Objective
The objective of the study was to determine whether vascular and other complications are more common in pregnant women with neurofibromatosis type 1 (NF1).
Study Design
We performed a population-based retrospective cohort study using the US Nationwide Inpatient Sample, 1988-2009, defining a cohort of pregnancy-related hospitalizations with an associated diagnosis of NF1 and comparing it with the control group not associated with NF1. Multivariable logistic regression was used to adjust for suspected confounders.
Results
Among 19 million pregnancy-related admissions between 1988 and 2009, we identified 1553 associated with NF1 (prevalence 0.008%). A diagnosis of NF1 in delivering mothers was associated with gestational hypertension (adjusted odds ratio [AOR], 1.6, 95% confidence interval [CI], 1.2–2.0), preeclampsia (AOR, 2.8, 95% CI, 2.3–3.4), intrauterine growth restriction (AOR, 4.6, 95% CI, 3.7–5.6), cerebrovascular disease (OR, 8.1, 95% CI, 2.6–25.4), preterm labor (AOR, 1.6, 95% CI, 1.4–1.9), and cesarean delivery (AOR, 2.0, 95% CI, 1.8–2.3). Women with NF1 were not significantly more likely to have deep venous thrombosis/pulmonary embolism, acute cardiac events, or stillbirth or to die during their hospitalizations compared with the general obstetric population.
Conclusion
NF1 was associated with increased maternal morbidity in pregnancy (including hypertensive and cerebrovascular complications) but not increased maternal mortality. Obstetricians should be aware of the potential for increased antenatal and peripartum complications among women with NF1.
Neurofibromatosis type 1 (NF1) is an autosomal dominant tumor suppressor syndrome characterized by histologically benign tumors of the peripheral and central nervous system. NF1 is among the most common neurogenetic disorders, with a birth incidence of approximately 1 in 3000. NF1 is associated with an incompletely understood effect on the vascular system, which may lead to earlier onset of cardiovascular disease and increased cardiac mortality. NF1 predisposes to pheochromocytoma (a catecholamine-secreting adrenal tumor) and renal artery stenosis, both of which cause early-onset secondary hypertension. NF1 is also associated with a broad range of cerebrovascular abnormalities including moyamoya syndrome, cerebral aneurysms, and stenotic or ectatic cerebral vessels, which can in turn predispose to stroke or cerebral hemorrhage.
Although the association between NF1 and vascular disorders is well recognized, it is not known whether NF1 predisposes pregnant women to a higher rate of vascular and other complications in the antenatal and peripartum period. To date, much of the relevant literature is limited to either isolated case reports or small series reporting catastrophic outcomes. Several small case series of up to 10 patients report higher rates of hypertension, preeclampsia, intrauterine growth restriction (IUGR), preterm labor, and cesarean delivery in NF1 patients. In contrast, a questionnaire-based study of 105 participants did not confirm higher rates of adverse pregnancy outcomes, and a recent review of the natural history of NF1 did not include information on pregnancy complications.
Because fertility in NF1 is thought to be normal, and life expectancy, although reduced by a decade or more, exceeds the prime reproductive years, many young women with NF1 will experience 1 or more pregnancies. Because serious adverse events during pregnancy can occur as a result of vascular problems related to hypertensive and embolic events, it is crucial to these patients, their families, and the specialists caring for them to accurately estimate the risks associated with pregnancy in women with NF1.
The goal of our study was to fill in the knowledge gap related to this uncommon neurocutaneous disorder by using a large population-based national registry to investigate pregnancy complications in women with NF1. We hypothesized that at least in part because of the higher prevalence of vascular disorders in NF1, women with NF1 would have higher rates of vascular and other complications during pregnancy and delivery.
Materials and Methods
Data source
The Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality (AHRQ; Rockville, MD), collects administrative and clinical data on US hospital discharges yearly. The 2009 NIS, the most recent available data set, contains discharge data from 1050 hospitals in 44 states, approximating a 20% stratified sample of all nonfederal hospitals. Information available from the NIS includes up to 15 International Classification of Diseases , ninth revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes for each hospitalization, geographic region, hospital characteristics, and payor information.
