Influence of Semiquantitative [18F]FDG PET and Hematological Parameters on Survival in HNSCC Patients Using Neural Network Analysis

Chemokine Receptor-4 Targeted PET/CT Imaging with 68Ga-Pentixafor in Head and Neck Cancer-A Comparison with 18F-FDG and CXCR4 Immunohistochemistry

BACKGROUND

Head and neck squamous cell carcinoma (HNSCC) is common, and its incidence is increasing, particularly in HIV-infected individuals who present with more aggressive disease. Despite aggressive treatment, the prognosis remains poor because of resistance to chemoradiation therapy. So far, studies report very low [68Ga]Ga-Pentixafor avidity in HNSCC. This study investigated the diagnostic performance of CXCR4-directed imaging of carcinoma of the oral cavity, oropharynx, and nasopharynx with positron emission tomography/computed tomography (PET/CT) using the radiolabelled chemokine ligand [68Ga]Ga-Pentixafor and explored its ability to quantify CXCR4 expression in vivo.

MATERIALS AND METHODS

In this prospective cross-sectional study, twenty-three (23) patients aged 52.9 ± 10.4 (19.6), 17 males and 6 females with primarily diagnosed (n = 17) or pre-treated (n = 6) SCC of the oral cavity (OCSCC, n = 11), oropharynx (OPSCC, n = 9), nasopharynx (NPSCC, n = 2) and unknown primary (n = 1) underwent imaging with [68Ga]Ga-Pentixafor-PET/CT. In 16/23 patients 2-[18F]fluoro-2-deoxy-D-glucose ([18F]F-FDG) served as a standard reference. All lesions were visually rated using a 5-point Likert scale. For both tracers, maximum standardized uptake values (SUVmax) and the total lesion uptake (TLU) were recorded and compared using the Wilcox-signed rank test. In addition, the tumor-to-background ratios were derived using the liver (TLR), spleen (TSR), and posterior cervical muscles (TMR) as background. The relationships between the SUVs of the two tracers were assessed using the Spearman correlation. CXCR4 immunohistochemistry (IHC) staining was correlated with 68Ga-Pentixafor-PET/CT in 21/23 patients.

RESULTS

Ninety-one percent (21/23) of tumors were visually detected on [68Ga]Ga-Pentixafor; however, [68Ga]Ga-Pentixafor was less intense compared with [18F]F-FDG-PET. Quantitative analysis showed higher [18F]F-FDG SUVmax in comparison with [68Ga]Ga-Pentixafor (16 ± 6.7 vs. 5.8 ± 2.6 g/mL, p = 0.011) and SUVmean (9.3 ± 4.1 vs. 3± 1.6 g/mL, p < 0.001) and TBR 4.9 ± 2.3 vs. 2.36 ± 1.4 p = 0.014. Nasopharyngeal cancer demonstrated more intense tracer accumulation than oropharyngeal and oral cavity malignancies. CXCR4 IHC staining was positive in 15/21 patients, and there was a statistically significant correlation between IHC staining and [68Ga]Ga-Pentixafor SUVmean r = 0.5 p = 0.027, and performance status r = 0.83 p = 0.0104.

CONCLUSIONS

In conclusion, although [68Ga]Ga-Pentixafor cannot replace [18F]F-FDG as a diagnostic tool because of its lower avidity, the correlation between CXCR4 targeted 68Ga-Pentixafor PET imaging and CXCR4 IHC staining indicates the potential of 68Ga-Pentixafor as an effective tool for selecting patients who may benefit from therapies targeting CXCR4. In addition, [68Ga]Ga-Pentixafor has no physiological brown fat uptake, which often obscures cervical lesions on [18F]F-FDG PET/CT imaging.

Artificial intelligence in head and neck surgery: Potential applications and future perspectives

Artificial intelligence (AI) has the potential to improve the surgical treatment of patients with head and neck cancer. AI algorithms can analyse a wide range of data, including images, voice, molecular expression and raw clinical data. In the field of oncology, there are numerous AI practical applications, including diagnostics and treatment. AI can also develop predictive models to assess prognosis, overall survival, the likelihood of occult metastases, risk of complications and hospital length of stay.

Determination of diagnostic and predictive parameters for vertical mandibular invasion in patients with lower gingival squamous cell carcinoma: A retrospective study

Vertical mandibular invasion of lower gingival squamous cell carcinoma (LGSCC) determines the method of resection, which significantly affects the patient's quality of life. Therefore, in mandibular invasion by LGSCC, it is extremely important to monitor progression, specifically whether invasion is limited to the cortical bone or has progressed to the bone marrow. This retrospective study aimed to identify the diagnostic and predictive parameters for mandibular invasion, particularly vertical invasion, to enable appropriate selection of the method of mandibular resection. Of the patients who underwent surgery for LGSCC between 2009 and 2017, 64 were eligible for participation in the study based on tissue microarrays (TMA) from surgical specimens. This study analyzed morphological features using computed tomography (CT), and metabolic characteristics using maximum standardized uptake value (SUVmax), peak value of SUV (SUVpeak), metabolic tumor volume (MTV), and total lesion glycolysis (TLG). Moreover, immunohistochemical analysis of proteins, including parathyroid hormone-related protein (PTHrP), interleukin-6 (IL-6), E-cadherin, and programmed cell death-1 ligand 1 (PD-L1), was performed. Statistical analysis was performed using univariate logistic regression analysis with the forward selection method. The present study showed that MTV (≥2.9 cm3) was an independent diagnostic and predictive factor for positivity of mandibular invasion. Additionally, TLG (≥53.9 bw/cm3) was an independent diagnostic and predictive factor for progression to bone marrow invasion. This study demonstrated that in addition to morphological imaging by CT, the volume-based parameters of MTV and TLG on fluorine-18 fluorodeoxyglucose positron emission tomography were important for predicting pathological mandibular invasion in patients with LGSCC. A more accurate preoperative diagnosis of vertical mandibular invasion would enable the selection of appropriate surgical procedure for mandibular resection.

The transformational potential of molecular radiomics

Conventional radiomics in nuclear medicine involve hand-crafted and computer-assisted regions of interest. Recent developments in artificial intelligence (AI) have seen the emergence of AI-augmented segmentation and extraction of lower order traditional radiomic features. Deep learning (DL) affords the opportunity to extract abstract radiomic features directly from input tensors (images) without the need for segmentation. These fourth-order, high dimensional radiomics produce deep radiomics and are well suited to the data density associated with the molecular nature of hybrid imaging. Molecular radiomics and deep molecular radiomics provide insights beyond images and quantitation typical of semantic reporting. While the application of molecular radiomics using hand-crafted and computer-generated features is integrated into decision-making in nuclear medicine, the acceptance of deep molecular radiomics is less universal. This manuscript aims to provide an understanding of the language and principles associated with radiomics and deep radiomics in nuclear medicine.