Prenatal sildenafil therapy improves cardiovascular function in fetal growth restricted offspring of Dahl Salt-Sensitive rats

Abstract

Fetal growth restriction (FGR) is associated with increased risk for cardiovascular and renal disorders in later life. Prenatal sildenafil improves birth weight in FGR animal models. Whether sildenafil treatment protects against long-term cardiovascular and renal disease in these offspring is unknown. The aim of this study is to test the hypothesis that prenatal sildenafil ameliorates cardiovascular and renal function in FGR offspring of Dahl salt-sensitive rats. Sildenafil citrate (60 mg/kg per day) or control gel diet (containing 0.3% salt) was administered from gestational day ten until birth. In male and female offspring, the mean arterial pressure was measured by telemetry in one subset from week five until week twenty. Echocardiographic parameters, glomerular filtration rate, and fractional electrolyte excretion were determined in another subset at week nine. Aortic and mesenteric artery rings were prepared to assess endothelial-dependent (acetylcholine) and -independent (sodium nitroprusside) vasorelaxation (week ten). The rise in mean arterial pressure per week was attenuated in treated versus untreated male offspring. Mesenteric arteries showed an increased endothelium-dependent relaxation and improved endothelium-independent relaxation in treated versus control male offspring. No differences in aortic relaxation, echocardiographic parameters or renal function were observed between groups. Prenatal sildenafil treatment subtly improves cardiovascular but not renal function in the offspring of this FGR rat model. Translationally, in utero treatment could be beneficial for cardiovascular programming in a sex-specific manner; however, caution is warranted since recent human trials have been halted because of potentially deleterious neonatal side effects when treating pregnancies complicated with severe FGR with sildenafil.

Introduction

Fetal growth restriction (FGR) is the inability of the fetus to reach its genetically determined growth potential and is mostly caused by placental insufficiency^1,2. FGR is associated with increased risk for renal and cardiovascular disorders in later life^3,4. It is believed that this risk origins in utero when organogenesis takes place, and the adverse intra-uterine environment negatively influences (epigenetic) programming⁵. In utero treatment strategies might therefore be a window of opportunity for prevention of long-term renal and cardiovascular dysfunction from the start.

Currently no treatment for FGR exists. In recent years, sildenafil - a phospodiesterase-5 (PDE5) inhibitor - has been a drug of interest to treat FGR⁶. Hypothetically, the reduced nitric oxide (NO) – cyclic guanosine monophosphate signalling observed in pregnancies complicated with FGR could be restored by sildenafil. In addition, this vasodilatory drug is believed to improve utero-placental blood flow towards the fetus^7–15. Sildenafil has been shown to improve endothelial dysfunction in myometrial vessels from women who have pregnancies complicated by FGR and increase dilation of placental arteries obtained from women with healthy pregnancies^16,17. Previous animal studies and small human trials report that prenatal sildenafil improves pregnancy outcomes, such as an increased birth weight of 10%, in FGR pregnancies¹⁸.

However, whether maternal sildenafil treatment ameliorates long-term renal or cardiovascular disease in the offspring has barely been investigated. Only one study tested the effect of prenatal sildenafil in the endothelial NO synthase (eNOS) knock-out mouse model and showed small beneficial effects on vascular function and metabolic outcomes in adult male offspring¹⁹. Animal studies with prenatal administration of other NO-stimulating agents (molsidomine and pentaerythritol tetranitrate) demonstrated that these therapies reduce blood pressure, glomerulosclerosis, and improve endothelial function in offspring from hypertensive dams^20,21. Also, epigenetic alterations in aorta has been reported in these offspring that could explain the observed upregulation of eNOS and improved vascular reactivity²⁰. The concept that prenatal treatment with NO stimulating agents could improve long-term health by preventing adverse epigenetic programming was elegantly demonstrated in a study by Itani et al.²². They report improved cardiovascular health by prenatal sildenafil administration in an FGR chick embryo model, so the observed benefits of sildenafil treatment in FGR cannot only be dependent upon improved utero-placental blood flow. So far, no human studies have reported on long-term effects of intrapartum sildenafil treatment, and there is great anticipation awaiting the finding of the long-term outcomes in human studies, such as the STRIDER (Sildenafil TheRapy In Dismal prognosis Early-onset intrauterine growth Restriction) initiative, over several years⁶.

