian MR, Tasche KK, Robinson RA, Pagedar NA, Sperry SM. Association of Main Specimen and Tumor Bed Margin Status With Local Recurrence and Survival in Oral Cancer Surgery. JAMA Otolaryngol Head Neck Surg 2016;142:1191-8. [Crossref] [PubMed]
3. Ghantous Y, Bahouth Z, Abu El-Naaj I. Clinical and genetic signatures of local recurrence in oral squamous cell carcinoma. Arch Oral Biol 2018;95:141-8. [Crossref] [PubMed]
4. Baldan F, Gnan C, Lazarevic M, Nikolic N, Mio C, Tepavcevic Z, Robiony M, Milasin J, Damante G. Somatic genomic imbalances in 'tumour-free' surgical margins of oral cancer. Int J Oral Maxillofac Surg 2023;52:831-8. [Crossref] [PubMed]
5. Ooms M, Ponke L, Puladi B, Winnand P, Heitzer M, Katz MS, Hölzle F, Modabber A. Impact of secondary tumor-free resection margins in soft tissue on local, regional, and distant recurrence in R0-resected oral squamous cell carcinoma. Head Neck 2022;44:1172-81. [Crossref] [PubMed]
6. Orosco RK, Tapia VJ, Califano JA, Clary B, Cohen EEW, Kane C, Lippman SM, Messer K, Molinolo A, Murphy JD, Pang J, Sacco A, Tringale KR, Wallace A, Nguyen QT. Positive Surgical Margins in the 10 Most Common Solid Cancers. Sci Rep 2018;8:5686. [Crossref] [PubMed]
7. Wong LS, McMahon J, Devine J, McLellan D, Thompson E, Farrow A, Moos K, Ayoub A. Influence of close resection margins on local recurrence and disease-specific survival in oral and oropharyngeal carcinoma. Br J Oral Maxillofac Surg 2012;50:102-8. [Crossref] [PubMed]
8. Noorlag R, de Bree R, Witjes MJH. Image-guided surgery in oral cancer: toward improved margin control. Curr Opin Oncol 2022;34:170-6. [Crossref] [PubMed]
9. Vonk J, de Wit JG, Voskuil FJ, Witjes MJH. Improving oral cavity cancer diagnosis and treatment with fluorescence molecular imaging. Oral Dis 2021;27:21-6. [Crossref] [PubMed]
10. Bortz MD, Khokar A, Winchester DJ, Moo-Young TA, Ecanow DB, Ecanow JS, Prinz RA. Radioactive iodine-125 seed localization as an aid in reoperative neck surgery. Am J Surg 2021;221:534-7. [Crossref] [PubMed]
11. Vlocskó M, Piffkó J, Janovszky Á. Intraoperative Assessment of Resection Margin in Oral Cancer: The Potential Role of Spectroscopy. Cancers (Basel) 2023;16:121. [Crossref] [PubMed]
12. Kok END, Eppenga R, Kuhlmann KFD, Groen HC, van Veen R, van Dieren JM, de Wijkerslooth TR, van Leerdam M, Lambregts DMJ, Heerink WJ, Hoetjes NJ, Ivashchenko O, Beets GL, Aalbers AGJ, Nijkamp J, Ruers TJM. Accurate surgical navigation with real-time tumor tracking in cancer surgery. NPJ Precis Oncol 2020;4:8. [Crossref] [PubMed]
13. de Wit JG, van Schaik JE, Voskuil FJ, Vonk J, de Visscher SAHJ, Schepman KP, van der Laan BFAM, Doff JJ, van der Vegt B, Plaat BEC, Witjes MJH. Comparison of narrow band and fluorescence molecular imaging to improve intraoperative tumour margin assessment in oral cancer surgery. Oral Oncol 2022;134:106099. [Crossref] [PubMed]
14. van Keulen S, van den Berg NS, Nishio N, Birkeland A, Zhou Q, Lu G, Wang HW, Middendorf L, Forouzanfar T, Martin BA, Colevas AD, Rosenthal EL. Rapid, non-invasive fluorescence margin assessment: Optical specimen mapping in oral squamous cell carcinoma. Oral Oncol 2019;88:58-65. [Crossref] [PubMed]
15. Seah D, Cheng Z, Vendrell M. Fluorescent Probes for Imaging in Humans: Where Are We Now? ACS Nano 2023;17:19478-90. [Crossref] [PubMed]