Study population
Using the NIS, we designed a retrospective cohort study, defining the population of interest as all pregnancy-related hospitalizations between 1988 and 2009. To identify these hospitalizations, we searched for ICD-9-CM codes related to pregnancy (630-679). To define the study group within this population, we searched for diagnosis codes for NF1 (237.71), neurofibromatosis type 2 (NF2; 237.72), and unspecified neurofibromatosis (NF NOS; 237.70). Prior to 1990, the fifth digit modifier was not consistently used, so NF1 was coded as 237.7.
Because NF1 is much more common than NF2, we assumed that NF NOS would nearly always be classified as NF1 if appropriate clinical criteria were applied. To test these assumptions, we estimated the specificity of the NF2 code by comparing the frequency of hospitalizations involving surgery for vestibular schwannoma (VS) resection in the various groups (because bilateral VS is pathognomonic for NF2).
Approximately 1 in 6 discharges coded as 237.72 were associated with a VS resection, compared with 0.1% in discharges coded as 237.70, 237.71, 237.7 (prior to 1990), or in the non- neurofibromatosis population. Thus, codes 237.70, 237.71, and 237.7 were all imputed as NF1. The control group was defined as all pregnancy-related hospitalizations not associated with a diagnosis of NF1, as described above.
Outcomes and covariates
We used ICD-9-CM diagnosis and procedure codes to identify all hospitalizations associated with the following complications of pregnancy: preeclampsia (mild, severe, or superimposed on preexisting hypertension and including codes for eclampsia as well); IUGR; intrauterine fetal demise (stillbirth); cerebrovascular disease (defined as but not limited to venous sinus thrombosis, vascular occlusive disease, transient ischemic attack, and cerebral hemorrhage); deep venous thrombosis (DVT) and pulmonary embolism (PE); and acute cardiac complications (myocardial infarction, systolic or diastolic heart failure, and pulmonary hypertension). We also identified hospitalizations associated with preterm labor (early or threatened labor), delivery by cesarean section, and mortality. A full list of relevant ICD-9-CM codes is available from the authors.
We selected potential confounders based on known risk factors for pregnancy complications identified by a review of the relevant literature: age, chronic and gestational diabetes, chronic and gestational hypertension, renal disease, multiple gestation, race/ethnicity, and socioeconomic status. Race/ethnicity was defined as white, black, Hispanic, or other. Socioeconomic status was defined as income quartile based on median total family income within the ZIP code of the patient’s primary residence.
For cesarean delivery, we additionally considered year (due to the observed increase in the cesarean delivery rate for the general population in the United States over the time period studied) and prior cesarean delivery (an important determinant of whether a patient is offered the procedure for subsequent deliveries). We also adjusted for potential confounders that could independently affect the likelihood of a surgical procedure being offered: hospital size (small, medium, or large); hospital type (rural, urban nonteaching, or urban teaching); and primary insurance payor (Medicare, Medicaid, privately insured, or self-pay/uninsured).
Statistical analysis
We eliminated observations with missing data for any of the variables of interest. These totaled less than 0.1% for all categories except race and income quartile. Because not every state reports data on race, ethnicity, and socioeconomic status (∼30% of race/ethnicity data are missing, we created a separate category (termed missing) for these variables and adjusted for it in the statistical analysis.
We used multivariable logistic regression to determine odds ratios (ORs) and 95% confidence intervals (CIs) for the association between NF1 and the outcomes of interest. Final adjusted models included subsets of the covariates described above to control for conditions such as preexisting hypertension and renal disease when needed. We used an automated score selection algorithm to determine the set of covariates that provided the best explanation of the variation in the data, allowing no more than 1 covariate for every 10 events. For rare outcomes (eg, mortality), we report only univariate analyses.
