PERSPECTIVE Bone Loss after Cardiac Transplantation creases osteoclast apoptosis and reduces the over- Shane et al. report a nonsignificant trend toward all number of active osteoclasts, thereby reducing fewer fractures in the treatment groups; the study, bone resorption. Alendronate may also reduce the however, was limited both by its small size and by frequency with which new units are activated. These the lack of randomized controls, so this observa- effects have been well documented in postmeno- tion is of unclear import. Although one may spec- pausal women with osteoporosis. In addition, alen- ulate that combination treatment with calcitriol and dronate may prolong the survival of osteoblasts. a bisphosphonate might confer an additive skele- Calcitriol therapy, which has also been investi- tal benefit, it is important to conduct further stud- gated in postmenopausal women with osteopo- ies in order to assess this possibility. Ultimately, rosis, reduces bone resorption at low doses but post-transplantation immunosuppressive proto- increases bone resorption at higher doses. The de- cols that minimize the use of glucocorticoids, cal- crease in resorption occurs through an indirect cineurin inhibitors, and other medications that have mechanism: the reduction of the circulating para- adverse effects on the skeleton will go a long way thyroid hormone level, as was seen in the study by toward reducing the prevalence of fractures after Shane et al. The primary effect of calcitriol is the cardiac transplantation. stimulation of intestinal calcium absorption — hence the requirement to control and monitor cal- From the Regional Bone Center, Helen Hayes Hospital, West Hav- cium excretion. erstraw, N.Y. Small-for-Gestational-Age Infants and Risk of Fetal Death in Subsequent Pregnancies Jun Zhang, Ph.D., M.D., and Mark A. Klebanoff, M.D., M.P.H. One of the triumphs of modern obstetrics has been Register to demonstrate that women whose first the dramatic reduction in late fetal mortality (de- infant was severely small for its gestational age had fined as the rate of antepartum or intrapartum fetal a significantly higher risk of late fetal death at 28 or death at 20 weeks of gestation or later) in the devel- more weeks of gestation in the subsequent preg- oped world during the second half of the 20th cen- nancy than women whose first infant was not small tury. The reduction can be attributed to better access for gestational age. The increased risk was further to vastly improved antepartum and intrapartum amplified if the first delivery was also preterm. For care; improved antepartum fetal surveillance; the example, as compared with women with a first de- virtual elimination of Rh isoimmunization; and the livery at term of an infant that was not small for revolution in neonatal care that enabled obstetri- gestational age, women who had a first delivery cians to deliver a fetus, particularly one remote from at term of an infant that was small for gestational term, when its health was in jeopardy. The fetal age had twice the risk of subsequent fetal death. mortality rate in the United States decreased from The odds ratio increased to 3.4 if the first delivery 18.4 per 1000 total births in 1950 to 6.7 per 1000 was also moderately preterm (occurring at 32 to 36 births in 2000. However, the pace of the decrease weeks of gestation) and to 5.0 if the first delivery has slowed substantially during the past 20 years. occurred before 32 weeks of gestation. Surkan et al. The number of fetal deaths is now equal to the num- defined an infant that was “small for gestational ber of infant deaths in the United States. The psy- age” as one with a birth weight that was more than chological effects of a late fetal loss on a woman and 2 SD below the mean at a given week of gestation her family can be as profound as those of the loss of (i.e., below the 2.5th percentile) — a cutoff that is a child. considered to indicate a severely small-for-gesta- In this issue of the Journal, Surkan et al. (pages tional-age infant by U.S. standards. 777–785) use data from the Swedish Medical Birth Women who have a small-for-gestational-age in- 754 n engl j med 350;8 www.nejm.org february 19, 2004 Downloaded from www.nejm.org on August 25, 2007 . Copyright © 2004 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Small-for-Gestational-Age Infants and Risk of Fetal Death in Subsequent Pregnancies Abruption 15 Anomalies 10 Infection 9 Diabetes 4 Hypertension 3 Labor 2 Intrauterine growth restriction 15 Other 17 Unexplained 25 Percent of Fetal Deaths Figure. Causes of Fetal Death. Data, for 61,957 pregnancies between 1978 and 1995, are from Fretts RC, Usher RH. Causes of fetal death in women of ad- vanced maternal age. Obstet Gynecol 1997;89:40-5. Fetal death was considered to have been caused by abruption if there was antepartum bleeding or a retroplacental blood clot; by an anomaly if there was a potentially lethal anomaly; by infec- tion if there was pathological evidence of infection in the fetus or the placenta; by diabetes if the mother was diabetic and the fetal death was otherwise