Primate maternal placental angiography
Article Outline
Abstract
Background
In humans, it is known that blood flow is directed to the gravid uterus from two (right and left) pelvic uterine arteries. The extent of supply from the tubo-ovarian anastomosis (joining of the ovarian and uterine arteries) is unknown. The aim of this study was to delineate the arterial blood supply to the placenta via systematic angiography in normal pregnancies in a non-human primate, the baboon (Papio hamadryas).
Methods
The assessment of the distribution of blood supply with single-shot 3-vessel angiography (aorta, right and left common iliac arteries), allowed assessment of bilateral supply and possible ovarian supply (n
=9). In 2-vessel pictures (aorta and left or right iliac), the contralateral supply was determined by subtraction of the ipsilateral supply from the total supply (n=7). The studies were all approved by the Institutional animal welfare committee and were conducted as part of a broader project investigating preeclampsia.
Results
The animals were 9 years of age and 140 days of gestation for the 3 vessel study and 154 days of gestation for the 2 vessel study. The angiograms were more likely to have cotyledons perfused by the left uterine artery (p=0.012) than the right. Overall, 55% of placentae had 5–44% of supply overlapping and 22% had 10–15% ovarian contribution to blood supply.
Discussion
This study demonstrates the variation in primate uteroplacental blood flow including the contribution of ovarian arteries and left and right collateralization. Similarity to human vascular anatomy strengthens the use of primate species as a model of human placentation.
Keywords: Primate, Placenta, Cotyledons, Blood flow, Angiography
1. Introduction
In humans, it is known that blood flow is directed to the gravid uterus from two (right and left) pelvic uterine arteries. Collateral supply from the tubo-ovarian anastomosis (joining of the ovarian and uterine arteries) is thought to be noncontributory except for pathological situations associated with uterine leiomyomata [1], [2]. The contribution of the ovarian arteries to placental blood flow is unknown.
Like humans, Papio hamadryas has a single discoid, haemomonochorial placenta [3], [4], [5] with predominantly singleton pregnancies (twinning rate 1/200). Baboons are used for ethical reasons for pregnancy-related research due to the similarity with human placental structure, and partly on the basis of presumed identical uterine blood supply characteristics, i.e. bilateral pelvic vessels with negligible ovarian contribution. No angiography studies have determined the possible ovarian artery contribution to uterine blood flow in either baboon or human pregnancy.
In the baboon, delineation of the blood supply depicts the left and right common iliac arteries coming off the aorta as in the human. Similarly, branches of the internal iliac artery (hypogastric artery) have both an anterior division and a posterior division. The posterior branches include the superior gluteal artery and anterior divisions including the inferior gluteal artery, the internal pudendal artery and the predominant supplier to the placenta, the uterine artery.
The aim of this study was to delineate the arterial blood supply to the placenta via systematic angiography in normal pregnancies in a non-human primate, the baboon (P. hamadryas). An analysis of angiograms performed for research into preeclampsia (a human condition caused by aberrant placental blood flow) was undertaken. The study explored the symmetry, the shared supply from the left and right uterine arteries, and also any blood flow unaccounted for by uterine vessels that originates from the ovarian arteries.
2. Methods
2.1. Pregnant animal selection
All P. hamadryas involved in the analysis were from the National Baboon Colony, NSW, Australia. These baboons have been out-bred in captivity. All animals were of reproductive age and part of the same basic social and reproductive unit. The animals were healthy and had been regularly screened for infectious diseases such as tuberculosis, and intestinal parasites. Animal age, gestation at angiography, prior surgical delivery, parity and fetal weight were recorded. All angiography was performed in normal pregnancies.
The experiments which lead to this collection of angiograms (1992–2006) was approved by the Sydney South Western Area Health services (formerly Central Sydney Area Health Service (CSAHS) Animal Welfare Committee and abided by the National Health and Medical Research Council's (NH&MRCs) Australian Code of Practice for the care and Use of Animals for Scientific Purposes and the Australian Code of Practice for the Use of Primates in Scientific Research. The primate studies were made possible by the support from the NH&MRC for the National Baboon Colony.
Placental angiograms from a number of baboons have been conducted as a part of other research projects [6], [7], [8]. The angiograms depicted arterial flow from the mid abdominal aorta and all the subsequent branches. Following the aortograms, selective right and/or left iliac angiograms were undertaken. Total blood supply from aortic arteries was quantified by a count of cotyledons. Cotyledons for the purposes of this study were classified as the well-defined ‘blush’ of contrast generated by each spiral vessel supplying the placenta.
