Observer variation in the diagnosis of superficial oesophageal adenocarcinoma.

BACKGROUND AND AIMS: When to perform oesophagectomy for neoplastic progression in Barrett's oesophagus is controversial. Some resect for high grade dysplasia whereas others defer treatment until intramucosal adenocarcinoma is diagnosed. Interobserver agreement for a diagnosis of high grade dysplasia or intramucosal adenocarcinoma remains unknown and may have therapeutic implications. METHODS: Histological slides from 75 oesophagectomy specimens with high grade dysplasia or T(1) adenocarcinoma were blindly reviewed by two gastrointestinal pathologists and one general surgical pathologist, and classified as high grade dysplasia, intramucosal adenocarcinoma, or submucosal adenocarcinoma. A subsequent re-review of all 75 cases by the same observers following establishment of uniform histological criteria was undertaken. Interobserver agreement was determined by kappa statistics. Coefficients <0.21, 0.21-0.40, 0.41-0.60, 0.61-0.80, and >0.80 were considered poor, fair, moderate, good, and very good agreement, respectively. RESULTS: Interobserver agreement among all pathologists and between gastrointestinal pathologists when comparing high grade dysplasia with intramucosal adenocarcinoma was only fair (k=0.42; 0.56, respectively) and did not substantially improve on subsequent re-evaluation following establishment of uniform histological criteria (K=0.50; 0.61, respectively). CONCLUSIONS: When evaluating resection specimens and after implementation of uniform histological criteria, even experienced gastrointestinal pathologists frequently disagree on a diagnosis of high grade dysplasia versus intramucosal adenocarcinoma. Treatment strategies based on the histological distinction of high grade dysplasia from intramucosal adenocarcinoma using limited biopsy specimens should be re-evaluated.

Immunohistochemistry in the differential diagnosis of acinar and endocrine pancreatic neoplasms.

Histologic differential diagnosis of acinar cell carcinoma (ACC), mixed acinar-endocrine cell carcinoma (MAEC), and pancreatic endocrine tumors (PET) can be difficult but is important because of differences in their clinical behavior. This study investigates the utility of immunohistochemistry (IHC) in this differential diagnosis using immunohistochemical stains that are available in most laboratories. IHC was performed on paraffin-embedded tissue in ACC (n = 6), MAEC (n = 2), and PET (n = 13), using synaptophysin (SYN), chromogranin (CHR), chymotrypsin (CHY), and alpha-1-antitrypsin (AAT). Electron microscopy (EM) was performed in all cases to confirm the diagnosis. Long-term follow-up and death of disease (DOD) was known in all patients. The ACCs stained as follows: CHY (4/6), AAT (3/6), SYN (4/6); CHR was negative in all cases. Both cases of MAEC stained with CHY, AAT, and SYN (2/2); CHR was negative. PET stained as follows: SYN (13/13), CHR (8/13), CHY (4/13), AAT (5/13). In the ACC/ MAEC group, six of eight patients were DOD at mean follow-up of 11 months. Among the PET, two of 16 patients were DOD at mean follow-up of 37 months. Considerable immunophenotypic overlap exists between ACC, MAEC, and PET. Consequently, one can neither confirm nor rule out a diagnosis of ACC or MAEC using generally available immunohistochemical stains alone. These findings support a role for EM in the evaluation of exocrine and endocrine pancreatic neoplasms.

Integrating data from legacy systems using object linking and embedding technology: development of a reporting system for heavy metal poisoning results.

Integrating data that reside in different systems remains an often laborious process, requiring either manual steps or complicated programming. This paper describes a method for state-mandated reporting of childhood blood lead testing results that makes use of object linking and embedding technology and readily available software products to pull together information from different legacy systems. A terminal session emulator employs object linking and embedding automation to extract host data, and Visual Basic routines specify the user interface and database manipulation. This system has significantly increased the efficiency and accuracy with which blood lead testing reports are provided to the local state health department. The system provides a model for a relatively easy solution for laboratories and other groups that need a way to integrate standard data sets that are distributed across legacy systems.

Information system issues facing clinical laboratories serving complex integrated delivery systems.

Laboratory information systems (LISs) have evolved into complex applications to meet the specialized needs of laboratories. As integrated delivery systems (IDSs) continue to emerge as a dominant model for healthcare delivery, clinical laboratories serving them face two imperatives that affect IS decisions. First, laboratories in IDSs must consolidate to reduce costs and duplication yet deliver service across a region in both inpatient and outpatient settings. Second, laboratories that successfully perform reference laboratory testing will increase revenue, generate referrals, and leverage excess capacity, and may provide a competitive advantage for the IDS. This article examines laboratory information management in complex IDSs, presents options for IS support of consolidating or integrating laboratory operations, and reviews functionality requirements for laboratory outreach activities.