Analyses were performed both on all admissions related to pregnancy and only those admissions associated with a delivery. Analysis of the subgroup of admissions associated with a delivery was done to eliminate effects from patients who were hospitalized repeatedly during 1 pregnancy and thus better estimate per-patient outcomes. Because preeclampsia, IUGR, stillbirth, and cesarean section are clinically relevant mainly for the delivering population, analyses for these outcomes are reported for deliveries only.
For all statistical analyses, we used SAS software (version 9.1; SAS Institute Inc, Cary, NC). Statistical significance is defined as P < .05; all P values are 2 tailed. Because obtaining accurate national estimates was not a focus of this study, the unweighted NIS data set was used, without accounting for its sampling design.
The Partners Institutional Review Board approved the use of the NIS and the internal database for this study. A waiver of informed consent was obtained for use of a publicly available, deidentified database.
Results
We identified 19,750,702 pregnancy-related admissions that occurred between 1988 and 2009. Of these, 1553 were associated with a diagnosis of NF1; 1248 (80.4%) of NF1 admissions were also associated with a delivery. The proportion of delivery-associated hospitalizations associated with a diagnosis of NF1 was approximately 1 in 10,000. Table 1 summarizes the demographic characteristics of both the NF1 and general obstetric populations. As expected, the majority of women (80.8%) were between 20 and 35 years of age with comparable proportions of women in the extremes of reproductive age (younger than 20 years or older than 35 years of age). However, pregnant women with NF1 were on average 1 year younger than women without it ( P < .001 for NF1 vs the control group).
| Characteristic | NF1, % (n) (n = 1553) | Control, % (n =19,750,702) | P value |
|---|---|---|---|
| Age, y | |||
| Mean (SD) | 26.4 (5.8) | 27.1 (6.1) | < .001 |
| Younger than 20 | 10.8 (167) | 11.8 | .24 |
| Older than 35 | 8.4 (131) | 9.9 | .06 |
| Race | < .001 | ||
| White | 40.9 (635) | 39.1 | |
| Black | 17.1 (266) | 10.3 | |
| Hispanic | 8.4 (130) | 14.6 | |
| Other | 3.5 (55) | 5.9 | |
| Missing | 30.1 (467) | 30.2 | |
| Income quartile | < .001 | ||
| 1 (least wealthy) | 31.6 (490) | 23.9 | |
| 2 | 27.3 (424) | 22.6 | |
| 3 | 19.9 (309) | 21.0 | |
| 4 (most wealthy) | 17.5 (271) | 30.0 | |
| Missing | 3.8 (59) | 2.6 | |
| Insurance payor | < .001 | ||
| Medicare | 3.2 (49) | 0.5 | |
| Medicaid | 56.1 (861) | 37.2 | |
| Private | 34.4 (527) | 53.9 | |
| Self-pay/other | 6.3 (47) | 8.5 | |
| Comorbidities | |||
| Chronic diabetes | 0.7 (11) | 1.1 | .21 |
| Chronic hypertension | 4.7 (73) | 1.8 | < .001 |
| Renal disease | 0.8 (13) | 0.3 | .002 |
| Multiple gestation | 1.74 (27) | 1.5 | .34 |
| Hospital setting | < .001 | ||
| Rural | 12.9 (200) | 11.6 | |
| Urban nonteaching | 33.4 (516) | 46.6 | |
| Urban teaching | 53.7 (830) | 41.8 | |
| Hospital size | < .001 | ||
| Small | 8.6 (133) | 11.5 | |
| Medium | 24.8 (383) | 28.2 | |
| Large | 66.6 (1030) | 60.3 | |
A large percentage of pregnant women with NF1 were from lower socioeconomic groups (quartiles 1 and 2) and covered by Medicaid or Medicare insurance. As expected, a higher percentage of pregnant women with NF1 had preexisting chronic hypertension (4.7% vs 1.8%) and renal disease (0.8% vs 0.3%), but these conditions were still present in only a small minority of NF1 patients. When compared with the control obstetrical population, women with NF1 were more likely admitted to either large or urban teaching hospitals. Despite the fact that the incidence of preexisting diabetes was similar in the 2 groups, NF1 patients had a significantly lower incidence of gestational diabetes (0.8% vs 3.7%, P < .001).
