Effect of internal iliac artery ligation on sonographic blood changes in the uterine arteries: A case–control study in women with postpartum hemorrhage in sub-Saharan Africa

Abstract

Objective

Internal iliac artery ligation due to severe postpartum hemorrhage has several pathophysiological consequences in terms of blood supply of the uterus. In this context, this study aimed to evaluate the sonographic changes in the uterine arteries after internal iliac artery ligation.

Methods

The population of this case–control study consisted of women unresponsive to conservative medical treatment who underwent internal iliac artery ligation for severe postpartum hemorrhage between January 2022 and December 2022. Women who underwent internal iliac artery ligation were included in the internal iliac artery ligation group, and those with similar postpartum age, parity, and body mass index were included in the control group. This retrospective case–control study included 62 patients with postpartum hemorrhage and ongoing bleeding despite all conservative medical treatments who underwent internal iliac artery ligation as well as 86 patients who were also in the postpartum period and had similar age, body mass index, and parity but did not have a history of postpartum hemorrhage at birth and did not undergo internal iliac artery ligation. The pulsatility index, resistance index, and maximal-to-minimal diastolic flow of the uterine arteries were calculated using color and pulsed Doppler imaging at the 6-month follow-up and compared between the two groups.

Results

There were 62 and 86 women in the internal iliac artery ligation and control groups, respectively. There was no significant difference between the two groups in terms of demographic and clinical characteristics and obstetric history (p > 0.05). There was also no significant difference between the two groups in terms of the pulsatility index, resistance index, and maximal-to-minimal diastolic flow of the uterine arteries determined at the 6-month follow-up (p > 0.05).

Conclusion

The study’s findings demonstrated that internal iliac artery ligation has no adverse effect on the sonographic blood flow indexes of the uterine arteries in women with postpartum hemorrhage.

Keywords

Postpartum hemorrhage internal iliac artery ligation uterine artery Doppler ultrasound imaging

Introduction

Obstetric hemorrhage is the leading preventable cause of maternal morbidity and mortality worldwide, with an incidence of approximately 140,000 deaths per year.¹ The vast majority of obstetric hemorrhages are caused by postpartum hemorrhage (PPH). In the current literature, massive PPH is defined as a total postpartum blood loss of ≥1000 mL, including intrapartum blood losses, regardless of the mode of delivery. It is observed in approximately 1%–5% of all births.²

Hysterectomy is a radical operation that has been performed throughout history in life-threatening cases with large blood volume loss due to severe PPH and still has a place in surgical treatment in the medical world.^3–6 Internal iliac artery ligation (IIAL), in which the internal iliac arteries are surgically ligated unilaterally or bilaterally, is an alternative surgical approach to peripartum hysterectomy in cases where conservative methods used to stop obstetric hemorrhage fail.^4–7 In addition, IIAL has been described as a uterine-preserving surgical approach that can be safely performed in young women with low parity or a desire for long-term fertility, unlike hysterectomy.^1,3–6 In studies evaluating cases of peripartum and postpartum hemorrhage, the IIAL procedure was effective in avoiding hysterectomy in approximately 50% of cases.^5–9

There are various uterus-sparing surgical procedures, including compression sutures, selective uterine devascularization or embolization, and ligation or embolization of the arteries supplying the female genital organs, used to treat patients with severe PPH.⁷ Rapidly controlling the obstetric hemorrhage without compromising the rest of the pelvic blood supply is the ultimate goal of uterus-sparing procedures.^4,7 In an experimental study, Burchell⁹ showed a 49% decrease in pelvic blood flow and an 85% decrease in pulse pressure due to IIAL. Nevertheless, preserving the genital organs with an intact blood supply may not always result in pregnancy due to hormonal and functional changes in the genital system.¹⁰

Blood flow to the uterus is provided mainly by the uterine arteries.¹¹ Not all vascular disturbances in pelvic organs result in ischemia associated with vascular anastomoses between the arteries and veins of the pelvic organs.^12–14 Detailed knowledge about the branching pattern of the internal iliac artery may shed light on the consequences of IIAL performed on the pelvic viscera.