16. De Ravin E, Venkatesh S, Harmsen S, Delikatny EJ, Husson MA, Lee JYK, Newman JG, Rajasekaran K. Indocyanine green fluorescence-guided surgery in head and neck cancer: A systematic review. Am J Otolaryngol 2022;43:103570. [Crossref] [PubMed]
17. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart LA. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. [Crossref] [PubMed]
18. Page MJ, Moher D, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ 2021;372: [Crossref] [PubMed]
19. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, Leeflang MM, Sterne JA, Bossuyt PM. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529-36. [Crossref] [PubMed]
20. Trikalinos TA, Balion CM, Coleman CI, Griffith L, Santaguida PL, Vandermeer B, Fu R. Chapter 8: meta-analysis of test performance when there is a "gold standard". J Gen Intern Med 2012;27:S56-66. [Crossref] [PubMed]
21. Ochodo EA, Reitsma JB, Bossuyt PM, Leeflang MM. Survey revealed a lack of clarity about recommended methods for meta-analysis of diagnostic accuracy data. J Clin Epidemiol 2013;66:1281-8. [Crossref] [PubMed]
22. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60. [Crossref] [PubMed]
23. Moore LS, Rosenthal EL, Chung TK, de Boer E, Patel N, Prince AC, Korb ML, Walsh EM, Young ES, Stevens TM, Withrow KP, Morlandt AB, Richman JS, Carroll WR, Zinn KR, Warram JM. Characterizing the Utility and Limitations of Repurposing an Open-Field Optical Imaging Device for Fluorescence-Guided Surgery in Head and Neck Cancer Patients. J Nucl Med 2017;58:246-51. [Crossref] [PubMed]
24. Stubbs VC, Jaffe S, Rajasekaran K, Cannady SB, Shanti RM, Lee JYK, Newman JG. Intraoperative Imaging with Second Window Indocyanine Green for Head and Neck Lesions and Regional Metastasis. Otolaryngol Head Neck Surg 2019;161:539-42. [Crossref] [PubMed]
25. Voskuil FJ, de Jongh SJ, Hooghiemstra WTR, Linssen MD, Steinkamp PJ, de Visscher SAHJ, Schepman KP, Elias SG, Meersma GJ, Jonker PKC, Doff JJ, Jorritsma-Smit A, Nagengast WB, van der Vegt B, Robinson DJ, van Dam GM, Witjes MJH. Fluorescence-guided imaging for resection margin evaluation in head and neck cancer patients using cetuximab-800CW: A quantitative dose-escalation study. Theranostics 2020;10:3994-4005. [Crossref] [PubMed]
26. Pan J, Deng H, Hu S, Xia C, Chen Y, Wang J, Wang Y. Real-time surveillance of surgical margins via ICG-based near-infrared fluorescence imaging in patients with OSCC. World J Surg Oncol 2020;18:96. [Crossref] [PubMed]
27. Steinkamp PJ, Voskuil FJ, van der Vegt B, Doff JJ, Schepman KP, de Visscher SAHJ, Kelder W, Jayalakshmi Y, Gao J, Sumer BD, van Dam GM, Witjes MJH. A Standardized Framework for Fluorescence-Guided Margin Assessment for Head and Neck Cancer Using a Tumor Acidosis Sensitive Optical Imaging Agent. Mol Imaging Biol 2021;23:809-17. [Crossref] [PubMed]
28. Krishnan G, van den Berg NS, Nishio N, Kapoor S, Pei J, Freeman L, Lee YJ, Zhou Q, van Keulen S, Farkurnejad S, Condon J, Baik FM, Martin BA, Rosenthal EL. Fluorescent Molecular Imaging Can Improve Intraoperative Sentinel Margin Detection in Oral Squamous Cell Carcinoma. J Nucl Med 2022;63:1162-8. [Crossref] [PubMed]
