The Problems and Promise of Central Pathology Review

Abstract

To the Editor, I read the interesting article concerning the central pathology review published in Pediatric and Developmental Pathology (Teot LA, et al. The Problems and Promise of Central Pathology Review: Development of a Standardized Procedure for the Children’s Oncology Group) [1]. The subject is of great importance and topical interest as a result of the increasing dissemination of this practice. Indeed, the central pathology review is considered a sine qua non condition in the majority of prospective and retrospective studies. The authors focused on the most important advantages related to the central review of the histological diagnosis and also underlined some possible disadvantages. The central pathology review better refines the target populations that have to be enrolled in a specific clinical trial. It is particularly helpful in case of rare diseases or in the case of diseases for which the interobserver reproducibility is particularly low. Furthermore, a pathologist skilled in a specific area is certainly better able to identify any morphological, immunohistochemical, ultrastructural, or molecular tumoral features, possibly those that are prognoses or therapy related. However, a diagnostic discrepancy between the institutional pathologist and the reviewer pathologist might elicit significant professional embarrassment or even legal risks. Further considerations could be made and discussed with the authors of the article and with the readers of the journal. A central review of the histological diagnosis could not completely annul the diagnostic variability. Indeed, well known is the possibility of significant intraobserver variability, even among pathologists considered to be expert in a specific field [2]. In addition, in the case of some tumors, objective diagnostic criteria are lacking, and consequently a nonreproducible diagnostic interpretation of the lesions is the rule among institutional pathologists as well as among reviewing pathologists [3]. In this regard, in some cases the molecular characterization of the tumor might be more helpful than the central pathology review. Moreover, the pathological diagnosis is just one of the many possible variables. Indeed, in order to properly stratify the patients, additional important variables, such as patient characteristics, the extent of the surgery, the timing of the adjuvant therapies, and the stadium of the disease, should always be centrally reviewed. The enrollment of a patient in a clinical trial need not deprive that patient of any chances of recovery. For this reason a quick diagnosis is mandatory. This statement opens a debate: when the diagnosis is complex or when it require ancillary studies (for example, immunohistochemistry, molecular study, or ultrastructural study), can the institutional pathologist take the necessary time that the case requires or should he immediately send the case to the reviewing pathologist? In this context we should decide between a horizontal and a vertical diagnostic algorithm. What is the true role of the institutional pathologist (To prepare the surgical samples? To make a provisional diagnosis? To send the samples to a reviewing pathologist?) and of the reviewing pathologist (To make the definitive and inappellable diagnosis?)? I believe that if a sufficient number of cases of a specific pathology are seen in an institution and if that institution is able to perform all the necessary ancillary tests, central pathology review could be avoided in the majority of cases. On the contrary, central pathology review should be strongly indicated if a specific pathology is rarely observed in an institution or if that institution is not able to perform all of the necessary ancillary tests. We see what we know and we know what we see.