The assessment of the distribution of blood supply with 3 vessel angiography (aorta, right and left common iliac arteries), allowed assessment of bilateral supply and possible ovarian supply (n=9). Ovarian supply was defined as the number of cotyledons supplied by the aortic angiogram and not seen on either the right or left angiogram. In the 2 vessel pictures (aorta and left or right common iliac arteries), the contralateral supply was determined by subtraction of the ipsilateral supply from the total supply (n=7).
The procedures for angiography are published elsewhere [6], [7], [8] and are summarized here. Angiography was performed between weeks 17–24 of normal gestation, i.e. near the end of pregnancy when the placental circulation would be fully established and when there would be little ability for the established placental circulation to re-model. All procedures involved sedative anesthesia with ketamine infusion. Analgesic was provided ad libitum. During the procedure the oxygen saturation was monitored peripherally using a portable oxygen saturation monitor. No complications of the anaesthetic were noted. Post-operative care and recovery involved analgesia and prophylactic penicillin.
All procedures involved cannulation of the femoral artery. A 0.3cm dermal incision was made to allow insertion of the needle and guidewire into the femoral artery. A 20G catheter was fed over the guidewire that permitted access for the angiography catheters to the arterial system (Universal Microintroducer Kit (4.5F<ce: hsp sp=”0.2”/>×<ce: hsp sp=”0.2”/>5cm) Bard Access Systems, Utah, USA).
An aortogram was performed first. A catheter (4F Soft-Vu Omni-Flush Non-braided, AngioDynamics Inc, New York, USA) and guidewire (Straight Starter Guidewire 0.035 inch, Boston Scientific International, Massachusetts, USA) was inserted into the aorta, at the level of the renal arteries and radio-opaque contrast (Ultravist 300, Schering, Berlin, Germany) was injected at 2ml/s for 6s. This allowed the visualization of the left and right uterine arteries, possible ovarian arteries, and the distribution of vascularisation within each placenta. An initial aortic picture was taken in each case, allowing the animal to be positioned with maximal “spread” of the cotyledons and to minimize those overlapping; the animal was then left in this position for all subsequent pictures. The animal was not repositioned in between subsequent artery injections so that the number and location of the cotyledons could be superimposed to determine the lateral contributions. The images were observed for 1–2min until all contrast was cleared from the bed to identify all perfused components and to ensure that subsequent injections represented new blood flow.
The position of the catheter in the aorta was confirmed at the beginning and end of this injection sequence. The catheter was then manipulated under direct vision into the right common iliac artery and contrast injected at 1ml/s for 6s (selective right uterine artery angiogram). This second procedure allowed visualization only of the areas of the placenta supplied by the right uterine artery. Caution was taken to ensure that no contrast overflowed in a retrograde fashion into the aorta, which would have resulted in the visualization of the areas of the placenta supplied by the left uterine artery as well as the right. The catheter position was confirmed at the end of this sequence of pictures. In the 3-vessel angiograms, the catheter was repositioned into the left common iliac artery, checked at the beginning and end of the procedure and direct screening was used to determine that there was no backflow into the right iliac system. In order to reduce the risk of arterial spasm and the risk of non-visualized cotyledon, a high flow rate of non-ionic contrast (equal for each selective arteriogram) was used. The sequence of arteries was random for each 3 vessel angiogram. Digital data acquisition was performed at two images per second (DICOM, Virginia, USA).
Each spiral vessel generated a well-defined ‘blush’ of contrast, which corresponded to dye entering the intervillous space of each cotyledon. Due to the nature of the research projects, not all animals required two selective angiograms. The animals were not moved at any point during the procedure to facilitate the accurate identification of vascular branches. The iliac vessels were visualized and the uterine artery identified at its point of origin from the internal iliac artery. One complication arose from the procedures with dissection of the common iliac vessel with occlusion. This animal was not included in the final analysis.
Statistical analysis included comparison of the symmetry using a Kruskal–Wallis analysis of total cotyledons, and Chi sq testing for percentage above 50%. The numbers of cotyledons are expressed as median and range. The gestations at testing and maternal age are expressed as a mean and s.e.m. The significance was set at p<0.05, using Minitab 15™.
3. Results
The animals were on average 9 years of age and 140 days of gestation for the 3 vessel study and 154 days of gestation for the 2 vessel study (p=0.01), reflecting the timing of the protocols relevant to these studies (Table 1).
Table 1. Characteristics of animals undergoing angiography.
| Mean and s.e.m. | Range | |
|---|---|---|
| n=16 | ||
| Maternal age | 9 years±1 | 4–18 years |
| Parity | ||
| Primiparous | 6 | |
| Multiparous | 10 | 1–7 pregnancies |
| Fetal weight at delivery (n=9) | 598g | 500–705g |
| Gestation at 3 vessel angiogram | 140±4 days | 123–164 days |
| Gestation at 2 vessel angiogram | 154±2.3 days | 146–165 days |
The angiograms were more likely to have cotyledons perfused by the left uterine artery than the right across all angiograms (p=0.012). The placentae were all fundal and were more likely to be positioned to the right on anterior–posterior view (Table 2).
