Tumor heterogeneities and resilient tumor sub-volumes that require individualized treatment planning and delivery for an improved outcome make head and neck carcinomas (HNC) a clinical challenge, with long-term survival in this patient group remaining unchanged over the past several decades. A new CT- and PET-imaging-based approach that entails applying big data to personalizing HNC treatment protocols patients is therefore needed to better identify which HNC patient subgroups respond to which specific therapies, and there is significant room for progress in this area.
That’s the conclusion of a study published online June 22 in Journal of the American College of Radiology. For the study, Loredana Marcu of the University of South Australia and colleagues sought to assess what they deemed “the current status of knowledge and practice utilizing big data toward personalized therapy in head and neck cancers based on CT and PET imaging modalities.”
Marcu et al. first conducted a search of Medline to identify studies concerning the use of big data mining and radiomics in assessing and devising treatment plans for HNC patients. The search was limited to studies that had been published in English from 2000 onward. Reference pearling was executed, with duplication studies and conference abstracts removed. A total of 25 articles were then retrieved, analyzed, and summarized in a tabulated manner.
The researchers found that, while studies based on big data in HNC are presently “limited,” development within the field continues, yielding valuable input for personalized treatment of patients with HNC. “Current research shows that big data are not commonly used in HNC research, but there is a growing interest,” they wrote. “The main sources for big data harvesting in HNC include genomics, radiomics, clinical studies, and epidemiology, with the most researched area being the medical imaging-based radiomics.”
Moreover, the researchers noted that, although current literature demonstrates the potential of big data and radiomics to allow for better stratification of HNC patients with HNC and consequently, improved personalization of treatment approaches, broader clinical implementation and changes in practice cannot occur until physicians can access larger, more robust datasets. The ideal model for decision support in HNC, they concluded, “should be based on human-machine collaboration, namely ... software-based algorithms, physician innovation collaboratives and clinician mix optimization.”