29. Richard C, White S, Williams R, Zaghloul T, Helmig S, Sheyn A, Abramson Z, Abdelhafeez H. Indocyanine green near infrared-guided surgery in children, adolescents, and young adults with otolaryngologic malignancies. Auris Nasus Larynx 2023;50:576-85. [Crossref] [PubMed]
30. Kim KW, Lee J, Choi SH, Huh J, Park SH. Systematic Review and Meta-Analysis of Studies Evaluating Diagnostic Test Accuracy: A Practical Review for Clinical Researchers-Part I. General Guidance and Tips. Korean J Radiol 2015;16:1175-87. [Crossref] [PubMed]
31. Jang JY, Choi N, Jeong HS. Surgical Extent for Oral Cancer: Emphasis on a Cut-Off Value for the Resection Margin Status: A Narrative Literature Review. Cancers (Basel) 2022;14:5702. [Crossref] [PubMed]
32. Jang JY, Choi N, Ko YH, Chung MK, Son YI, Baek CH, Baek KH, Jeong HS. Differential Impact of Close Surgical Margin on Local Recurrence According to Primary Tumor Size in Oral Squamous Cell Carcinoma. Ann Surg Oncol 2017;24:1698-706. [Crossref] [PubMed]
33. Nishio N, van den Berg NS, van Keulen S, Martin BA, Fakurnejad S, Zhou Q, Lu G, Chirita SU, Kaplan MJ, Divi V, Colevas AD, Rosenthal EL. Optimal Dosing Strategy for Fluorescence-Guided Surgery with Panitumumab-IRDye800CW in Head and Neck Cancer. Mol Imaging Biol 2020;22:156-64. [Crossref] [PubMed]
34. Brennan PA, Dylgjeri F, Coletta RD, Arakeri G, Goodson AM. Surgical tumour margins and their significance in oral squamous cell carcinoma. J Oral Pathol Med 2022;51:311-4. [Crossref] [PubMed]
35. Dinnes J, Deeks J, Kirby J, Roderick P. A methodological review of how heterogeneity has been examined in systematic reviews of diagnostic test accuracy. Health Technol Assess 2005;9:1-113. iii. [Crossref] [PubMed]
36. Shim SR. Meta-analysis of diagnostic test accuracy studies with multiple thresholds for data integration. Epidemiol Health 2022;44:e2022083. [Crossref] [PubMed]
37. Fernandes JR, Dos Santos LCF, Lamers ML. Applicability of autofluorescence and fluorescent probes in the trans-surgical of oral carcinomas: A systematic review. Photodiagnosis Photodyn Ther 2023;41:103238. [Crossref] [PubMed]
38. Gao RW, Teraphongphom N, de Boer E, van den Berg NS, Divi V, Kaplan MJ, Oberhelman NJ, Hong SS, Capes E, Colevas AD, Warram JM, Rosenthal EL. Safety of panitumumab-IRDye800CW and cetuximab-IRDye800CW for fluorescence-guided surgical navigation in head and neck cancers. Theranostics 2018;8:2488-95. [Crossref] [PubMed]
39. Lin N, Gao J, Wang H, Zhang H, Xiang Y, Ma X. Diagnostic value of indocyanine green for sentinel lymph node mapping and lymph node metastasis in oral/oropharyngeal carcinoma. Oral Oncol 2021;122:105563. [Crossref] [PubMed]
40. Bhatia A, Burtness B. Treating Head and Neck Cancer in the Age of Immunotherapy: A 2023 Update. Drugs 2023;83:217-48. [Crossref] [PubMed]
41. Nakano K. Progress of molecular targeted therapy for head and neck cancer in clinical aspects. Mol Biomed 2021;2:15. [Crossref] [PubMed]
42. Day KE, Sweeny L, Kulbersh B, Zinn KR, Rosenthal EL. Preclinical comparison of near-infrared-labeled cetuximab and panitumumab for optical imaging of head and neck squamous cell carcinoma. Mol Imaging Biol 2013;15:722-9. [Crossref] [PubMed]