IIAL is one of the techniques used to diminish blood flow to the uterus in patients with PPH.^11,15 Considerable variations in the vascular pattern of the internal iliac artery have been noted in the literature.¹⁴ A limited number of studies with small sample sizes have assessed the changes in the collateral blood supply of the uterus following IIAL.^10,11,16,17 Hence, controversy remains regarding the changes in the uterine arteries following uterus-sparing surgical procedures. In this context, this study was carried out to evaluate the sonographic changes in the uterine arteries after IIAL due to PPH.

Materials and methods

Study design

The population of this case–control study consisted of women who underwent bilateral IIAL due to severe PPH at the Department of Obstetrics and Gynecology, Turkish-Nigerian Friendship Hospital, Niamey, Niger, between January 2022 and December 2022. The study protocol was approved by the local and ethical clinical research committees of the Ministry of Health of the Republic of Turkey (Approval Date & No: 30.05.2023 & E-774.99-238414145). The study was conducted in accordance with the principles stated in the 1975 Helsinki Declaration, revised in 2013. Written informed consent was obtained from all participants included in the study. All patient information was de-identified.

The reporting of this study conforms to the STROBE guidelines.¹⁸

Study sample

Consecutive female patients aged between 17 and 47 years who were unresponsive to conservative medical management during their most recent delivery and subsequently underwent IIAL due to severe PPH were included in the study. Our study is a retrospective case–control study including 62 patients with PPH and ongoing bleeding despite all conservative medical treatments who underwent IIAL as well as 86 patients who were also in the postpartum period and had similar age, body mass index (BMI), and parity but did not have a history of PPH at birth and did not undergo IIAL.

The diagnostic criterion for PPH was an estimated blood loss exceeding 500 mL following vaginal delivery or 1000 mL during cesarean section. Conservative treatment options included interventions such as uterine massage, bimanual compression, administration of uterotonics, intravenous hydration, and blood transfusion.^19,20 The decision to switch from conservative management to interventional or surgical procedures was made by the attending physician based on the assessment of unresponsiveness to conservative measures.

The type of delivery, either spontaneous vaginal delivery or cesarean section, was not an exclusion criterion. Thus, patients who had undergone other uterus-sparing techniques (e.g. Hayman or B-Lynch sutures) were not excluded from the study. Conversely, patients who had undergone hysterectomy for PPH, those with systemic or endocrine disorders, a history of previous chemotherapy or radiotherapy, endometriosis, polycystic ovary syndrome, or incomplete follow-up data were excluded from the study. The IIAL techniques and other uterus-sparing methods used within the scope of the study have been detailed previously.^11,20–23

Women with IIAL who met the inclusion criteria constituted the IIAL group, and women without IIAL whose postpartum age, parity, and BMI were close to those of cases that underwent IIAL constituted the control group. We attempted to create a control group similar to the study group in terms of demographic data.

Follow-up procedure

The participants included in the IIAL and control groups received standardized follow-up care at the Department’s outpatient clinics after IIAL. At the 6-month follow-up, color and pulsed (duplex) Doppler imaging were conducted during the mid-follicular phase (typically day 5 or 6 of the menstrual cycle) between 10:00 AM and 12:00 PM to assess the pulsatility index (PI), resistance index (RI), and maximal-to-minimal diastolic flow (S/D) of the uterine arteries. An endo-cavitary convex transducer (8 MHz, General Electric Logic5 Pro, Milwaukee, Wisconsin, USA) was used for imaging.^7,10 The average of the three consecutive measurements for each variable was calculated and recorded. The radiologists were blinded to the women’s clinical history and previous treatments.

Data collection and variables

Participants’ demographic characteristics, including age, clinical characteristics (BMI, comorbidities, and smoking history), and obstetric history (gravidity, parity, abortion, dilatation and curettage, and contraception) were recorded. The PI, RI, and S/D values were calculated for each participant.

Statistical analysis

In our investigation, the primary endpoint was the variations observed in the Doppler blood flow metrics of the uterine arteries. To ensure a comprehensive understanding of our dataset, we employed a range of descriptive statistics. Continuous variables that adhered to a Gaussian distribution are articulated as mean ± standard deviation. In instances in which the data deviated from a normal distribution, the median and range (minimum and maximum values) are reported. Categorical variables are systematically delineated using frequency counts (n) and their associated percentages (%).

