To the EditorsDespite advances in imaging technology, including endoscopic ultrasound (EUS), computed tomography (CT), and positron emission tomography (PET), accurate prediction of tumor node metastasis (TNM) staging still remains suboptimal. A high pretreatment prediction rate is very important for surgeons and oncologists. Patients with upper gastrointestinal (GI) tract cancer and distant metastases or unresectable tumors and those with nodal status suggesting noncurative resection (R1, R2) benefit more from primary systemic chemotherapy or chemoradiotherapy than from surgery. A robust pretreatment stage suggesting the potential for complete tumor resection (R0) allows better designing for the extent of surgery and can improve survival.Could the combination of endoscopic and laparoscopic ultrasonography (EUS-LUS) in the pretreatment setting benefit patients with upper GI tract cancer? The clinical utility of a pretreatment EUS-LUS was evaluated for 936 GI tract cancer patients reported in the March issue of Surgical Endoscopy [1]. The TNM category and resectability groups predicted by EUS-LUS for patients with esophageal (n = 256), gastric (n = 273), or pancreatic (n = 407) cancer were associated with a statistically significant survival difference. The authors concluded that the pretherapeutic EUS-LUS was useful for the patient's selection strategy and provided a prognostic outcome similar to data from CT-based populations.Given the need for improving pretreatment TNM status prediction, this large-scale retrospective study can potentially provide useful clinical data. However, more information does not necessarily translate into clinically correct surgical or oncologic decisions. To draw conclusions, comparative-effectiveness research is required to assess, according to strict protocols, whether pretreatment EUS-LUS compared with CT or PET can increase the rates of robust prediction and medical decisions, and whether such decisions translate into improved oncology and quality of life for patients. Such conclusions cannot be drawn from the current study.It has been become clear that apart of imaging technology advances, we need genetic, genomic, molecular diagnostic, and predictive tools. Such biomarkers are urgently needed to predict the pretreatment TNM stage, prognosis, and treatment response for selection of the best possible multimodal treatment for the individual patient. Promises of such personalized decisions and tailored therapy are now provided by the latest advances in cancer genome sequencing and first steps toward understanding gene-expression regulation. By identifying genetic alterations and epigenetic aberrations underlying cancer development and metastases, we can understand the role of transcriptome and micro-RNAs. This exploration of both structure variation and function perturbation of the cancer genome opens one of the most exciting scientific ways for the development of biomarkers with clinical utility .