Interpregnancy interval and abnormally invasive placentation

Abstract

Introduction:

The presence of a previous uterine scar is a strong risk factor for developing abnormally invasive placentation (AIP). We sought to determine whether a short interpregnancy interval predisposes to AIP. We hypothesized that a short interpregnancy interval after a previous cesarean delivery increases the risk of AIP in comparison with a longer interpregnancy interval.

Material and methods:

We performed a retrospective cohort study of women with a histological diagnosis of AIP and a history of a previous cesarean section. Women were included in the control group if they had a previous cesarean section with a placenta underlying the previous uterine scar or an anterior previa. The time interval between pregnancy and AIP data was analyzed using the chi-square test and two-tailed Fisher’s exact test.

Results:

There was no statistical difference in the interpregnancy interval between women who had AIP vs the control group. Gravidity and parity were found to be significantly higher in the women with AIP vs the controls.

Conclusions:

These results suggest that a short interpregnancy interval may not increase the risk of developing AIP.

1 ∣. INTRODUCTION

Abnormally invasive placentation (AIP), which encompasses placenta accreta, increta and percreta, is a life-threatening complication of pregnancy that has been steadily increasing in incidence over the past 40 years, with current estimates at one in 533 pregnancies.¹ Placenta accreta occurs by trophoblastic invasion into a layer of the decidua basalis, the Nitabuch layer.² Histopathologically, a placenta accreta will show a direct apposition of villous tissue into the myometrium without intervening decidua.³ An increta occurs when the villous tissue invades the myometrium, and a percreta occurs when the villous tissue and abnormal vessels invade through the myometrium into the serosal layer.⁴

The strongest risk factors identified for AIP include the number of prior cesarean sections (CS) and having a placenta previa;^5,6 however, other known risk factors include advanced maternal age, in vitro fertilization pregnancy, history of postpartum hemorrhage and other prior uterine surgeries including dilation and curettage, and hysteroscopic surgery for uterine anomalies.^5,6 History of prior CS or uterine instrumentation/manipulation can lead to a functional paucity of decidua, which is one of the proposed etiologies of AIP.

Repeat hysterotomy incisions can cause fibrosis and scarring of the uterine wall, leading to decreased vascularity and poor wound healing.⁷ In most women, the scar will be well-healed, but in others there is presence of a uterine wall defict.⁷ This has been corroborated by pathologic evaluation of the uterine scar, suggesting granulation tissue and fibrosis formation rather than regeneration of the myometrium.⁸

In a normally developing embryo, the proliferating trophoblastic tissues plug the lumina of the decidual vessels, which leads to a physiologic hypoxia. This local low oxygen level stimulates cytotrophoblastic invasion into the endometrium,⁹ secondary to a persistence of the invasive phenotype of the cytotrophoblastic cells rather than the proliferative phenotype. For that reason, the avascular and acellular area of a fibrotic CS scar can actually be a preferential location of the developing embryo, which is another proposed etiology of AIP.

The surface area of the scarring increases after multiple CS. This ultimately leads to a greater chance of implantation to that area.⁷ The risk of AIP in one study showed a 7-fold increased risk after one prior CS to a 56-fold increased risk after three or more CS.⁶

Inadequate healing of the previous hysterotomy incision after a short interpregnancy interval of less than 18 months can lead to a higher rate of uterine rupture during a trial of labor vs intervals of 24 months and longer (4.8% vs 1.3%, respectively).⁸ Because of the strong association between the presence of a uterine scar and risk of AIP, we sought to determine whether a short interpregnancy interval, which has been shown to be associated with inadequate healing of the previous uterine scar, predisposes to AIP if the placenta implants over the previous scar. We hypothesized that a short interpregnancy interval in women with a history of previous cesarean delivery would increase the risk of AIP in comparison with a long interpregnancy interval.

2 ∣. MATERIAL AND METHODS

Women met the inclusion criteria of the study group if they had a histologically proven diagnosis of AIP between 2006-2017 (placenta accreta, increta or percreta) and a history of a previous CS. These subjects were identified via records from the Center for Abnormal Placentation at Hackensack Meridian Health-Hackensack University Medical Center, a multidisciplinary unit established to care for these complex cases. Cases and controls were identified by performing an electronic query from the ultrasound reporting system of the maternal fetal medicine department using the keywords “anterior placenta”, “placenta previa”, “previous C-section scar”, and from the Pathology department records using keywords “cesarean hysterectomy”, “accreta”, “increta”, “percreta”, “placenta previa”, “adherent placenta”. Demographics including maternal age, gravidity, parity, number and date of previous CS, last menstrual period, and prepregnancy weight and height were abstracted from medical records. Women were included in the control group if they had a prior CS with a placenta previa where the anterior component of the placenta was underlying the previous uterine scar, or an anterior placenta underlying the previous uterine scar.

