Chronic hypertension, perinatal mortality and the impact of preterm delivery: a population-based study.

OBJECTIVES: To estimate the association between chronic hypertension and perinatal mortality and to evaluate the extent to which risks are impacted by preterm delivery. DESIGN: Cross-sectional analysis. SETTING: United States, 2015-18. POPULATION: Singleton births (20-44 weeks of gestation). EXPOSURE: Chronic hypertension, defined as elevated blood pressure diagnosed before pregnancy or recognised before 20 weeks of gestation. MAIN OUTCOMES AND MEASURES: We derived the risk of perinatal mortality in relation to chronic hypertension from Poisson models, adjusted for confounders. The impacts of misclassification and unmeasured confounding were assessed. Causal mediation analysis was performed to quantify the impact of preterm delivery on the association. RESULTS: Of the 15 090 678 singleton births, perinatal mortality rates were 22.5 and 8.2 per 1000 births in chronic hypertensive and normotensive pregnancies, respectively (adjusted risk ratio 2.05, 95% CI 2.00-2.10). Corrections for exposure misclassification and unmeasured confounding biases substantially increased the risk estimate. Although causal mediation analysis revealed that most of the association of chronic hypertension on perinatal mortality was mediated through preterm delivery, the perinatal mortality rates were highest at early term, term and late term gestations, suggesting that a planned early term delivery at 37-386/7 weeks may optimally balance risk in these pregnancies. Additionally, 87% (95% CI 84-90%) of perinatal deaths could be eliminated if preterm deliveries, as a result of chronic hypertension, were preventable. CONCLUSIONS: Chronic hypertension is associated with increased risk for perinatal mortality. Planned early term delivery and targeting modifiable risk factors for chronic hypertension may reduce perinatal mortality rates. TWEETABLE ABSTRACT: Maternal chronic hypertension is associated with increased risk for perinatal mortality, largely driven by preterm birth.

Cutaneous melanoma arising from a cesarean delivery skin scar.

A 41-year-old pregnant African-American woman noticed rapid growth of her cesarean delivery skin scar beginning at 14-week gestation. Skin biopsy, which was performed at 31 weeks, revealed poorly differentiated cutaneous melanoma. At 34 weeks, she underwent repeat cesarean delivery with tumor excision, pelvic lymphadenectomy and abdominal wall reconstruction. Locally advanced disease and anatomical limitations prevented attainment of negative surgical margins. Despite adjuvant chemotherapy and radiation, she died 1 year after diagnosis. Deferring biopsy of a suspicious skin lesion during pregnancy may have delayed the diagnosis of melanoma in this case and possibly affected the long-term outcome.

Infection with a periodontal pathogen induces procoagulant effects in human aortic endothelial cells.

BACKGROUND: Multiple studies have demonstrated a link between periodontal infections and vascular disease. Porphyromonas gingivalis, a major periodontal pathogen, has been shown to adhere to and invade endothelial cells. OBJECTIVE: In order to dissect mechanisms underlying these observations, we assessed the role of P. gingivalis infection in modulating properties of endothelial cells linked to atherothrombosis. METHODS: Primary human aortic endothelial cells (HAEC) were infected with either P. gingivalis 381 or its non-invasive fimbriae-deficient mutant, DPG3. Markers of coagulation and thrombosis were assessed 8 h and 18 h postinfection in cell lysates and supernatants. RESULTS: Infection with P. gingivalis 381 significantly enhanced tissue factor expression and activity, and suppressed levels of tissue factor pathway inhibitor. Furthermore, P. gingivalis infection decreased levels and activity of tissue plasminogen activator, and enhanced plasminogen activator inhibitor-1 antigen and activity. Consistent with an important role for bacterial adhesion/invasion in this setting, infection with DPG3 failed to induce procoagulant properties in HAEC. Most of the above effects of P. gingivalis 381 were more apparent at the later time point (18 h postinfection). This suggests that P. gingivalis infection, rather than having an immediate and direct effect, might activate pathways that, in turn, trigger endothelial procoagulant mechanisms. CONCLUSIONS: Taken together these data demonstrate for the first time that infection with a periodontal pathogen induces procoagulant responses in HAEC.