To rigorously assess the distribution of our numeric variables, we employed a trio of tests: Shapiro–Wilk, Kolmogorov–Smirnov, and Anderson–Darling tests. These tests were chosen for their sensitivity in detecting deviations from normality, which is crucial for the subsequent selection of appropriate inferential tests. For the comparison of two independent groups, the independent samples t-test was our method of choice for data that conformed to normal distribution, given its robustness in detecting mean differences. For non-normally distributed data, the Mann–Whitney U test was employed, a nonparametric alternative that makes fewer assumptions about the data distribution. Categorical data comparisons between our groups were facilitated using Pearson’s chi-square test, renowned for its efficacy in detecting associations in categorical data. In specific scenarios for which expected cell frequencies were low, Fisher’s exact test was judiciously employed to ensure accurate P value computation.

Our statistical computations were executed using two renowned software platforms: Jamovi version 2.2.5.0 and JASP version 0.16.1. Both tools are recognized for their precision and versatility in handling complex datasets. A probability (p) value threshold of ≤0.05 was set a priori as the benchmark for statistical significance, in line with conventional scientific standards.

Results

There were 62 and 86 women in the IIAL and control groups, respectively. There was no significant difference between the two groups in terms of demographic and clinical characteristics and obstetric history (p > 0.05) (Table 1).

Table 1.

Demographic and clinical characteristics and obstetric history of the study groups.

	Groups
	IIAL	Control	p
	(n = 62)	(n = 86)	p
Age (years)^†	31.0 ± 9.7	28.0 ± 8.2	0.055
BMI (kg/m²)^‡	22.3 [17.3–38.9]	24.1 [17.6–36.8]	0.328
Hemoglobin (pre–post)^‡	3.6 [1.7–6.6]	1.1 [0.4–3.8]	<0.001
Hematocrit (pre–post)^‡	11.4 [4.6–20.0]	1.6 [0.0–10.0]	<0.001
Mode of delivery^§
Normal birth	16 (25.8%)	25 (29.1%)	0.832
Normal birth with intervention	26 (41.9%)	32 (37.2%)
C/S birth/C/S plus myomectomy	20 (32.3%)	29 (33.7%)
Comorbidities^§	27 (43.5%)	19 (22.1%)	0.009
Hypertension	16 (25.8%)	12 (14.0%)	0.109
Diabetes mellitus	11 (17.7%)	8 (9.3%)	0.206
Asthma	11 (17.7)	6 (7.0%)	0.078
Others	4 (6.5%)	3 (3.5%)	0.453
Smoking^§	7 (11.3%)	6 (7.0%)	0.535
Obstetric history
Gravidity^‡	5.5 [1.0–13.0]	5.0 [1.0–14.0]	0.710
Parity^‡	4.0 [0.0–11.0]	4.5 [1.0–12.0]	0.654
Abortion^‡	1.0 [0.0–5.0]	1.0 [0.0–4.0]	0.508
Patients with dilatation & curettage^§	7 (11.3)	13 (15.1)	0.669
Contraception use^§	0 (0.0)	3 (3.5)	0.265

†

: mean standard deviation, ^‡: median [min–max], ^§: n (%).

BMI: body mass index; C/S: cesarean section, IIAL: internal iliac artery ligation.

Assisted vaginal (41.9%) and vaginal deliveries (25.8%) were the most frequent modes of delivery among patients in the IIAL group. Uterine atony (56.5%) and placental invasion anomalies (17.7%) were the most common indications for IIAL. Compression sutures were not used in seven (11.3%) patients. B-Lynch and Hayman’s sutures were performed in 44 (71.0%) and 11 (17.7%) patients, respectively. Participants’ other characteristics are detailed in Table 2.

Table 2.

Perioperative characteristics of the patients with IIAL.

Feature	Group IIAL (n = 62)
Mode of delivery^§
Vaginal delivery	16 (25.8)
Assisted vaginal delivery	26 (41.9)
Cesarean section	17 (27.4)
Cesarean section plus myomectomy	3 (4.8)
Indication for IIAL^§
Uterine atony	35 (56.5)
Placental invasion anomalies	11 (17.7)
Uterine rupture	6 (9.7)
Retroperitoneal hematoma	4 (6.5)
Vulvar hematoma	2 (3.2)
Deep vaginal/cervical tear	4 (6.5)
Compression sutures^§
No compression suture	7 (11.3)
B-Lynch	44 (71.0)
Hayman	11 (17.7)

†

: mean standard deviation, ^‡: median [min–max], ^§: n (%)

IIAL: internal iliac artery ligation.