Although the Dahl Salt Sensitive (S) rat has been used extensively as an experimental model to study salt-sensitive hypertension, we have previously showed that it can be utilized as a model to study the mechanisms mediating superimposed preeclampsia²³. Specifically, non-pregnant females under normal-salt diet conditions develop proteinuria and hypertension that worsens during pregnancy and we have shown that prenatal treatment with sildenafil improves uteroplacental blood flow, and thereby ameliorates the superimposed preeclampsia phenotype (increase of pre-existent blood pressure and proteinuria) and rescues FGR that has been observed in this model^8,24. The objective of this study was to investigate whether prenatal sildenafil prevents renal and cardiovascular dysfunction in young adult FGR offspring of Dahl S rats. Renal function was measured with fluorescein-isothiocyanate (FITC)-inulin clearance, creatinine clearance, and fractional electrolyte excretion, and cardiovascular function was assessed with echocardiogram and speckle tracking technology, blood pressure, and vascular reactivity.

Methods

All supporting data are available within the article and its online supplementary file.

Animals

Twenty-two virgin female Dahl S rats (229 ± 4 g; 17 ± 1 weeks old) were acquired from a colony maintained by Dr. Michael Garrett at the University of Mississippi Medical Center, Jackson, MS. Rats were housed with two to three rats per cage in controlled conditions (23°C; 12–12 hour light-dark cycle). They had free access to water and normal chow (TD7034, 0.3% NaCl, Teklad, Envigo, Indianapolis IN) or, starting from gestational day ten until delivery, a gel diet made from powdered TD7034 with or without treatment. Gestational day one was determined by the presence of sperm on vaginal smears. From day 19 of gestation, dams were single housed and remained with their litter until weaning. Litters were culled to eight pups (four males and four females, if possible) on postnatal day three to standardize lactation intake between litters. Per litter, one male and one female were followed up for renal and cardiovascular measurement (renal clearance, echocardiogram, vascular reactivity) and studied at ten weeks old. In one different male and one different female per litter blood pressure was measured. The offspring used for these analyses were selected by a manually generated random number list. Three of the eleven control Dahl S litters were excluded, because all the pups died before follow-up was complete, and one of the eleven treated Dahl S litters was excluded, because of undernutrition due to insufficient lactation. All procedures included in this study were approved by the Animal Care and Use Committee at the University of Mississippi Medical Center and in accordance by the National Institutes of Health Guidelines for Use and Care of Laboratory Animals.

Study protocol

Sildenafil citrate treatment

Pregnant dams were allocated to a treatment group according to their urine protein excretion one week prior to pregnancy to create comparable groups with this accepted marker of preeclampsia. Dams received gel diet either with or without sildenafil citrate (60 mg/kg body weight per day, BIOTANG, Inc) from gestational day ten until birth. The dose was based on previous literature to achieve an optimal gain in birth weight by sildenafil^8,18. The gel diet contained all the required nutrients and was made by dissolving 242 g of powdered TD7034 chow and 6 g of agar (BD Biosciences) in 275 ml of water. We added 0.15 g of sildenafil citrate to the gel diet in the sildenafil (SILD) treated group. No specific vehicle agent was added to control (CON) gel diet.

Pregnancy and birth weight

On day 19 of gestation, dams were weighed, and a 24-hour urine sample was collected to determine urine protein excretion with Bradford assay (BioRad Laboratories)²⁵. Birth weight and litter size were determined within 24 hours after birth, and all pups were numbered for follow up. Offspring body weight was measured every two weeks until ten weeks of age.

Mean arterial pressure (MAP)

Telemetry devices (HD-S10, Data Sciences, Inc.) were implanted via the left femoral artery in one male and one female per litter under sterile conditions while under anaesthesia with isoflurane when they were five week old. Carprofen (5 mg/kg) was injected once intraperitoneally prior to surgery). We were able to implant the device in 5 male and 5 female control offspring and 8 female and 9 male sildenafil treated offspring. We excluded one control and one sildenafil treated female because the MAP read improperly in the beginning. In some of the other animals, MAP measurement was not completed through week twenty because of poor quality measurements. We did not exclude these animals, but to perform proper sex-specific analysis, we added offspring from three additional litters (with similar study protocol). MAP was measured every 3 hours for ten seconds from week five until week twenty. Per week an average was calculated for the whole day, for day-time (6:00 – 17:00) and night-time (18:00–05:00).