The women in the control group had no clinical or histological diagnosis of AIP. The resulting sample size was 73 AIP subjects and 88 control subjects. Six and two extreme outliers (interpregnancy interval >2 SD above the mean for the group) in case and control groups were excluded, leaving sample sizes of 67 AIP and 86 control for analysis. We define interpregnancy interval as the date of previous cesarean delivery to the estimated date of conception of the affected or control pregnancy. To determine the estimated date of conception, 14 days were added to the calculated or known last menstrual period. Our primary outcome was the presence of AIP.

2.1 ∣. Statistical analyses

The women were stratified based on their interpregnancy interval, expressed in months. For yearly intervals, the data were analyzed using a chi-square test. Intervals of 18 months were analyzed using a two-tailed Fisher’s exact test. Interpregnancy interval reported as a continuous variable was not normally distributed and was therefore tested using the Mann–Whitney U test.

The subgroup with placenta previa and the subgroup with an anterior placenta underlying previous CS scar were combined into a single control group, as no statistical difference was found in any demographic categories or interpregnancy intervals.

Prior to analysis, normality of the demographic data for both the control and AIP group was determined utilizing the D’Agnostino–Pearson normality test. Maternal age was the only demographic category which passed normality for both our control and study group and was, therefore, analyzed using an unpaired, two-tailed t test with Welch’s correction. Gravidity, parity, number of prior CS and prepregnancy body mass index did not pass normality and were analyzed using the two-tailed Mann–Whitney U test. P < .5 were considered statistically significant. Data analysis was performed using GraphPad PRISM 7.03. The mean and standard deviation of the previa control and AIP interpregnancy interval were used to perform an a priori power analysis using a Wilcoxon–Mann–Whitney test within G*POWER 3.1 with an effective size of .21.

2.2 ∣. Ethical approval

Institutional Review Board approval was obtained for this retrospective study: reference number Pro00001951 (2008).

3 ∣. RESULTS

The final sample size for this retrospective study included 67 women in the study group and 86 women in the control group. The interpregnancy interval for each subject is expressed in months, and subsequently graphed with the mean and standard deviation noted (Figure 1). The composite interpregnancy interval (expressed in months) of each group was compared and showed no significant difference (Table 1). An interpregnancy interval of less than 18 vs 18 months and greater was used to stratify within the two groups. There was no statistical difference in the number of women with a short interpregnancy interval (less than 18 months) who had AIP vs women with a short interpregnancy interval who did not have AIP. Similar results were obtained for interpregnancy intervals >18 months. We further stratified each group into pregnancy intervals of <1, 1-2 and >2 years. Again, there was no statistically significant difference in the number of women with AIP vs controls in any group. Interpregnancy interval also showed no difference when stratifying the AIP cases by accreta and severe AIP, increta and percreta (P = .3354). Accretas (n = 21), the least invasive of the AIP diagnosed, had an interpregnancy interval median of 26 months (interquartile range 18-43 and full range 4-68). Incretas (n = 6) and percretas (n = 26) had an interpregnancy interval of 32 months (interquartile range 21-52 and full range 3-119).

FIGURE 1.

The interpregnancy interval for each subject expressed in months, including mean and standard deviation

TABLE 1.

Composite and stratified analysis of interpregnancy interval association of abnormally invasive placenta compared with placenta previa control. Composite interpregnancy intervals are expressed as median, [interquartile range] and (full range). Stratified interpregnancy intervals are expressed as count and (percent proportion)

Interpregnancy interval	Placenta previa control (n = 86)	Abnormally invasive placenta (AIP) (n = 67)	P value
Composite interpregnancy interval (mo)	29 [17-58] (3-173)	30 [19-47] (3-102)	0.8841
<18 mo	23 (26.7%)	19 (28.4%)	0.8565
≥18 mo	63 (73.3%)	48 (71.6%)
<1 yr	12 (14.0%)	70 (1.4%)	0.7722
1-2 yr	24 (27.9%)	18 (26.9%)
≥2 yr	50 (58.1%)	42 (62.7%)

Demographics were analyzed and revealed that gravidity and parity were significantly higher in the women with AIP than in the controls (Table 2).

TABLE 2.