There was no significant difference between the two groups in terms of Doppler blood flow characteristics of the uterine arteries, including PI, RI, and S/D (p > 0.05) (Table 3, Figure 1).

Table 3.

Doppler blood flow characteristics of the uterine arteries at the 6-month follow-up examination.

	Groups
Parameter	IIAL (n = 62)	Control (n = 86)	p
PI^‡	1.6 [1.3–2.2]	1.7 [1.3–2.4]	0.172
RI^‡	0.6 [0.5–1.0]	0.7 [0.5–1.0]	0.652
S/D^‡	2.7 [2.1–3.9]	2.6 [2.2–3.1]	0.163

‡

: median [min–max], ^§: n (%).

IIAL: internal iliac artery ligation, PI: pulsatility index, RI: resistance index, S/D: maximal-to-minimal diastolic flow.

Figure 1.

Doppler blood flow metrics of the uterine arteries at 6-month follow-up. The panel illustrates the median values with range (min–max) for the pulsatility index (PI), resistance index (RI), and maximal-to-minimal diastolic flow (S/D) for both the internal iliac artery ligation (IIAL) and control groups.

When the Doppler blood flow characteristics of the uterine arteries were examined in the 6-month follow-up examinations by matching cases with similar delivery methods as subgroups, no statistically significant difference was observed between the uterine artery Doppler blood flows of the two groups when cesarean delivery and normal delivery were compared (Table 4).

Table 4.

Doppler blood flow characteristics of uterine arteries at 6-month follow-up examinations according to birth.

		Group						Total
	Mode of delivery	IIAL			Control			Total
	Mode of delivery	n	Med. [min–max]	p	n	Med. [min–max]	p	N	Med. [min–max]	p
PI	Cesarean section	20	1.7 [1.4–2.2]	0.910	29	1.6 [1.3–2.0]	0.184	49	1.7 [1.3–2.2]	0.278
PI	Vaginal delivery	42	1.6 [1.3–2.0]	0.910	57	1.7 [1.4–2.4.]	0.184	99	1.7 [1.3–2.4]	0.278
RI	Cesarean section	20	0.6 [0.5–1.0]	0.821	29	0.6 [0.5–1.0]	0.568	49	0.6 [0.5–1.0]	0.737
RI	Vaginal delivery	42	0.6 [0.5–1.0]	0.821	57	0.7 [0.5–1.0]	0.568	99	0.7 [0.5–1.0]	0.737
S/D	Cesarean section	20	2.7 [2.1–3.1.]	0.603	29	2.5 [2.2–3.0]	0.418	49	2.6 [2.1–3.1]	0.344
S/D	Vaginal delivery	42	2.7 [2.1–3.9]	0.603	57	2.6 [2.2–3.1]	0.418	99	2.6 [2.1–3.9]	0.344

IIAL: internal iliac artery ligation, PI: pulsatility index, RI: resistance index, S/D: maximal-to-minimal diastolic flow.

Discussion

Our findings did not reveal any significant change in the blood flow characteristics of the uterine arteries following IIAL. Similarly, Yildirim et al.¹⁰ found no significant difference in the vascular indices for the uterine, arcuate, and ovarian arteries after IIAL. They concluded that IIAL has no negative impact on pelvic circulation. Other researchers have reported similar findings.^12,17,24 We did not observe any significant change in the Doppler characteristics of the groups. Thaler et al.²⁵ and Chitrit et al.¹⁷ demonstrated a variable number of physiologic variabilities in the blood flow of uterine arteries. Taken together, these findings indicate that IIAL does not block or impair uterine blood perfusion. The variability of the anomalies and anastomoses of the internal iliac arteries is vital to prevent future ischemic changes in patients with IIAL or other related procedures.

The effects of IIAL on flow velocity waveforms from uterine arteries during pregnancy have been addressed previously.^26,27 Placental and uterine hypervascularization in pregnancy due to pelvic collateral circulation or recanalization of the internal iliac artery might be the reason for maintaining near-normal perfusion pressure following IIAL.^16,27 These findings might be attributed to the normal reproductive function in patients with a history of IIAL. Domingo et al.¹¹ suggested that the absence of uterine revascularization may reduce the possibility of pregnancy after IIAL. However, they evaluated the reproductive potential of only three patients. Therefore, large-scale studies are needed to establish the long-term effects of IIAL on fertility.