Echocardiography

At nine weeks of age, echocardiography was performed with Vevo 3100 (FUJIFILM VisualSonics, Inc) using a 30-Hz transducer while rats were under anesthesia (isoflurane 2.5%/1L O₂) and placed on a heating pad (37.5°C). Heart rate, breathing frequency, temperature, and electrocardiogram were monitored continuously. We obtained conventional echocardiographic parameters from four views; aortic, parasternal long axis (PSLAX), parasternal short axis (SAX) and apical 4-chamber. Speckle tracking technology was used to measure global longitudinal and circumferential strain on B-mode images to detect subclinical cardiac dysfunction. All parameters were analyzed with Vevo Lab and Strain software, followed by post analysis calculations to normalize data to body weight.

Vascular function of aorta and third-order mesenteric arteries

Aortas and third-order mesenteric arteries were isolated from ten-week-old offspring and collected in physiological saline solution (PSS: 118.3 NaCl, 4.7 KCl, 2.5 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, and 11.1 dextrose, in mmol/l). Aortic and third-order mesenteric rings (length ~2.5 mm) were mounted in a wire myograph (model 620M; Danish Myo Technology, Aarhus, Denmark). Blood vessel integrity was assessed in response to a dose of phenylephrine (Phe) [2E-6M] to induce vasoconstriction followed by vasorelaxation produced by a dose of acetylcholine (Ach) [2E-3M]. Vessel segments were washed with PSS, constricted with Phe, and cumulative concentration-response curves were generated for Ach [1E-10M to 3E-5M] (Sigma). Segments were washed again with PSS and response curves generated for sodium nitroprusside (SNP) [1E-10M to 3E-5M] (Sigma). These curves allow the study of endothelial-dependent and -independent vasorelaxation, respectively.

Renal function measurement

24-hour urine samples were collected at ten weeks. Sodium azide (Fisher Chemical, Inc) was added to prevent bacterial growth. Assays were run to determine proteinuria (Bradford Assay, BioRad Laboratories), urea (DiaSys), sodium (Na⁺) and potassium (K⁺, Sherwood Scientific), and creatinine (DiaSys). Fractional excretion (FE) of Na⁺ and K⁺ was calculated with the following formula: FE_electrolye (%) = 100 x (urine_electrolyte (mM) x plasma_creatinine(mg/dl)) / (plasma_electrolye (mM) x urine_creatinine (mg/dl)). To calculate fractional Na and K excretion, plasma levels were set to 140 and 4 mM respectively. Creatinine clearance (C_Cr) was calculated as follows: C_Cr (ml/min) = (urine_creatinine (mg/ml) x 24-hour urine volume (ml))/(plasma_creatinine (mg/ml) x 1440 (min)).

Glomerular filtration rate (GFR)

At ten weeks of age, glomerular filtration rate (GFR) was measured by transcutaneous method and device from MediBeacon (Mannheim, Germany)²⁶. One or two days before measurement, a catheter was inserted in the right jugular artery while the rat was under isoflurane anaesthesia. The catheters were accessible at the nape and filled with sterile 3% heparin-saline. At the day of measurement, hair on the nape was shaved, and the device was placed and immobilized by a harness (Instech Lab) under isoflurane anaesthesia. A bolus of FITC-sinistrin (35 mg/ml) was administered through the jugular catheter. The half-life (t_½) of FITC-sinistrin clearance by the kidneys was measured transcutaneously while the rats were conscious, by which the GFR was calculated with the following formula: GFR (ml/min/100 g BW) = [31.26 (ml/100 g BW)/t_1/2 of FITC-s (min)] x BW as described²⁶.

Histology

Paraffin embedded longitudinal sections of the left kidney and a midsection of the heart from ten-week-old offspring were placed in buffered in 10% formalin for 48 hours, after which they were immersed in 70% ethanol until embedment in paraffin. The sections were stained with Haematoxylin and Eosin (H&E) and Periodic acid-Schiff (PAS). The investigator who scored the histology slides was blinded to the treatment. In the heart, we investigated whether fibrotic signs were present, and in the kidney, we assessed interstitial fibrosis and glomerulosclerosis.