Table 2 and Figures 1 and 2 summarizes the odds ratios and 95% confidence intervals for all the outcomes of interest. In univariate analyses, complications such as gestational hypertension and preeclampsia were more frequent in the NF1 cohort, with ORs of 1.6 (95% CI, 1.2–2.2) and 3.0 (95% CI, 2.5–3.6), respectively, in the delivering population. IUGR was also more common in the NF1 cohort (OR, 5.6; 95% CI, 4.5–6.8, reported for deliveries only). Although the odds ratio for stillbirth was 1.9 (95% CI, 1.0–3.4) in the NF1 cohort, this difference did not reach statistical significance. NF1 was also more commonly associated with cerebrovascular events during pregnancy and in the peripartum period. DVT and PE were not more common in the NF1 cohort and no acute cardiac complications were observed in the group (baseline prevalence for general population, 0.04%).
| Outcome | Admission type | NF1, % (n) a | Control, % | Univariate OR (95% CI) | Multivariate OR (95% CI) | P value |
|---|---|---|---|---|---|---|
| Preeclampsia | Delivery only | 9.7 (121) | 3.5 | 3.0 (2.5–3.6) | 2.8 (2.3–3.4) b | < .001 |
| IUGR | Delivery only | 7.9 (98) | 1.5 | 5.6 (4.5–6.8) | 4.6 (3.7–5.6) c | < .001 |
| Stillbirth | Delivery only | 0.9 (11) | 0.5 | 1.9 (1.0–3.4) | d | .06 |
| Cesarean delivery | Delivery only | 43.3 (540) | 25.7 | 2.2 (2.0–2.5) | 2.0 (1.8–2.3) e | < .001 |
| Gestational hypertension | All admissions | 3.5 (55) | 2.3 | 1.6 (1.2–2.1) | 1.6 (1.2–2.0) f | .001 |
| Delivery only | 3.8 (47) | 2.4 | 1.6 (1.2–2.2) | 1.6 (1.2–2.1) g | .002 | |
| Preterm labor | All admissions | 13.5 (209) | 9.1 | 1.6 (1.4–1.8) | 1.4 (1.2–1.6) h | < .001 |
| Delivery only | 12.4 (155) | 6.9 | 1.9 (1.6–2.3) | 1.6 (1.4–1.9) h | < .001 | |
| Cerebrovascular disease | All admissions | <10 i | 0.06 | 10.0 (5.2–19.3) | d | < .001 |
| Delivery only | <10 | 0.03 | 8.1 (2.6–25.4) | d | .01 | |
| DVT/PE | All admissions | <10 | 0.14 | 1.9 (0.7–5.0) | d | .17 |
| Delivery only | <10 | 0.06 | 1.3 (0.2–9.1) | d | .54 | |
| Maternal mortality | All admissions | <10 | 0.02 | 4.3 (0.6–30.3) | d | .21 |
a n = 1553 for all admissions; n = 1248 for delivery only
b Adjusted for age, chronic and gestational diabetes, chronic and gestational hypertension, renal disease, multiple gestation, race, and socioeconomic status
c Adjusted for age, preeclampsia, gestational diabetes, chronic and gestational hypertension, renal disease, multiple gestation, race, and socioeconomic status
d Not enough events to perform multivariable logistic regression
e Adjusted for age, preeclampsia, chronic and gestational diabetes, chronic and gestational hypertension, renal disease, multiple gestation, race, socioeconomic status, insurance payor, hospital size, hospital type, prior cesarean delivery, and year performed
f Adjusted for chronic and gestational diabetes, multiple gestation, race, and socioeconomic status
g Adjusted for chronic and gestational diabetes, multiple gestation, and race
h Adjusted for age, preeclampsia, chronic and gestational diabetes, chronic and gestational hypertension, renal disease, multiple gestation, race, and socioeconomic status
i Because of privacy considerations, Healthcare Cost and Utilization Project policy prohibits reporting of cell sizes with fewer than 10 events.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
