Over a period of 2 years, 77 patients, strongly suspected of having pancreatic cancer, had abdominal CT as part of the diagnostic work-up. The CT images were reviewed by 2 radiologists who did not know the clinical course, the surgical procedure, or the final diagnosis. The positive predictive value of CT for the diagnosis of pancreatic cancer was 92% (82-97%) and the negative predictive value 69% (41-89%). The 95% confidence limits of the positive predictive value of CT in staging the extent of periampullary cancer in 52 patients varied between 21 to 79% and 59 to 100%, with the lowest values in diagnosis of liver metastases. The confidence limits for the negative predictive values in staging varied between 17 to 56% and 67 to 94%, with the highest values concerning liver metastases. We conclude that CT cannot be used as the only diagnostic procedure for confirming or excluding the diagnosis of pancreatic or periampullary cancer, and that the staging of periampullary cancer by CT alone is too inaccurate for the evaluation of resectability.

No abstract

No abstract

Human blood lymphocytes were stimulated in vitro by four Staphylococcus aureus strains. Activation of immunoglobulin‐secreting cells was determined by a reverse plaque forming cell (PFC) assay, and proliferation by quantitation of thymidine incorporation. Whole killed 5. aureus were slightly more efficient than water‐soluble preparations in the form of sonicated extracts and culture supernatants. Two S. aureus strains rich in protein A (Cowan I and E 2371) and one S. aureus strain deficient in protein A (E 1369) were potent B‐lymphocyte stimulators inducing maximal activity on day 6 of culture. Another S. aureus strain deficient in protein A (Wood 46) did not possess human lymphocyte stimulating capacity.