Statistical analysis

Statistical analysis was performed with IBM SPSS Statistics for Windows, version 25 (IBM Corp., Armonk, NY). Differences between treatments in dams were tested with an independent t-test. For offspring data, a 2-way analysis of variance (ANOVA) was performed for the effects treatment, sex, and their interaction, followed by Bonferroni’s multiple comparison post-hoc test. Maximal effective concentration (E_max), sensitivity (logEC₅₀) and area under the curve (AUC) were calculated with GraphPad Prism version 7.04 and subsequently tested with 2-way ANOVA. We used a linear mixed-model statistical analysis for telemetry data with individuals as a random effect nested within the groups and compound symmetry as (repeated) covariance type²⁷. The interaction between treatment group and weeks as repeated interval was modelled. All data are expressed as mean ± standard error of the mean (s.e.m.) unless otherwise specified. A two-sided p-value < 0.05 was considered significant.

Results

Pregnancy outcomes

Urine protein excretion at gestational day 19 did not significantly differ in the treated versus control dams (Table 1). Average birth weight per litter was significantly higher in the treated group compared to the non-treated group, while litter size did not change (Table 1).

Table 1:

Pregnancy outcomes.

Pregnancy outcome	CON (n=8)	SILD (n=10)	P-value
UPE (mg/day) at GD19	75.0 ± 43.8	63.5 ± 21.8	0.13
Birth weight combined sex (g)	6.23 ± 0.05	6.83 ± 0.18	0.02
Birth weight females (g)	6.16 ± 0.10	6.78 ± 0.17	0.02
Birth weight males (g)	6.39 ± 0.07	6.96 ± 0.20	0.04
Total litter size	9.3 ± 0.8	10.9 ± 1.0	0.23
Viable litter size (PD 3)	7.9 ± 1.0	9.6 ± 0.9	0.23

Mean ± s.e.m. CON, control; GD, gestational day; PD, postnatal day; SILD, sildenafil; UPE, urine protein excretion. Three CON and one SILD litters were excluded due to lost to follow up of the offspring.

MAP in offspring

MAP was similar between groups at the beginning, but diverged, as the animals aged; treated animals have a lower MAP progression calculated via slope (r) from week five to twenty compared to untreated animals (Fig 1 and Table 2). This effect was only significant in male offspring. The observed improvement in 24-hour MAP progression per week in combined sexes and in male offspring appeared to be due to the difference observed during day-time only.

Figure 1: Mean arterial pressure in five-till-twenty-week-old offspring.

Mean arterial pressure (MAP) during day- and night-time per week from week 5-20 in sildenafil (SILD) and control (CON) treated A) male (n=12 vs 8) and female (n=10 vs 7) offspring combined; B) in male offspring only; C) in female offspring only. Significant differences in slope per week between prenatal sildenafil (solid line) and control (dotted line) gel diet is indicated in the panels. Mean ± s.e.m.

Table 2:

MAP progression per week from week five until week twenty.

Sex	Parameter	Group	Intercept	Pintercept	r	Pr
All	MAP	CON	97.1 (90.2; 104.0)	0.60	2.4 (2.0; 2.8)	0.03
		SILD	98.2 (95.6; 100.8)		2.1 (2.0; 2.3)
	MAPday-time	CON	92.1 (85.5; 98.8)	0.13	2.6 (2.2; 3.0)	0.001
		SILD	95.4 (92.9; 97.9)		2.2 (2.0; 2.3)
	MAPnight-time	CON	102.8 (94.7; 111.0)	0.43	2.2 (1.7; 2.7)	0.88
		SILD	100.8 (97.7; 103.9)		2.2 (2.0; 2.3)
Male	MAP	CON	95.3 (84.8; 105.8)	0.68	2.6 (2.0; 3.2)	0.02
		SILD	96.7 (92.8; 100.6)		2.1 (1.9; 2.3)
	MAPday-time	CON	90.4 (80.6; 100.3)	0.25	2.8 (2.1; 3.5)	0.002
		SILD	94.1 (90.5; 97.7)		2.1 (1.9; 2.3)
	MAPnight-time	CON	102.1 (89.3; 114.9)	0.45	2.3 (1.5; 3.1)	0.67
		SILD	99.0 (94.3; 103.7)		2.2 (1.9; 2.5)
Female	MAP	CON	98.0 (89.3; 106.8)	0.46	2.3 (1.8; 2.7)	0.34
		SILD	100.0 (96.6; 100.4)		2.2 (2.0; 2.3)
	MAPday-time	CON	93.1 (84.3; 102.0)	0.17	2.5 (2.0; 3.0)	0.11
		SILD	96.8 (93.4; 100.2)		2.2 (2.0; 2.4)
	MAPnight-time	CON	102.9 (93.0; 113.1)	0.98	2.1 (1.9; 2.4)	0.89
		SILD	103.0 (99.1; 106.9)		2.1 (1.6; 2.0)

Mean arterial pressure (MAP) progression in prenatal sildenafil treated (SILD; n=22) and control (CON; n=15) offspring for combined sexes and separated for male (SILD: n=12 vs CON: n=8) and female (SILD: n=10 vs CON: n=7) offspring as analyzed with linear mixed model. Mean (95% confidence interval). The intercept is at five weeks of age. r = slope per week, from week five until twenty.