Table 2. Sided distribution of blood supply to placentae with angiography. 3 vessel picture allowed assessment of bilateral uterine supply and possible ovarian supply. In the 2 vessel picture, the opposite supply was determined by subtraction of the ipsilateral supply from the total supply.
| 3V (n=9) Median (Range) | 2V (n=7) Median (Range) | P | |
|---|---|---|---|
| Total number of cotyledons per animal | 14 (7–21) | 13 (7–20) | |
| Average number of cotyledons from left uterine artery per placenta | 9 (5–13) | 7 (4–1) | 0.105 |
| Average number of cotyledons from right uterine artery per placenta | 6 (1–10) | 6 (2–10) |
Overall, 77% of placentae had either bilateral (overlapping) blood supply (55%) or ovarian contribution to blood supply (22%) (Table 3). In the placentae with unaccounted supply from the iliac and uterine arteries, 2/20 and 1/7 cotyledons, i.e. 10% and 14%, of blood supply was attributable to ovarian supply. In the 5 placentae with overlapping supply, 11%, 43%, 5%, 44% and 14% of the total supply was contributed to by both left and right uterine arteries. This represented between 1 and 4 cotyledons in any given placenta. The placentograms performed at 140 days were no more likely to have left dominance than those done at 154 days of gestation (p=0.072).
Table 3. Contribution of blood supply by left or right sided dominance, overlapping supply or ovarian supply. The position of the placenta is indicated. NS= not significant.
| Number of animals (%) | Number of animals (%) | ||
|---|---|---|---|
| Left dominant (>50% of cotyledons) | 6 (66%) | 3 (43%) | 0.01 |
| Right dominant (>50% of cotyledons) | 2 (22%) | 2 (29%) | |
| Equal number of left and right supplied cotyledons | 1 (11%) | 2 (29%) | |
| Number of animals with overlapping supply | 5 (55%) | Undetermined | |
| Number with supply unaccounted for by the separate left and right iliac injections | 2 (22%) | Undetermined | |
| Placental position | |||
| Central | 3 | 2 | NS |
| Right | 5 | 2 | |
| Left | 1 | 3 | |
Fig. 1 demonstrates a placenta with left dominant blood supply, overlap of a large cotyledon and 1 cotyledon unaccounted for by the iliac studies. Fig. 2 demonstrates a larger cotyledon with ovarian supply in a central location, not supplied by either the left or right uterine arteries.
Fig. 1
A placenta with left dominant blood supply. Panel A, aortogram; Panel B, left iliac angiograms showing 6 cotyledons; and Panel C, right iliac angiograms showing 1 cotyledon. The angiograms display that the same cotyledon is supplied by both the left and right uterine arteries (overlapping supply 1/7, thin arrow). There is one cotyledon which is present in the aortogram, unaccounted for in the selective iliac angiograms, depicting evidence of a collateral supply (from the ovarian arteries, broad arrow). RT
=radiotelemeter; FS=fetal skeleton. Anatomical landmarks, aorta, left ureter, aortic catheter position and left iliac position are also marked. The fading cotyledon is encircled in image B.
Fig. 2
A placenta with ovarian blood supply. Panel A, aortogram showing 18 cotyledons (collateral supply, thick arrow); Panel B, left iliac angiograms showing 11 cotyledons (thin arrow depicts shared supply); and Panel C, right iliac angiograms showing 7 cotyledon (thin arrow depicts shared supply).
4. Discussion
This study demonstrates that the primate haemomonochorial placenta is supplied by a dominant left uterine artery in 67% of cases. There is cross over or dual supply of between 1 and 4 placenta cotyledons from the left and right uterine arteries in 55% of placentae. Up to 22% have probable ovarian artery contribution of up to 20% of cotyledons in individual placentae. The uterus is rotated to the right in the AP plane; the placenta is mostly fundal and located on the right side of the mid-abdomen. Nine out of the sixteen animals had a greater than 50% supply from the left, rather than the right, uterine artery. The data also shows that the ‘average baboon’ at a gestation of 147 days had 7–21 cotyledons present.
The 2 vessel angiography was from earlier work which attempted to limit the dosing of contrast before it was identified that this procedure was safe for a healthy pregnancy outcome. The subsequent angiograms have included the additional contralateral side angiogram so that collateral supply could be identified. All available angiograms (given the complexity in their collection) were included for completeness and contribute to the overall picture of baboon angiography. It in fact demonstrates that a more comprehensive angiogram is possible and that detail about tubo-ovarian supply can be obtained.
