RESUMEN
BACKGROUND: The assessment of hidradenitis suppurativa (HS) severity requires detailed, and error-prone lesion counts. This proof-of-concept study aimed to automatically classify HS disease severity using machine learning of clinical smartphone images. METHODS: 777 ambient-light and size-controlled images were used to build a class-balanced synthetic dataset (n = 7675). Convolutional neural networks (CNN) were used for automated severity classification (scale 0-3), and to assess disease-dynamics. International Hidradenitis Suppurativa Severity Score System (IHS4) served as reference. A U-NET algorithm was implemented for automated localization of diseased skin. RESULTS: CNNs were able to distinguish no/mild from moderate/severe disease with an overall prediction accuracy of 78% [receiver operating curve (AUC) 0.85]. Correct IHS4 classification was achieved with an overall accuracy of 72% (AUC 0.84-0.89). In addition, disease dynamics using IHS4 numerical values aligned with CNN outputs (NRMSE 0.262). The UNET algorithm localized lesions with a pixel accuracy of 88.1% and test loss of 0.42. LIMITATIONS: Limitations in assessing tattooed and hairy skin. Limited number of patients with dark skin colour and Hurley I. CONCLUSION: CNNs were able to distinguish no/mild from moderate/severe disease, classify disease severity over time, and automatically identify diseased skin areas and the skin phototype. This study breaks new grounds for fast, reliable, reproducible and easy-to-use HS severity assessments using clinical images.
Asunto(s)
Hidradenitis Supurativa , Humanos , Hidradenitis Supurativa/diagnóstico por imagen , Índice de Severidad de la Enfermedad , Gravedad del Paciente , Redes Neurales de la ComputaciónRESUMEN
The initiation and maintenance of a malignant phenotype requires complex and synergistic interactions of multiple oncogenic signals. The Hedgehog (HH)/GLI pathway has been implicated in a variety of cancer entities and targeted pathway inhibition is of therapeutic relevance. Signal cross-talk with other cancer pathways including PI3K/AKT modulates HH/GLI signal strength and its oncogenicity. In this study, we addressed the role of HH/GLI and its putative interaction with the PI3K/AKT cascade in the initiation and maintenance of chronic lymphocytic leukemia (CLL). Using transgenic mouse models, we show that B-cell-specific constitutive activation of HH/GLI signaling either at the level of the HH effector and drug target Smoothened or at the level of the GLI transcription factors does not suffice to initiate a CLL-like phenotype characterized by the accumulation of CD5(+) B cells in the lymphatic system and peripheral blood. Furthermore, Hh/Gli activation in Pten-deficient B cells with activated Pi3K/Akt signaling failed to enhance the expansion of leukemic CD5(+) B cells, suggesting that genetic or epigenetic alterations leading to aberrant HH/GLI signaling in B cells do not suffice to elicit a CLL-like phenotype in mice. By contrast, we identify a critical role of GLI and PI3K signaling for the survival of human primary CLL cells. We show that combined targeting of GLI and PI3K/AKT/mTOR signaling can have a synergistic therapeutic effect in cells from a subgroup of CLL patients, thereby providing a basis for the evaluation of future combination therapies targeting HH/GLI and PI3K signaling in this common hematopoietic malignancy.
Asunto(s)
Proteínas Hedgehog/fisiología , Leucemia Linfocítica Crónica de Células B/etiología , Proteínas Oncogénicas/fisiología , Fosfatidilinositol 3-Quinasas/fisiología , Transducción de Señal/fisiología , Transactivadores/fisiología , Animales , Antígenos CD19/análisis , Linfocitos B/inmunología , Antígenos CD5/análisis , Humanos , Leucemia Linfocítica Crónica de Células B/tratamiento farmacológico , Ratones , Ratones Endogámicos C57BL , Proteínas Oncogénicas/antagonistas & inhibidores , Fosfohidrolasa PTEN/fisiología , Inhibidores de las Quinasa Fosfoinosítidos-3 , Receptores Acoplados a Proteínas G/fisiología , Receptor Smoothened , Transactivadores/antagonistas & inhibidores , Proteína con Dedos de Zinc GLI1RESUMEN
The Hedgehog (Hh) pathway regulates cell proliferation and survival and contributes to tumorigenesis. We investigated the expression and function of this pathway in B-cell chronic lymphocytic leukemia (CLL) cells and in healthy B lymphocytes. Profiling of cognate Hh pathway members revealed reduced expression of two key Hh signaling effectors, Smoothened (SMOH) and GLI, in CLL cells, whereas transcription levels of other investigated members resembled normal B-lymphocyte levels. Examining the functional role of SMOH and GLI in cell survival, we found that CLL cells were hardly sensitive toward specific SMOH inhibition, but showed an unspecific decline in cell viability in response to high concentrations of the SMOH antagonist cyclopamine. In contrast, treatment with the novel GLI antagonist GANT61 reduced expression of the target gene Patched and preferentially decreased the viability of malignant cells. Specific RNA interference knockdown experiments in a CLL-derived cell line confirmed the autonomous role of GLI in malignant cell survival. GANT61-induced apoptosis in primary leukemic cells was partly attenuated by protective stromal cells, but not soluble sonic hedgehog ligand. In summary, our data show a downregulation of the classical Hh pathway in CLL and suggest an intrinsic SMOH-independent role of GLI in the ex vivo survival of CLL cells.