Echocardiography in offspring

In the parameters derived from the parasternal long axis (PSLAX) and parasternal short axis (SAX), the view differed between male and female offspring, regardless of treatment (supplemental Table S1). The area of the left ventricle was significantly smaller in the treatment group (independent of sex) or between sex only, but this treatment effect disappeared when sex was taken into consideration. Prenatal treatment with sildenafil resulted in an increased stiffness in the posterior base region in female offspring, while this was decreased in male offspring.

Vascular function in ten-week-old offspring

There were no differences in maximum (E_max), total (AUC), or sensitivity (-logEC₅₀) to the vasorelaxation induced by Ach or SNP in the aorta among treatment groups (independent of sex) or between sex only (Fig S1 and Table 3). However, in the mesenteric arteries, the lower logEC₅₀ and greater AUC indicated an increased endothelium-dependent relaxation of the sildenafil treated versus control groups (independent of sex). In addition, the mesenteric arteries showed an improved endothelium-independent vasorelaxation indicated by AUC in male offspring compared to female offspring (interaction treatment*sex) (Fig 2 and Table 3).

Table 3:

Vasorelaxation in aorta and SMA in ten-week-old offspring.

Vessel test	Paramet er	Male		Female		p-value
		CON (n=7)	SILD (n=10)	CON (n=7)	SILD (n=10)	Treatment	Sex	Treatment * sex
Aorta Ach	Emax, %	90.5 ± 2.7	90.1 ± 3.4	92.6 ± 2.2	93.8 ± 2.9	0.90	0.34	0.80
	logEC50, M	−6.9 ± 0.4	−6.8 ± 0.6	−6.9 ± 0.5	−6.5 ± 0.3	0.61	0.72	0.80
	AUC	232 ± 29	247 ± 49	266 ± 27	216 ± 24	0.61	0.98	0.35
Aorta SNP	Emax	97.4 ± 1.2	98.3 ± 3.2	90.7 ± 3.4	98.9 ± 2.1	0.12	0.28	0.21
	logEC50, M	−8.1 ± 0.3	−7.9 ± 0.4	−8.2 ± 0.3	−8.3 ± 0.2	0.84	0.40	0.60
	AUC	344 ± 27	351 ± 33	347 ± 25	372 ± 17	0.55	0.66	0.74
SMAAch	Emax, %	101.7 ± 6.0	99.3 ± 3.0	103.0 ± 2.3	100.5 ± 0.8	0.48	0.71	0.99
	logEC50, M	−8.0 ± 0.2	−9.3 ± 0.4	−7.6 ± 0.5	−8.2 ± 0.3	0.01	0.06	0.42
	AUC	345 ± 20	449 ± 25	355 ± 35	365 ± 23	0.04	0.19	0.09
SMASNP	Emax, %	99.8 ± 0.3	100.0 ± 0.7	100.6 ± 2.8	96.2 ± 3.3	0.38	0.52	0.33
	logEC50, M	−7.9 ± 0.2	−8.9 ± 0.5	−8.5 ± 0.4	−8.2 ± 0.1	0.31	0.79	0.09
	AUC	358 ± 14	415 ± 32	410 ± 25	344 ± 8	0.87	0.71	0.03

Ach, acetylcholine; AUC, area under the curve; CON, control; E_max, maximum relaxation at 10^−4.5; logEC₅₀, sensitivity in M; SILD, sildenafil; SMA, third-order (small) mesenteric arteries; SNP, sodium nitroprusside. Mean ± s.e.m.

Figure 2: Vasorelaxation in third-order mesenteric arteries in ten-week-old offspring.

