Suboptimal and inadequate quantification: an alarming crisis in medical applications of PET.

Dear Sir, In addition to its remarkable capability for unraveling the underlying molecular characteristics of a multitude of disorders with unprecedented sensitivity, positron emission tomography (PET) provides the most quantitative data that can be generated by any imaging technique. In spite of the vast amount of research that has been carried out in this domain over the years, the most reliable approaches that can reliably generate reasonably accurate results for either research or clinical purposes have yet to be clearly defined. Ideally, these approaches should be simple, reproducible, and can be disseminated readily worldwide for optimal management of patients in many settings. We strongly believe that an effective and decisive move in this direction would further enhance the role of this powerful imaging modality among physicians from different disciplines and will provide justification for replacing outdated technologies with PET in the daily practice of medicine. We believe that among several issues related to disease quantification with PET, certain items should be addressed with high priority by the PET imaging groups [1, 2]. Recent literature describing the issues related to standardization of quantification in PET imaging [3] have been published, which highlights the wide variability in patient preparation, image acquisition, and processing parameters in PET/CT centers. This variation and lack of a clear-cut consensus and understanding could be regarded as the major stumbling block in employing PET as a standard modality for either diagnosis or treatment monitoring in medical oncology and the other domains. It is important to recognize that the numbers generated superficially can lead to mismanagement of patients in a big way. In this communication, we would like to state some serious concerns about the evolution of this imaging modality as a quantitative technique over the past decade and would like to share our thoughts with the community so that the utility of this incredible and revolutionary methodology will be further enhanced in the future. Firstly, the magical threshold standardized uptake value (SUV) of 2.5, which has been promulgated in the literature to distinguish benign from malignant disease, is a major source of concern. As fluorodeoxyglucose (FDG) PET and PET/CT are utilized widely around the globe, the flaw in this concept is increasingly becoming apparent by the practitioners of the field. This is partly due to the fact that SUV of FDG changes substantially (but variably among lesions) over time and therefore, this temporal factor should be taken into consideration in generating the numbers for clinical and research purposes [2, 4, 5]. Furthermore, the SUV varies significantly among abnormal tissues depending on the underlying biology of the intended lesion, and this can be quite nonuniform [6]. Secondly, the impact of the partial volume effect upon SUV measurement is substantial, particularly for small lesions, but is often ignored entirely in clinical practice [1, 7]. We believe S. Basu Radiation Medicine Centre (BARC), Tata Memorial Hospital Annexe, Parel, Bombay 400012, India