Analysis of parenchymal density on mammograms in 1353 women 25-79 years old.

OBJECTIVE: This study was undertaken to determine the frequency and distribution of dense mammograms. Factors that may affect parenchymal density of breasts among the diverse age groups of women who undergo mammography were also studied. MATERIALS AND METHODS: Mammograms of 1353 women between 25 and 79 years old who were grouped in 5-year age cohorts were reviewed. Breast density and the presence of benign calcifications were analyzed. RESULTS: Parenchymal density on mammograms decreased progressively in the patient cohorts 25-29 years old through 75-79 years old (Spearman correlation, p < .01). In the cohort of 25- through 29-year-old patients, 38% had predominantly (> 50%) fatty breasts. In the cohort of 75- through 79-year-old patients, 76% had predominantly fatty breasts. Increased parenchymal density mammograms were more common in women who had smaller breasts, had had fewer than two pregnancies, and underwent hormone replacement (p < .01). Forty-nine percent of women 50-79 years old undergoing hormone replacement had predominantly dense breasts, a percentage similar to that (48%) of the patient cohort of women 40-44 years old. Prevalence of benign calcifications also increased with age, from 8% at ages 25-29 to 86% at ages 75-79 (p < .01). CONCLUSION: In our study, a significant percentage (38%) of women who were 25-39 years old had predominantly fatty breast tissue that should not impede selective mammographic screening or diagnostic efforts in this age group.

Detection of pelvic recurrence of colorectal carcinoma: prospective, blinded comparison of Tc-99m-IMMU-4 monoclonal antibody scanning and CT.

PURPOSE: To prospectively compare the accuracy of imaging with technetium-99m-labeled Fab' fragment of the anti-carcinoembryonic antigen antibody (CEA) IMMU-4 with that of computed tomography (CT) for the detection of pelvic recurrence of colorectal carcinoma. MATERIALS AND METHODS: In 61 patients, blinded interpretations of both modalities were correlated with surgical-pathologic (n = 23) or clinical and CT follow-up findings (n = 38). RESULTS: Sensitivity and specificity with antibody scanning alone and combined with CT (79% and 84% vs 83% and 81%, respectively) were not significantly different from those values for CT alone (66% and 97%, respectively). Sensitivity of antibody scanning was greater for recurrences larger than 2 cm (94% vs 55% [P = .02]), serum CEA more than 2.5 ng/mL (91% vs 40% [P = .03]), and combined planar and single photon emission CT antibody scanning compared with planar alone (79% vs 48% [P = .03]), without a significant decrease in specificity. CONCLUSION: Antibody scanning does not improve on findings at CT alone for recurrent colorectal carcinoma but can help differentiate recurrent tumor from fibrosis.

Strong binding of single-stranded DNA by stem-loop oligonucleotides.

We report that oligodeoxynucleotides which form stem-loop hairpin structures and which have pyrimidine-rich loops can form strong complexes with complementary single-stranded DNA sequences. Stem-loop oligonucleotides were constructed with a 25-nt T-rich loop and with variable Watson-Crick stems. The complexes of these oligomers with the sequence dA8 were studied by thermal denaturation. Evidence is presented that the complexes are one-to-one, bimolecular complexes in which the pyrimidine loop bases comprise the outer strands in a pyr.pur.pyr triplex, in effect chelating the purine strand in the center of the loop. Melting temperatures for the loop complexes are shown to be up to 29 degrees C higher than Watson-Crick duplex of the same length. It is shown that the presence of a stem increases stability of the triplex relative to an analogous oligomer without a stem. The effect of stem length on the stability of such a complex is examined. Such hairpin oligomers represent a new approach to the sequence-specific binding of single-stranded RNA and DNA. In addition, the finding raises the possibility that such a complex may exist in natural RNA folded sequences.