Vasorelaxation curves to cumulative concentrations of acetylcholine (Ach) in third-order mesenteric arteries from prenatal sildenafil (SILD) and control (CON) treated A) male (n=10 vs n=7) and B) female (n=10 vs n=7) offspring. Vasorelaxation curves to cumulative concentrations of sodium nitroprusside (SNP) in treated and untreated C) male (n=10 vs n=7) and D) female (n=10 vs n=7) offspring. Mean ± s.e.m. Phe, phenylephrine.

GFR, proteinuria, urea, and fractional electrolyte excretions in ten-week old offspring

The GFR expressed per 100 g body weight was higher in males compared to females, but no effect of the treatment was observed (supplemental table S2). Most of the plasma or urine laboratory parameters for renal function were higher in males compared to females, as would be expected because of their greater muscle mass (creatinine excretion) and food intake (K excretion).

Heart and kidney morphology in of ten-week-old offspring

Body weight was much lower and relative heart weight was higher in females compared to males (Table S3 and Fig. S2). Body weight and relative organ weight per body weight was similar between treated and untreated animals. The midsections of the heart did not show any signs of fibrosis. In the kidney, we also found practically no (less than 5%) interstitial fibrosis or glomerulosclerosis with no differences among groups (Fig S3).

Discussion

Prenatal administration of sildenafil increases birth weight in our FGR model in Dahl S rats. Our main finding is that in utero treatment with sildenafil dampens the development of high blood pressure in later life of male offspring. In addition, the mesenteric arteries showed increased and sex-independent endothelium-dependent relaxation in treated versus control offspring and improved endothelium-independent relaxation in treated male offspring. While prenatal treatment with sildenafil had the above-mentioned beneficial effects on cardiovascular health, it did not significantly affect renal function, and no differences in echocardiographic parameters were observed between treated and control young adult Dahl S offspring.

Previous studies investigated the effect of maternal sildenafil on outcomes during pregnancy, such as birth weight and maternal blood pressure¹⁸. A meta-analysis showed that there was an overall beneficial effect on birth weight in preeclampsia and/or FGR models, including the Dahl S rat^8,18. To the best of our knowledge, we are the first to investigate the effect of prenatal sildenafil administration on long-term renal function in male and female offspring of a spontaneous FGR animal model. Only one study investigated the effect of maternal sildenafil on long-term cardiovascular function in male offspring of the eNOS knock-out mouse model¹⁹. While they were unable to confirm that sildenafil could improve fetal growth and sex-specific difference might have been missed, they found small beneficial effects on endothelial-dependent relaxation of second-order mesenteric arteries and no effect on blood pressure in adulthood. In a chick embryo model of FGR, an improvement of cardiovascular function was reported²².

The beneficial effect of sildenafil on the development of hypertension may be due to several mechanisms. The most likely explanation is the subtle improvement of vascular sensitivity, as we found that prenatal sildenafil enhanced dilatation in resistance arteries: endothelium-dependent relaxation was enhanced in third-order mesenteric arteries of treated rats in both males and females, and endothelium-independent relaxation was enhanced in treated male offspring compared to treated female offspring. This finding agrees with the potential involvement of the NO pathway that sildenafil targets. Itani et al. confirmed that sildenafil improves endothelial-dependent vascular relaxation in a NO dependent manner²². No differences in echocardiographic parameters were found in our study that could explain the positive effect on blood pressure.

While the vascular reactivity data presented here might partly explain the lower blood pressure in male offspring of sildenafil-treated dams, we speculate that the underlying mechanism might also partly be attributable to the reversal of adverse epigenetic programming. Wu et al. linked epigenetic alterations in aorta of prenatal NO donor treated offspring, including enhanced histone 3 lysine 27 acetylation and histone 3 lysine 4 trimethylation, that could explain the observed upregulation of endothelial NO synthase and improved vascular reactivity²⁰. It is quite likely that such an alteration is systemic. If this is indeed the case, it will also be present in resistance arteries, and therefore could have contributed to our findings. Future studies are planned to investigate these potential mechanisms, with a focus on potential changes in target organs (vasculature, kidney and heart) to connect the epigenetic patterns to the clinical outcomes. As part of reversed programming hypothesis, the prenatal treated offspring might be relatively protected to a second hit (such as high salt intake).

We believe that one of the strengths on this study was the investigation of long-term effects of prenatal sildenafil in an mammalian FGR model. This is valuable information, especially in the face of human trials having been initiated without knowledge of long-term effects. Another strength is that we used the Dahl S rat, which is one of the better models for FGR offspring studies, with the origin of FGR lying in spontaneous placental insufficiency, and not surgically- or pharmacologically-induced FGR.