Serum percent-free PSA does not predict extraprostatic spread of prostate cancer.

Percent-free prostate-specific antigen (proportion of free prostate-specific antigen [PSA] to total PSA) has been shown recently in studies on frozen serum samples to be more useful than total PSA alone in distinguishing prostate cancer from benign conditions of the prostate gland. The primary purpose of our study was to determine whether percent-free PSA could predict extraprostatic spread of prostate cancer. We also sought to evaluate the freeze-thaw stability of free PSA. Percent-free PSA values in fresh serum samples were compared with those in aliquots subjected to one to five freeze-thaw cycles. Percent-free PSA values in frozen serum samples from 130 men undergoing radical prostatectomy for clinically localized prostate cancer were compared across pathologic stages. Free PSA levels remained stable for up to five freeze-thaws. Great overlap was found in percent-free PSA values for men with organ-confined disease and those with extraprostatic spread. These results indicate that multiple freeze-thaw cycles do not significantly affect free PSA levels and percent-free PSA is not useful in identifying ideal candidates for radical prostatectomy.

Dedifferentiated liposarcoma: a clinicopathological analysis of 155 cases with a proposal for an expanded definition of dedifferentiation.

We studied 155 cases of dedifferentiated liposarcoma to define its clinicopathologic features and behavior, in particular how the extent and grade of dedifferentiation affected outcome. Tumors occurred in late adult life (median, 61.5 years: range, 21-92 years), most commonly in the retroperitoneum (106 cases), extremities and trunk (32 cases), and scrotum/spermatic cord (13 cases). The majority of dedifferentiated liposarcomas presented as de novo lesions, whereas the remainder developed as a late complication of a preexisting well-differentiated liposarcoma after an average interval of 7.7 years. At the time of presentation, most of the dedifferentiated liposarcomas displayed extensive areas of high-grade dedifferentiation resembling malignant fibrous histiocytoma or high-grade fibrosarcoma, whereas a minority contained only areas of low-grade dedifferentiation resembling fibromatosis or well-differentiated fibrosarcoma. Divergent myosarcomatous or osteosarcomatous differentiation was observed focally in six cases. The behavior of dedifferentiated liposarcomas was that of a high-grade sarcoma with a local recurrence rate of 41%, a metastatic rate of 17%, and disease-related mortality of 28%. The most important prognostic factor was location in that retroperitoneal tumors had significantly worse survival than those in other sites. Tumors were divided into those having less than or those with more than 25% dedifferentiation, and dedifferentiated zones were classified into low grade or high grade. Neither low-grade dedifferentiation nor a low percentage of dedifferentiation was associated with an improved outcome for the tumors examined in this study; however, in no cases was the absolute size of the dedifferentiated focus <2 cm. Therefore, this study did not determine a minimum, or threshold, amount of dedifferentiation below which outcome was more favorable. The behavior of liposarcomas in which the dedifferentiated component was entirely low grade was more similar to that of traditional dedifferentiated liposarcoma than to that of well-dedifferentiated liposarcoma. Our study supports the expansion of the definition of dedifferentiated liposarcoma to include tumors with low-grade dedifferentiation and also suggests that low-grade dedifferentiation represents a precursor lesion of high-grade dedifferentiation.

Determination of the "reflex range" and appropriate cutpoints for percent free prostate-specific antigen in 413 men referred for prostatic evaluation using the AxSYM system.