43. de Kleijn BJ, Heldens GTN, Herruer JM, Sier CFM, Piazza C, de Bree R, Guntinas-Lichius O, Kowalski LP, Vander Poorten V, Rodrigo JP, Zidar N, Nathan CA, Tsang RK, Golusinski P, Shaha AR, Ferlito A, Takes RP. Intraoperative Imaging Techniques to Improve Surgical Resection Margins of Oropharyngeal Squamous Cell Cancer: A Comprehensive Review of Current Literature. Cancers (Basel) 2023;15:896. [Crossref] [PubMed]
44. de Wit JG, Vonk J, Voskuil FJ, de Visscher SAHJ, Schepman KP, Hooghiemstra WTR, Linssen MD, Elias SG, Halmos GB, Plaat BEC, Doff JJ, Rosenthal EL, Robinson D, van der Vegt B, Nagengast WB, van Dam GM, Witjes MJH. EGFR-targeted fluorescence molecular imaging for intraoperative margin assessment in oral cancer patients: a phase II trial. Nat Commun 2023;14:4952. [Crossref] [PubMed]
45. Lin CK, Leu SW, Tsai YH, Zhou SK, Lin CM, Huang SY, Chang CC, Ho MC, Lee WC, Chen MC, Hung MS, Lin YC, Li JR, Lin BS. Increased tissue water in patients with severe sepsis affects tissue oxygenation measured by near-infrared spectroscopy: a prospective, observational case-control study. Quant Imaging Med Surg 2022;12:4953-67. [Crossref] [PubMed]
46. Aoki T, Murakami M, Koizumi T, Matsuda K, Fujimori A, Kusano T, Enami Y, Goto S, Watanabe M, Otsuka K. Determination of the surgical margin in laparoscopic liver resections using infrared indocyanine green fluorescence. Langenbecks Arch Surg 2018;403:671-80. [Crossref] [PubMed]
47. Xu Z, Qian J, Wu H, Meng C, Ding Q, Tao W, Ling CC, Chen J, Li P, Yang Y, Ling Y. Novel pH-activatable NIR fluorogenic spray mediated near-instant and precise tumor margins identification in human cancer tissues for surgical resection. Theranostics 2023;13:4497-511. [Crossref] [PubMed]
48. Cai X, Hong H, Pan W, Chen J, Jiang L, Du Q, Li G, Lin S, Chen Y. Does Using Indocyanine Green Fluorescence Imaging for Tumors Help in Determining the Safe Surgical Margin in Real-Time Navigation of Laparoscopic Hepatectomy? A Retrospective Study. Ann Surg Oncol 2023;30:1981-7. [Crossref] [PubMed]
49. Tashiro Y, Aoki T, Hirai T, Koizumi T, Mansou DA, Kusano T, Matsuda K, Yamada K, Nogaki K, Hakozaki T, Wada Y, Shibata H, Tomioka K, Yamazaki T, Saito K, Fujimori A, Enami Y, Hoffman RM, Murakami M. Pathological Validity of Using Near-infrared Fluorescence Imaging for Securing Surgical Margins During Liver Resection. Anticancer Res 2020;40:3873-82. [Crossref] [PubMed]
50. Kravchenko Y, Sikora K, Wireko AA, Lyndin M. Fluorescence visualization for cancer DETECTION: EXPERIENCE and perspectives. Heliyon 2024;10:e24390. [Crossref] [PubMed]
51. Xu D, Li L, Chu C, Zhang X, Liu G. Advances and perspectives in near-infrared fluorescent organic probes for surgical oncology. Wiley Interdiscip Rev Nanomed Nanobiotechnol 2020;12:e1635. [Crossref] [PubMed]
52. Li MM, Puram SV, Silverman DA, Old MO, Rocco JW, Kang SY. Margin Analysis in Head and Neck Cancer: State of the Art and Future Directions. Ann Surg Oncol 2019;26:4070-80. [Crossref] [PubMed]
53. Young K, Bulosan H, Kida CC, Bewley AF, Abouyared M, Birkeland AC. Stratification of surgical margin distances by the millimeter on local recurrence in oral cavity cancer: A systematic review and meta-analysis. Head Neck 2023;45:1305-14. [Crossref] [PubMed]