Given the similarities between human and baboon arterial supply, it is important to note that a high rate of anastomosis exists in single disc monochorial placentas, and the arteries may well originate from opposing uterine vessels [9]. The fetal baboon architecture has been defined elsewhere [10] and is similar to the human. In other primates such as the rhesus monkey, the overlap of uterine arteries was not seen; they have two disc placentas [11] and anastomoses usually occur at the level of the radial arteries, with multiple feeders into a single spiral artery.
This study revealed instances of maternal uteroplacental supply unaccounted for by the uterine arteries. Cotyledons in this category are likely to be perfused from the ovarian arteries. These images were taken as single injection studies and it is possible that there is sporadic flow into individual cotyledons, as seen in the rhesus [9], [12], [13], which may result in some cotyledons not being filled during injection of the aorta and uterine arteries. However, the blood flow was remarkably constant in this study of baboons. There were no instances in the sixteen angiograms of “additional” cotyledons arising from the iliac injections, that weren't seen on the aortic view. Spasm after the aortic injection could lead to a lower cotyledon count on the subsequent angiogram but low ionic contrast was used in these studies to reduce this risk and the injection rate was similar for all images.
In humans, it has been noted that the ovarian artery can provide significant collateral flow to the uterus as demonstrated in fibroid disease [1], [2], [14]. Anecdotal reports of ovarian contribution appear in the treatment of post-partum haemorrhage [15], [16]. Although the course of normal ovarian arteries is well described [14] – arising from the juxtarenal aorta as paired ventral structures – the ovarian arteries are rarely demonstrated in aortography due to their less than 1mm diameter. However, the ovarian arteries are dilated in the presence of uterine or other pelvic pathology. Therefore for physiological uterine supply to be from the ovarian arteries, the aortogram must have been taken from above the juxtarenal aorta and also some form of dilation must be present, as in pregnancy.
This study has potential implications in the treating of a bleeding uterus in primate pregnancy. This study demonstrates the variation in primate uteroplacental blood flow including the contribution of ovarian arteries and left and right collateralization. Similarity to human vascular anatomy strengthens the use of primate species as a model of human placentation.
References
- . Long-term efficacy and safety of uterine artery embolization in young patients with and without uteroovarian anastomoses. J Vasc Interv Radiol. 2008;19:195–200
- . Angiographic classification of ovarian artery-to-uterine artery anastomoses: initial observations in uterine fibroid embolization. Radiology. 2002;224:707–712
- . Metabolism of 17alpha-hydroxyprogesterone caproate by hepatic and placental microsomes of human and baboons. Biochem Pharmacol. 2008;75:1848–1857
- . Placental villous vascular endothelial growth factor expression and vascularization after estrogen suppression during the last two-thirds of baboon pregnancy. Endocrine. 2007;31:260–267
- . Animal models of human placentation – a review. Placenta. 2007;28, Suppl A:S41–S47
- . Maintenance of high risk pregnancies: role of prostaglandins and other mediators. Aust N Z J Obstet Gynaecol. 1994;34:351–356
- . Low-dose nitro-l-arginine administration in baboon (Papio hamadryas) pregnancy. Clin Exp Pharmacol Physiol. 1999;26:849–852
- Uteroplacental ischemia results in proteinuric hypertension and elevated sFLT-1. Kidney Int. 2007;71:977–984
- . Radioangiographic measurement of the internal diameters of the uteroplacental arteries in rhesus monkeys. Am J Obstet Gynecol. 1979;135(2):247–251
- . Observations on the vasculature of the baboon placenta (Papio sp.) with special reference to the transverse communicating artery. Folia Primatol (Basel). 1968;9:68–77
- . Radioangiographic measurement of the internal diameters of the uteroplacental arteries in rhesus monkeys. Am J Obstet Gynecol. 1979;135:247–251
- . Intermittent functioning of the uteroplacental arteries. Am J Obstet Gynecol. 1964;90:819–823
- . Placental radioangiography as a method for the study of uteroplacental blood flow in the rhesus monkey. J Med Primatol. 1979;8(1):48–56
- . Successful embolization of collaterals from the ovarian artery during uterine artery embolization for fibroids: a case report. J Vasc Interv Radiol. 2000;11:607–610
- . Serious primary post-partum hemorrhage, arterial embolization and future fertility: a retrospective study of 46 cases. Hum Reprod. 2008;23:1553–1559
- . Management of life-threatening postpartum hemorrhage: indications and technique of arterial embolization. J Radiol. 2006;87:533–540
PII: S0143-4004(09)00332-4
doi:10.1016/j.placenta.2009.10.009
© 2009 Elsevier Ltd. All rights reserved.