unexplained; by hypertension if the mother was hypertensive and the fetal death was oth- erwise unexplained; by labor if there was unexplained intrapartum asphyxia; and by intrauterine growth restriction (“fetal malnutrition”) if the death was asphyxial or unexplained and the weight of the fetus was below the third percentile; a fetal death was attributed to intrauterine growth restriction only if there was no other identified cause, except in the case of hypertension (if there was both intrauterine growth restriction and maternal hypertension, the death was attributed to the former condition). The “other” category included the following causes: a prolapsed cord, knots, placenta previa, hydrops, twin–twin transfusion, maternal disease, trauma, and Rh disease. fant, a preterm delivery, or a fetal death tend to have plained even after careful pathological examination these outcomes repeated in subsequent pregnan- of the fetus and the placenta. When one considers cies. Surkan et al. further demonstrate that a histo- that conditions such as placental abruption, hyper- ry of delivering a small-for-gestational-age infant, tensive disorders, and idiopathic growth restriction particularly one delivered remote from term, is also themselves often have unknown causes, our lack of associated with an increased risk of late fetal death understanding of the underlying causes of late fe- in the succeeding pregnancy. This association pre- tal death becomes even more apparent. In order sumably occurs because reduced fetal growth and to focus on this long-neglected area, the National fetal death often share underlying causes, such as Institute of Child Health and Human Development congenital anomalies, placental defects, and ma- recently launched the Stillbirth Collaborative Re- ternal vascular disease. Abnormalities in placental search Network, a cooperative network of five sites implantation, vascularization, and function play a in the United States that are using standardized pro- particularly important role. As the fetus grows larg- tocols to investigate the scope and causes of still- er with advancing gestation, the demands on the birth in a given population. placenta increase. Poor growth in a fetus at less than Unfortunately, many of the underlying causes of 32 weeks’ gestation suggests that, even remote from fetal death can neither be predicted accurately nor term, the maternal–placental unit has already ex- be prevented. Although recent literature suggests hausted its reserve. The data presented by Surkan that Doppler velocimetry may help to prevent fetal et al. demonstrate that the conditions that cause this death through the earlier detection and confirma- exhaustion tend to recur and can have manifold tion of fetal compromise in high-risk pregnancies, effects. its low predictive value implies that many high-risk The broad categories of obstetrical conditions women would need to be monitored in order to pre- that are associated with fetal death are shown in the vent one fetal death. In addition, because abnormal Figure. Fully one quarter of fetal deaths are unex- findings on Doppler velocimetry are not specific for n engl j med 350;8 www.nejm.org february 19, 2004 755 Downloaded from www.nejm.org on August 25, 2007 . Copyright © 2004 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Small-for-Gestational-Age Infants and Risk of Fetal Death in Subsequent Pregnancies immediate fetal jeopardy, its use may lead to unnec- restriction for another. This variability may partial- essary early delivery, with its attendant risks. Once ly explain why, in the Swedish study, most of the a compromised fetus is discovered and the compro- women who had a fetal death (84 percent) had giv- mise is deemed to be severe enough to place the en birth at term to an infant that was not small for fetus at high risk of imminent death, there are few gestational age in the previous pregnancy. At the specific therapeutic options other than delivery, same time, pregnant women and their physicians which is undertaken after the administration of cor- should be reassured by the fact that even when the ticosteroids to the mother to promote fetal lung previous delivery occurred at less than 32 weeks of maturity if the fetus is remote from term. gestation and the infant was severely small for ges- Although being severely small for gestational age tational age, the probability of a late fetal death in is generally considered to be pathologic, a weight the current pregnancy was still less than 2 percent. above the cutoff point may not necessarily mean that From the Division of Epidemiology, Statistics, and Prevention Re- the fetal growth is normal. Since fetuses have vary- search, National Institute of Child Health and Human Develop- ing potential for growth, a given birth weight may ment, National Institutes of Health, Department of Health and Hu- be appropriate for one fetus but indicative of growth man Services, Bethesda, Md. 756 n engl j med 350;8 www.nejm.org february 19, 2004 Downloaded from www.nejm.org on August 25, 2007 . Copyright © 2004 Massachusetts Medical Society. All rights reserved.