Sex-specific differences in both fetal/neonatal and long-term health have been observed in numerous clinical and preclinical studies investigating the developmental-origin-of-disease hypothesis²⁸. Preclinical FGR studies often report hypertension development in male offspring only^28–30. Male FGR offspring could therefore benefit more or earlier from prenatal interventions. Several underlying mechanisms for the sex-specific difference in blood pressure observed in FGR offspring have been suggested, such as differences in placental development regulation, renin-angiotensin system regulation mediated by testosterone or renal oxidative stress, and epigenetic programming^28–30.

We acknowledge some limitations. We only followed the animals up until a relatively young adult age. The subtle differences that we observed might have diverged more when the offspring aged more or when a second hit occurred at older age. For instance, the cardiac ultrasound measurements were performed prior to the manifestation of blood pressure differences between treated and control rats. Secondly, we did not investigate what the most optimal time window of treatment would be. Thirdly, we used a higher dose of sildenafil than human clinical trials, because a dose-response curve based on previous literature suggested that a higher dose is needed to gain optimal effect¹⁸. However, sildenafil has not yet been given in this dose in humans, and therefore translation to clinical practice needs further exploration.

Implementation of in utero treatment with PDE5-inhibitors into the clinical practice still faces several challenges. As mentioned, this beneficial experimental dose exceeds the dose that has been used in clinical practice. Most importantly, the Dutch STRIDER initiative was abruptly halted due to potential adverse effects in combination with the lack of beneficial effects of sildenafil on FGR neonates (see perspectives)³¹. However, our study shows that targeting the NO pathway is still an interesting therapeutic option to prevent adult diseases from the start. Therefore, we would recommend future studies to focus on development of modified PDE5-inhibitors or NO donors that do not cross the placenta in order to avoid deleterious fetal side effects as such agents currently do not exist.

In conclusion, prenatal sildenafil subtly ameliorates long-term cardiovascular health in our FGR young adult offspring in a sex-dependent manner. No harmful effects were observed in the kidneys or cardiovascular system of the offspring. While in utero treatment with agents that increase NO bioavailability may be a strategy to prevent long-term health problems in FGR offspring, more research is needed to develop a safe administration strategy to avert potential neonatal side effects.

Perspectives

FGR is associated with increased risk for cardiovascular diseases in adulthood. Cardiovascular disease is the most common cause of death worldwide. In utero treatment with sildenafil (or other PDE-5 inhibitors) could prevent long-term cardiovascular health problems in FGR offspring as our study shows that it reduces the development of hypertension (especially in males). The implementation of sildenafil in clinical practice still faces challenges. Even though prenatal sildenafil does appear to have beneficial cardiovascular effects, and we did not observe any overt harmful effect on the offspring’s health, we acknowledge that great caution must be exercised. The Dutch STRIDER trial - a randomised control trial in which women whose pregnancies were complicated by severe early-onset FGR received either sildenafil or placebo - has recently stopped after an interim analysis showed a potential associated risk of neonatal death (in combination with futility)³¹. These adverse effects have not been reported in other studies of prenatal administration of sildenafil, including the other completed branches of the STRIDER trial in the UK and New-Zealand³². Until the interim results of the Dutch Strider trial are further analysed and clarified, we do not recommend administrating sildenafil to pregnant women. Nevertheless, the results in the present study, do not, as yet, completely rule out the potential of in utero treatment with NO stimulating agents in preventing diseases in adult life. Our study shows that, in principal, targeting the NO pathway could be an effective therapeutic strategy. Future studies should develop safer therapies; unknown toxic or adverse effects in neonatal offspring could for instance be prevented by administering promising drugs that are modified so that they do not cross the placenta³³.

Novelty and significance.

What is new?

• This is the first study to investigate the long-term effect of prenatal sildenafil on cardiovascular and renal function in both male and female offspring of a spontaneous model of fetal growth restriction.

What is relevant?

• We confirm that prenatal sildenafil improves birth weight in Dahl S rats as a model for FGR.

• Prenatal sildenafil results in lower blood pressure in male offspring.

• No harmful effects of prenatal sildenafil on the renal or cardiovascular system were observed.

Summary.

This study shows that prenatal sildenafil lowers the progression of hypertension development in fetal growth restricted rat offspring in a sex-dependent manner.