OBJECTIVES: Prostate-specific antigen (PSA) exists in the serum in two clinically important molecular forms: free PSA and PSA complexed to alpha 1-antichymotrypsin. Total PSA approximates the sum of the free and complexed forms. Preliminary investigations have illustrated the potential benefits of using percent free PSA to enhance the clinical utility of PSA in distinguishing benign prostate disease from prostate cancer. The current study defines the optimal range of total PSA for measuring percent free PSA (reflex range) and generates appropriate cutpoints for percent free PSA within this range. METHODS: A total of 413 patients, 225 (54%) with benign prostate disease (mean age, 67 years) and 188 (46%) with prostate cancer (mean age, 66 years), who had PSA values between 2.0 and 20.0 ng/mL participated in the investigation. All patients underwent a sextant biopsy to establish the diagnosis. The serum specimens were assayed with the AxSYM PSA assay (total PSA) and AxSYM Free PSA assay (Abbott Laboratories; Abbott Park, IL). Percent free PSA was calculated for all patients. Receiver operating characteristic (ROC) curves were generated for various ranges of total PSA to determine the reflex range that maximized the increase in sensitivity and specificity of percent free PSA over total PSA. Within the optimal range, the ROC curves were utilized to generate cutpoints for percent free PSA to be used in clinical practice. RESULTS: The appropriate reflex range for the utility of percent free PSA was 3.0 to 10.0 ng/mL. The appropriate cutpoint for percent free PSA when the total PSA value was 3.0 to 4.0 ng/mL to achieve 90% sensitivity for the detection of prostate cancer was 0.19. This approach resulted in a biopsy rate of 73% and a cancer detection rate of 44% in men with a total PSA value between 3.0 and 4.0 ng/mL. The appropriate cutpoint for percent free PSA when the total PSA value was 4.1 to 10.0 ng/mL to ensure 95% sensitivity for detection of prostate cancer was 0.24. Within the range of 4.1 to 10.0 ng/mL, this approach resulted in 13% fewer negative biopsies and failure to detect 5% of the cancers. CONCLUSIONS: Percent free PSA should be utilized in patients with a total serum PSA value between 3.0 and 10.0 ng/mL. In patients with a total PSA value between 3.0 and 4.0 ng/mL, percent free PSA enhanced the detection of prostate cancer (improving sensitivity). In patients with a total PSA concentration ranging from 4.1 to 10.0 ng/mL, negative biopsies were eliminated (improving specificity).

Half-life determination of serum free prostate-specific antigen following radical retropubic prostatectomy.

OBJECTIVES: Prostate-specific antigen (PSA) continues to be the the most clinically useful tumor marker for prostate cancer. Recently, several molecular forms of PSA have been detected and characterized. These specific forms, including free PSA and PSA complexed to alpha 1-antichymotrypsin, can be measured and their proportions determined. In doing so, the sensitivity of PSA as a tumor marker can be maintained while the specificity is improved. In order to maximize the clinical utility of free PSA, the half-life and elimination kinetics of free PSA from the serum were determined. METHODS: Twenty-five patients, ages 43-74 years (mean 60 years) with biopsy proven, organ-confined adenocarcinoma of the prostate who underwent anatomic radical retropubic prostatectomy, were identified. For each patient, venous blood samples were obtained preoperatively, and at 60-minute intervals beginning 1 hour after the prostate was removed. The specimens were handled and stored in a consistent fashion. Using the AxSYM immunoassay analyzer (Abbott Diagnostics, Abbott Park, IL), the serum free PSA values were determined and plotted as a function of time for each patient. From the 25 individual elimination curves that were generated, the half-life of serum free PSA was determined. RESULTS: The mean half-life of serum free PSA was 110 minutes +/- 18.6 minutes (SD). Analysis of the individual and cumulative elimination curves indicates that the elimination of free PSA from the serum following radical prostatectomy follows a biphasic pattern. CONCLUSIONS: Unlike PSA, which has a half life of 2-3 days, the half-life of serum free PSA is 110 minutes (1.83 hours). This short half-life may have significant implications for the use of percentage of free PSA as a clinically useful tool in distinguishing patients with early, curable prostate cancer from men with benign prostatic hyperplasia (BPH) only.

Using hub technology to facilitate information system integration in a health-care enterprise.

The deployment and maintenance of multiple point-to-point interfaces between a clinical information system, such as a laboratory information system, and other systems within a healthcare enterprise is expensive and time consuming. Moreover, the demand for such interfaces is increasing as hospitals consolidate and clinical laboratories participate in the development of regional laboratory networks and create host-to-host links with laboratory outreach clients. An interface engine, also called a hub, is an evolving technology that could replace multiple point-to-point interfaces from a laboratory information system with a single interface to the hub, preferably HL7 based. The hub then routes and translates laboratory information to other systems within the enterprise. Changes in application systems in an enterprise where a centralized interface engine has been implemented then amount to thorough analysis, an update of the enterprise's data dictionary, purchase of a single new vendor-supported interface, and table-based parameter changes on the hub. Two other features of an interface engine, support for structured query language and information store-and-forward, will facilitate the development of clinical data repositories and provide flexibility when interacting with other host systems. This article describes the advantages and disadvantages of an interface engine and lists some problems not solved by the technology. Finally, early developmental experience with an interface engine at the University of Michigan Medical Center and the benefits of the project on system integration efforts are described, not the least of which has been the enthusiastic adoption of the HL7 standard for all future interface projects.