Kasper et al., pp. 961–973 The interplay between pregnancy and breast cancer risk is well established. In general, women who have multiple pregnancies especially at a young age have a low risk of developing breast cancer. However, this protection might not apply to women with mutations in the BRCA1/2 genes. These women are at high risk (80%) to develop breast cancer during their lifetime. Since hereditary breast cancer often occurs during premenopausal years, it was hypothesized that the increase in estrogen and progesterone levels experienced during pregnancy might in fact increase the breast cancer risk in these women by promoting proliferation of mammary tissue. The new study by Cullinane and colleagues partially supports this hypothesis. In a large matched case-control study, the authors analyzed data from 1,260 pairs of women with known BRCA1 or BRCA2 mutations. They found that the risk of breast cancer was increased by 70% in women carrying the BRCA2 mutation. This elevated risk was only observed in women younger than 50 years of age and in the 2 years following pregnancy. The risk dropped if the last birth was 5 or more years in the past (OR = 1.24; 95% CI = 0.79–1.95). In contrast, the number of pregancies in women with a BRCA1 mutation had little or no effect. Women with a BRCA1 mutation and 4 or more children can expect a 38% decrease in breast cancer risk as compared to childless BRCA1 carriers. The differential effect of pregnancy on BRCA1 versus BRCA2 mutation carriers is interesting. Although the difference remains unexplained, the outcome of this study is likely to influence the clinical management of women with BRCA2 mutations. Tighter surveillance or prophylactic breast surgery might be indicated in the 5 years following pregnancy. Mizuno et al., pp. 1058–1059 The flagellated corkscrew-shaped bacterium Helicobacter pylori infects half of the world's population. It survives in the acid environment of the stomach and duodenum by hiding in the mucus and neutralizing stomach acid. Compelling evidence has linked the infection with H. pylori to the development of chronic gastritis, peptic ulcers and gastric cancer. This link has changed the treatment of these diseases. Instead of prescribing acid blockers to reduce stomach acidity, antibiotics are used in the successful treatment of peptic ulcers and the prevention of gastric cancer. To examine whether H. pylori infection is also associated with colon cancer, Mizuno and colleagues analyzed data from 334 patients who underwent high-resolution colonoscopy. H. pylori infection was diagnosed with serologic testing. They found that seropositivity was significantly higher in patients with adenomatous hyperplasias (p < 0.0001) and significantly lower in subjects with normal colonoscopic findings (p < 0.0005). While they observed highly significant differences in adenomatous polyps, no difference in hyperplastic polyps was observed in seropositive patients. Hyperplastic polyps are not considered neoplastic, while most colorectal cancers arise from adenomatous polyps. One consequence of this study could be that seropositive patients with adenomatous polyps will be treated with antibiotics in the future to prevent development of colon cancer. Grosse-Hovest et al., pp. 1060–1064 Full activation of T cells requires both a signal via the T cell receptor and so-called costimulatory receptors such as CD28. However, under certain circumstances triggering the CD28 costimulatory receptor alone is sufficient to induce “supraagonistic” T cell activation, which occurs in the apparent absence of T cell receptor-mediated signals. In this study, Hovest and colleagues use a bispecific, single-chain antibody (r28M) directed against CD28 and a melanoma-associated proteoglycan (MAPG) present on glioblastoma cells. R28M can be isolated in high concentrations from serum of cloned cows, which were engineered to produce the recombinant protein. Purified antibodies, when added to peripheral blood mononuclear cells, induced antigen-restricted T cell proliferation and tumor cell killing in the presence of MAPG-positive cells. Natural killer cells significantly contributed to its lytic activity. In vivo, a single dose of r28M was sufficient to suppress tumor growth in the striatum of nude mice inoculated with U87MG glioblastoma cells. At day 50, survival was 100% in r28M-treated mice coinjected with freshly isolated peripheral blood mononuclear cells and 0% in the control groups. The authors point out that the production of the antibodies in transgenic animals is reliable and will facilitate pilot clinical studies in which the protein will be applied to the treat patients with glioblastoma. The r28M antibody consists of 2 monospecific VH-VL single-chain FV fragments (scFV) linked by a particular linker sequence (L2) derived from the upper elbow of the CH1 domain. The bispecific antibody is directed against CD28 (α-CD28-scFV) and against a melanoma-associated proteoglycan (α-MAPG-scFV).