Analysis of demographics. Displayed as the mean ± SD with the range of the data

Characteristic	Placenta previa control (n = 86)	Abnormally invasive placenta (AIP) (n = 67)	P value
Maternal age (yr)	34.1 ± 4.7	33.3 ± 5.7	0.3403
Gravidity	3 ± 2 (2-20)	5 ± 3 (2-14)	<0.0001
Parity	2 ± 2 (1-15)	3 ± 2 (1-11)	0.0006
Cesarean section	1 ± 1 (1-3)	2 ± 1 (1-5)	0.0031
Prepregnancy mass index	26.5 ± 5.5 (18.6-45.3)	26.7 ± 5.8 (17.2-42.0)	0.6592

4 ∣. DISCUSSION

In our cohort of women, we found there is no significant difference in the interpregnancy interval between women with AIP and the control group. These results suggest that interpregnancy interval is not a factor that increases the risk of developing AIP in women with a previous cesarean delivery. Our demographic data analysis revealed that women with AIP have a higher mean number of CS, which corroborates previously published data⁵ (P = 0.0031). Mean gravidity and parity were also significantly increased in the women with AIP compared with those in the control group, regardless of the number of cesarean deliveries (Table 2, P ≤0.0001, P = 0.0006, respectively).

The association between short interpregnancy interval and increased risk of inadequate healing of the uterine scar has been concluded based on studies that show increased risk of uterine rupture during trial of labor after cesarean delivery when there is less than 18 months between cesarean delivery and conception compared with interpregnancy intervals of 24 months and longer (4.8% vs 1.3%, respectively).⁸ Building on this, we chose to stratify our sample by 18-month intervals and even further stratified into 1-year intervals to better elucidate whether there is an interpregnancy interval that does put women more at risk of abnormal placentation; none of these showed a significant difference within our cohort of women.

There are several key differences between our study and other studies that assessed the interpregnancy interval and the presence of AIP. We identified matched control subjects according to location of the placenta. While questioning the integrity of a previous uterine scar on the development of invasive placentation, we wanted subjects whose placentation developed in a similar environment: under the previous uterine scar. This included women with a complete previa (64% of control group), as well as those with an anterior placenta underlying the previous uterine scar (36% of control group). Our study did not include women with a posterior placenta previa. To our knowledge, there have been three other studies for which similar data were analyzed. In a study conducted by Sumigama et al, one secondary outcome was the interpregnancy interval on risk of placenta accreta.¹⁰ The conclusion was that there was no significant difference in interpregnancy interval and risk of AIP; however, these women were not matched by location of placenta and it is therefore difficult to interpret these results.¹⁰ Another study by Wax et al, which was underpowered, sought to determine cesarean to conception interval and the effect on development of abnormally adherent placenta.¹¹They concluded that cesarean to conception mean intervals are shorter in women with abnormally adherent placentas (35 months) vs controls (48 months); however, this was not statistically significant. The study by Bowman et al found no difference in the interpregnancy interval of women with accreta compared with the interpregnancy interval of those without.¹² The placental location for this control group was not specified.¹² In that study, it was also unclear whether women were matched by the location of their placenta previa, whether the location was marginal or central and whether these women included posterior placenta previa. Another study by To et al which sought to determine the association between previous cesarean delivery and development of placenta previa with accreta looked at interpregnancy intervals; however, their control group was composed of women with a placenta previa and no prior cesarean delivery, therefore no cesarean to conception interval was determined.¹³ In our present study, there was no significant difference in the mean interpregnancy interval (expressed in months) in our subjects vs controls.

One of the strengths of our study was that our control and study groups were evenly matched. In our study, we focused on the specific placental location, since this important variable has not been previously reported. The limitations of our study include its retrospective design and that in this type of study, a power analysis cannot be done. However, using a priori power analysis of the monthly interpregnancy interval mean and standard deviation of placenta previa control and AIP cases, 95% power could be achieved with a total sample size of 1208 (604 controls and 604 AIP cases). A larger, multi-centered prospective study would have to be done to identify any other differences. Other risk factors could be evaluated in future studies such as CS performed electively or during labor. Prior to making any changes to recommendations on appropriate interpregnancy interval, further studies should be conducted. As always, performing a cesarean delivery is at the discretion of the provider and should be done on appropriate indications only.

5 ∣. CONCLUSION

The results of this study suggest that a short interpregnancy interval in patients with a previous cesarean delivery may not be associated with an increased risk of abnormally invasive placentation.

Key Message.

A short interpregnancy interval may not increase the risk of developing abnormally invasive placentation. Gravidity and parity were found to be significantly higher in the women with abnormally invasive placentation vs the controls.

Abbreviations:

AIP

abnormally invasive placentation

CS

cesarean section.