54. Jimenez-Lillo J, Villegas-Tovar E, Momblan-Garcia D, Turrado-Rodriguez V, Ibarzabal-Olano A, De Lacy B, Diaz-Giron-Gidi A, Faes-Petersen R, Martinez-Portilla RJ, Lacy A. Performance of Indocyanine-Green Imaging for Sentinel Lymph Node Mapping and Lymph Node Metastasis in Esophageal Cancer: Systematic Review and Meta-Analysis. Ann Surg Oncol 2021;28:4869-77. [Crossref] [PubMed]
55. Villegas-Tovar E, Jimenez-Lillo J, Jimenez-Valerio V, Diaz-Giron-Gidi A, Faes-Petersen R, Otero-Piñeiro A, De Lacy FB, Martinez-Portilla RJ, Lacy AM. Performance of Indocyanine green for sentinel lymph node mapping and lymph node metastasis in colorectal cancer: a diagnostic test accuracy meta-analysis. Surg Endosc 2020;34:1035-47. [Crossref] [PubMed]
56. Rosenthal EL, Moore LS, Tipirneni K, de Boer E, Stevens TM, Hartman YE, Carroll WR, Zinn KR, Warram JM. Sensitivity and Specificity of Cetuximab-IRDye800CW to Identify Regional Metastatic Disease in Head and Neck Cancer. Clin Cancer Res 2017;23:4744-52. [Crossref] [PubMed]
57. van Heumen S, Riksen JJM, Bramer WM, van Soest G, Vasilic D. Imaging of the Lymphatic Vessels for Surgical Planning: A Systematic Review. Ann Surg Oncol 2023;30:462-79. [Crossref] [PubMed]
58. Gleysteen JP, Newman JR, Chhieng D, Frost A, Zinn KR, Rosenthal EL. Fluorescent labeled anti-EGFR antibody for identification of regional and distant metastasis in a preclinical xenograft model. Head Neck 2008;30:782-9. [Crossref] [PubMed]
59. van Keulen S, Nishio N, Fakurnejad S, Birkeland A, Martin BA, Lu G, Zhou Q, Chirita SU, Forouzanfar T, Colevas AD, van den Berg NS, Rosenthal EL. The Clinical Application of Fluorescence-Guided Surgery in Head and Neck Cancer. J Nucl Med 2019;60:758-63. [Crossref] [PubMed]
60. Poupore NS, Chen T, Nguyen SA, Nathan CO, Newman JG. Transoral Robotic Surgery for Oropharyngeal Squamous Cell Carcinoma of the Tonsil versus Base of Tongue: A Systematic Review and Meta-Analysis. Cancers (Basel) 2022;14:3837. [Crossref] [PubMed]
61. Boehm F, Graesslin R, Theodoraki MN, Schild L, Greve J, Hoffmann TK, Schuler PJ. Current Advances in Robotics for Head and Neck Surgery-A Systematic Review. Cancers (Basel) 2021;13:1398. [Crossref] [PubMed]
62. Ma Z, Guo Z, Ding Z, Cao C, He J, Tang H, Hua Y, Hong J, Shen Q, Lubamba GP, Wang X, Yang Z, Zhu G, Li C. Evaluation of a newly developed oral and maxillofacial surgical robotic platform (KD-SR-01) in head and neck surgery: a preclinical trial in porcine models. Int J Oral Sci 2024;16:51. [Crossref] [PubMed]
63. Maza G, Sharma A. Past, Present, and Future of Robotic Surgery. Otolaryngol Clin North Am 2020;53:935-41. [Crossref] [PubMed]
64. Achterberg FB, Sibinga Mulder BG, Meijer RPJ, Bonsing BA, Hartgrink HH, Mieog JSD, Zlitni A, Park SM, Farina Sarasqueta A, Vahrmeijer AL, Swijnenburg RJ. Real-time surgical margin assessment using ICG-fluorescence during laparoscopic and robot-assisted resections of colorectal liver metastases. Ann Transl Med 2020;8:1448. [Crossref] [PubMed]
65. Koch M, Glatz J, Ermolayev V, de Vries EG, van Dam GM, Englmeier KH, Ntziachristos V. Video-rate optical flow corrected intraoperative functional fluorescence imaging. J Biomed Opt 2014;19:046012. [Crossref] [PubMed]
66. Ma Z, Wang F, Wang W, Zhong Y, Dai H. Deep learning for in vivo near-infrared imaging. Proc Natl Acad Sci U S A 2021;118:e2021446118. [Crossref] [PubMed]