Fully automated 3D machine learning model for HPV status characterization in oropharyngeal squamous cell carcinomas based on CT images.

BACKGROUND: Human papillomavirus (HPV) status plays a major role in predicting oropharyngeal squamous cell carcinoma (OPSCC) survival. This study assesses the accuracy of a fully automated 3D convolutional neural network (CNN) in predicting HPV status using CT images. METHODS: Pretreatment CT images from OPSCC patients were used to train a 3D DenseNet-121 model to predict HPV-p16 status. Performance was evaluated by the ROC Curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and F1 score. RESULTS: The network achieved a mean AUC of 0.80 ± 0.06. The best-preforming fold had a sensitivity of 0.86 and specificity of 0.92 at the Youden's index. The PPV, NPV, and F1 scores are 0.97, 0.71, and 0.82, respectively. CONCLUSIONS: A fully automated CNN can characterize the HPV status of OPSCC patients with high sensitivity and specificity. Further refinement of this algorithm has the potential to provide a non-invasive tool to guide clinical management.

Two single nucleotide polymorphisms in IL13 and IL13RA1 from individuals with idiopathic Parkinson's disease increase cellular susceptibility to oxidative stress.

The human genes for interleukin 13 (IL-13) and its receptor alpha 1 (IL-13Rα1) are in chromosomal regions associated with Parkinson's disease (PD). The interaction of IL-13 with its receptor increases the susceptibility of mouse dopaminergic neurons to oxidative stress. We identified two rare single SNPs in IL13 and IL13RA1 and measured their cytotoxic effects. rs148077750 is a missense leucine to proline substitution in IL13. It was found in individuals with early onset PD and no other known monogenic forms of the disease and is significantly linked with PD (Fisher's exact test: p-value = 0.01, odds ratio = 14.2). rs145868092 is a leucine to phenylalanine substitution in IL13RA1 affecting a residue critical for IL-13 binding. Both mutations increased the cytotoxic activity of IL-13 on human SH-SY5Y neurons exposed to sublethal doses of hydrogen peroxide, t-butyl hydroperoxide or RLS3, an inducer of ferroptosis. Our data show that both rs148077750 and rs145868092 conferred a gain-of-function that may increase the risk of developing PD.

Skin cancer-associated genodermatoses in skin of color patients: a review.

Skin cancers are associated with a large number of genodermatoses. Existing knowledge and guidelines on the presentations of these genodermatoses focus disproportionately on White patients. Our goal is to identify notable characteristics in location, frequency, and severity of cutaneous findings along with the median age of skin cancers in skin-of-color (SOC) patients with skin-cancer-associated genodermatoses to improve diagnosis rates. We searched for genodermatoses on six databases. Each case report or case series was reviewed, including reports, published in English, containing adult patient descriptions. Duplicate manuscripts were removed using EndNote. The following case-level data were collected from the manuscripts: age, gender, patient country or region of origin, author country/continent of residence, skin cancer-related, and other key dermatologic features. 381 published articles, with a total of 578 SOC patients, met criteria for inclusion. SOC patients can present with fewer classic findings, such as a lower incidence of basal cell carcinomas (44%) in SOC Gorlin syndrome patients than palmar pits (66%) and mandibular cysts (66%). Differences between SOC populations were also noted, such as leukoplakia being more common in Asian dyskeratosis congenita patients (80%) in comparison to African dyskeratosis congenita patients (44%). SOC patients also have varying onset of skin cancer depending on the genodermatosis, from a median of 25 years of age in Rothmund-Thomson syndrome to 53 in Muir-Torre syndrome. In this review, SOC patients with genodermatoses can have varying presentations. Being cognizant of these characteristics may lead to earlier diagnosis and interventions to mitigate skin-cancer-related morbidity in SOC patients.

Skin cancer in renal transplant recipients: outcomes from a safety net hospital in Boston.

BACKGROUND: Renal transplant recipients (RTRs) are prone to skin cancer due to the immunosuppression required to maintain graft function. Existing studies of skin cancer in RTRs focus on patients with Fitzpatrick skin types I-II, with limited documentation of incidence in skin types III-VI. This study seeks to better characterize skin cancers in RTRs with skin types III-VI. PRIMARY AIMS: Compare the incidence of skin cancer in RTRs of skin types I-II with skin types III-VI. SECONDARY AIMS: Explore the association between the development of skin cancer and other contributing factors in RTRs of skin types I-VI. METHODS: Retrospective chart review of RTRs at a single institution between January 1, 2000 and December 31, 2022. Patients were followed from the date of transplant to the last clinical follow-up or death. 777 RTRs were included in the study, including 245 patients with Fitzpatrick skin types I-II and 532 with skin types III-VI. A total of 48 patients developed NMSCs, 2 patients developed melanoma, and 3 patients developed Kaposi sarcoma. RESULTS AND CONCLUSIONS: There is a higher incidence of skin cancer in RTRs with Fitzpatrick skin types III-VI compared to the reported incidence among non-transplant recipients of the same skin types, but the incidence remains considerably lower compared to RTR of skin types I-II.

Multiparametric machine learning algorithm for human papillomavirus status and survival prediction in oropharyngeal cancer patients.

BACKGROUND: Human papillomavirus (HPV) status influences prognosis in oropharyngeal cancer (OPC). Identifying high-risk patients are critical to improving treatment. We aim to provide a noninvasive opportunity for managing OPC patients by training multiple machine learning pipelines to determine the best model for characterizing HPV status and survival. METHODS: Multi-parametric algorithms were designed using a 492 OPC patient database. HPV status incorporated age, sex, smoking/drinking habits, cancer subsite, TNM, and AJCC 7th edition staging. Survival considered HPV model inputs plus HPV status. Patients were split 4:1 training: testing. Algorithm efficacy was assessed through accuracy and area under the receiver operator characteristic curve (AUC). RESULTS: From 31 HPV status models, ensemble yielded 0.83 AUC and 78.7% accuracy. From 38 survival models, ensemble yielded 0.91 AUC and 87.7% accuracy. CONCLUSION: Results reinforce artificial intelligence's potential to use tumor imaging and patient characterizations for HPV status and outcome prediction. Utilizing these algorithms can optimize clinical guidance and patient care noninvasively.