Abstract
Asvold [1] and others reported a prospective observational study evaluating concentrations of human chronic gonadotropin (HCG) in consecutive early pregnancies following IVF and the subsequent risk of developing preeclampsia. Research has been conducted assessing the cause and etiology of preeclampsia, where one area of investigation that is showing potential is the early markers of preeclampsia [2].
It is well known that preeclampsia involves some aspect of vascular pathology and studies have confirmed that impaired implantation can potentially lead to subsequent preeclampsia [3]. HCG is produced by placental trophoblast cells from the time of implantation and has multiple functions in pregnancy. These include promotion of corpus luteal progesterone production, uterine angiogenesis and vasculogenesis, cytotrophoblast differentiation, umbilical cord development and growth and differentiation of fetal organs [4].
Because of the importance of this hormone in pregnancy growth and development, it has been theorized that deficient levels of HCG can be associated with adverse pregnancy outcomes. Romero [3] found that when spiral artery remodeling was incomplete several disorders could occur, including fetal growth restriction, preeclampsia, preterm labor, placental abruption and preterm premature rupture of membranes. Others studies have looked at the specific subunithyperglycosolated HCG. Low levels are associated with early pregnancy loss and early onset preeclampsia [5].
One of the difficulties in evaluating HCG as a predictor in clinical studies is ensuring reliable estimates of gestational age. This limitation is not present with IVF because gestational ages are very accurate. The authors used this precision in gestational age dating to their advantage when selecting their cohort population. All of their subjects underwent IVF with autologous oocytes. The status of the embryos (frozen or thawed) and the addition of intracytoplasmic sperm injection (ICSI) were not significant. By utilizing this population, they could accurately obtain day 12 HCG levels. This is very significant, because HCG levels can change dramatically over 1–2 days in early gestation.
The outcome of interest was the development of preeclampsia and the authors chose very traditional definitions for this outcome. Interestingly, they did not use 24 hour urine collections or spot protein/creatinine ratios for determination of proteinuria, although these are the accepted gold standard [6]. In addition, for severe preeclampsia, there is no mention of clinical signs and symptoms of persistent CNS symptoms, pulmonary edema or seizures. There is also no mention of additional lab abnormalities, such as increased liver enzymes, low platelets or evidence of microangiopathic hemolytic anemia.
Because this is a cohort study, the authors were able to calculate odds ratios, and did so using multivariate logistical regression. The hypothesis is that low HCG levels are associated with preeclampsia; therefore, the highest category of HCG level was used as the reference value. HCG levels were divided into four categories (<50, 50–99, 100–149, and ≥150 IU/I), and all analyses were based on these four categories.
The data set that was available had additional information for all patients from December 1998 onwards, so the authors performed separate analysis on this portion of the data set. This comprised 86.4% of the entire data set. In addition to analyzing the association with preeclampsia, the authors also looked at the risk of gestational hypertension, preterm delivery and small for gestational age delivery. Supplementary analysis included associations of HCG with preterm and term preeclampsia, and risk of preeclampsia based on type of IVF treatment, cause of infertility and number of transferred embryos. Analyses were also performed after defining HCG levels as multiples of the median. Finally, the authors evaluated if HCG measurements may have changed over the timeframe from the December 1998 onwards portion of the data set.
Results
The results of the study were significant for preeclampsia and severe preeclampsia. Only the women with HCG less than 50 IU/L had a significantly increased risk of developing preeclampsia and severe preeclampsia. Otherwise, none of the other analyses reached statistical significance.
The authors mention several strengths of their study, with which we agree. As mentioned above, due to the nature of IVF, consistent day 12 HCG values were obtained, and the same laboratory was used to determine the measurements. In addition, while not specifically stated, it appears these day 12 HCG values were not available to the obstetricians caring for the patients, and therefore were not a factor in diagnosing preeclampsia. Exogenous HCG was administered for ovulation induction, but as the authors point out, Damewood et al. [14] demonstrated that after 14 days exogenous serum HCG was no longer detectable. All the patients in this study had levels obtained 16 days after induction of ovulation.
The primary weakness in this study is the lack of generalization to the typical non-IVF pregnant population. This study looks at HCG levels much earlier in pregnancy than other studies. Similar findings were reported by Keikkala [7], but these were based on HCG levels obtained from 8 to 13 weeks gestation. This is an important distinction, because while there is less variance among HCG levels at very early gestations, only patients undergoing IVF are consistently seen this early in pregnancy. The results are statistically significant, but may be of limited clinical significance. Another potential weakness is the fact that there is some evidence that women conceiving via IVF are more likely to develop preeclampsia than those with spontaneous conceptions [8].
Significance
Preeclampsia and hypertensive disorders occur in approximately 4% of pregnancies, but account for 14% of maternal mortality. A recent study showed that hypertensive disorders are the fifth leading cause of maternal mortality in developed countries [9]. With such a high prevalence, identifying screening and prevention tools is of the utmost importance.
Currently at the initial prenatal visit, aside from identifying risk factors, there is no effective means of screening for the development of preeclampsia. Thilaganathan et al. [10] looked at Cystatin C levels in the first trimester, and found a statistically significant association between elevated levels and the subsequent development of preeclampsia. Poon et al. [11] found that maternal history and uterine artery pulsatility index were useful first trimester screening tools for preeclampsia, but that incorporating values of pregnancy associated plasma protein A (PAPP-A) did not increase their detection rate. Bills et al. [12] looked at VEGF165b levels throughout pregnancy and found that low levels in the first trimester predicted the development of preeclampsia in the third trimester. Grill et al. [2] performed a nice review of biochemical markers in the prediction of preeclampsia, including the findings from the above studies. Asvold's current study increases this body of knowledge by adding another marker that can potentially help predict future preeclampsia.
There is significantly less useful evidence available for methods of preventing preeclampsia. Many studies have been done evaluating antiplatelet drugs in the prevention of preeclampsia, with mostly negative findings. However, a large systematic review showed aspirin to confer a moderate (15%) decrease in the risk of developing preeclampsia. The authors of this study also acknowledge that prior large trials and meta-analyses failed to show any benefit from antiplatelet therapy, and therefore recommendations should be tailored to individual patients [13].
Future perspective
The current research trend makes it likely that some combination of biomarkers will allow the early prediction of preeclampsia. Going forth from this study, a prospective observational study looking at HCG levels early in pregnancy among all-comers, and not specifically IVF patients, could provide more generalizable data. There would be no means of obtaining consistent gestational age levels, but possibly correlating the HCG levels with sonographic crown rump length data might permit reasonable assessments.
Financial & competing interest disclosures
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.
Executive summary
Although the etiology of preeclampsia is not well understood, it appears that it becomes established at the earliest stages of pregnancy and implantation.
A large cohort study of IVF pregnancies found an association between low day 12 HCG levels and subsequent development of preeclampsia.
The results of this study are consistent with the literature in demonstrating the significance of low HCG levels for predicting preeclampsia, although at a much earlier gestational age than might be practical for spontaneous conceptions.