The Author reply:
     We thank Mr Bhanot for his interest in this case report and his thoughtful comments (Bhanot UK. Pancreatic ductal adenocarcinoma associated with pancreatic ductal intraepithelial neoplasia (Letters to the Editor). Hepatobiliary Pancreat Dis Int 2008;7:106-107.). We fully agree with him that the current grading system for PanINs into PanIN-1, PanIN-2 and PanIN-3 lack interobserver reproducibility.^[1] At the Consensus meeting on classification of pancreatic intraepithelial neoplasia and intraductal papillary neoplasia held at The Johns Hopkins Hospital in 2003^[2] also it was felt that though the three-tiered classification systems for PanINs and IPMNs have been invaluable in the genetic analyses of these precursor neoplasms, there is considerable interobserver variability in the assigning of histologic grades and there is morphometric evidence that a two-tier system (low-grade/high-grade) may be more reproducible.^{[1, 3]} But, still they have recommended that till the time further genetic and morphologic studies refine the histologic grading of PanINs, and that  existing three tier grading should be adhered to for the sake of consistency in grading.
    We also fully agree with Mr Bhanot that practically in surgical decision-making, PanIN can be clubbed in two groups--PanIN-1 and -2 in one group, as these are usually incidental findings and they do not require further resection and PanIN-3 in the other group, as further resection is required in these patients to prevent recurrence.^[4]
    Though Hisa et al^[5] have shown that PanIN-3 might extend up to 25 mm, it is sometimes  difficult to differentiate between  PanIN and secondary ductal involvement by invasive carcinoma (i.e. cancerization of ducts).^{[6, 7]} Practically for deciding the extent of pancreatic resection, most surgeons would send the pancreatic resection margin for frozening and proceeding accordingly.