Several studies investigated the changes in the ovarian and clitoral arteries after IIAL or uterine artery ligation. Although Raba and Baran¹² found significant decreases in PI and S/D of the ovarian arteries after IIAL, they did not detect any change in PI, RI, and S/D of the uterine arteries. They also observed the reversed blood flow in the ovarian branches of the uterine arteries after IIAL.¹⁵ Hence, the post-IIAL changes in the ovarian arteries may be related to the changes in the ovarian reserve in patients with IIAL. Kaya et al.⁴ reported that IIAL did not affect the blood flow of the dorsal clitoral artery between 3 and 6 months postpartum. The ovarian blood supply was not affected secondary to uterine artery ligation.^7,28 However, the authors failed to explain the lack of a significant decrease in the uterine blood flow after its ligation due to the low recanalization rate. Increased collateral circulation with a possible recanalization remains the most probable pathophysiological change in these patients.

Limitations

The relatively large sample size was the major strength of our study. The study’s primary limitation was the lack of data about the patients’ fertility outcomes. Other study limitations include the unequal distribution of the patients with different compression sutures in the study group and the relatively short follow-up time.

Conclusion

The study’s findings demonstrated that IIAL had no adverse effect on the Doppler imaging findings of the uterine arteries. Prospective studies with fertility outcomes are needed to clarify the controversies regarding the functional changes of the pelvic organs after IIAL.

Footnotes

Acknowledgements

The authors wish to acknowledge all study participants, study staff, and staff at the health research center and study health facilities.

Author contributions

IK, OB, and FC conceived the study, designed the work, acquired and interpreted the data, and drafted the manuscript. OB, IK, and FC were involved in the design of the work, data acquisition, and interpretation of data. OB and IK participated in the design of the work and analysis and interpretation of data. All authors approved the submitted version.

Data availability statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declaration of conflicting interests

Authors have no conflicts of interest to declare.

Ethics committee approval

The study protocol was approved by the local and ethical clinical research committees of the Ministry of Health of the Republic of Turkey (Approval Date & No: 30.05.2023 & E-774.99-238414145). The study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants included in the study.

Funding

The authors declare that this study received no financial support.

Informed consent

Written informed consent was obtained from all participants included in the study.

ORCID iD

Özer Birge

References

Maheshwari

SLK

Kumari

Ahmad

SN.

Role of bilateral internal iliac artery ligation in severe obstetric and gynaecological hemorrhage. Int J Reprod Contracept Obstet Gynecol 2018; 7: 1090–1093.

Menard

Main

Currigan

SM.

Executive summary of the reVITALize initiative: standardizing obstetric data definitions. Obstet Gynecol 2014; 124: 150–153.

Nizard

Barrinque

Frydman

, et al. Fertility and pregnancy outcomes following hypogastric artery ligation for severe post-partum haemorrhage. Hum Reprod 2003; 18: 844–848.

Kaya

Usluogullari

Yurttutan

, et al. Does ligation of internal iliac artery for postpartum hemorrhage affect clitoral artery blood flow and postpartum sexual functions? Eur J Obstet Gynecol Reprod Biol 2017; 219: 124–128.

Sziller

Hupuczi

Papp

Hypogastric artery ligation for severe hemorrhage in obstetric patients. J Perinat Med 2007; 35: 187–192.

Unal

Kars

Buyukbayrak

, et al. The effectiveness of bilateral hypogastric artery ligation for obstetric hemorrhage in three different underlying conditions and its impact on future fertility. J Matern Fetal Neonatal Med 2011; 24: 1273–1276.

Kaplanoglu

Karateke

, et al. Evaluation of uterine artery recanalization and Doppler parameters after bilateral uterine artery ligation in women with postpartum hemorrhage. Int J Clin Exp Med 2015; 8: 7823–7829.

Singh

Kishore

Saxena

SS.

Ligating internal iliac artery: success beyond hesitation. J Obstet Gynaecol India 2016; 66: 235–241.

Burchell

RC.

Physiology of internal iliac artery ligation. J Obstet Gynaecol Br Commonw 1968; 75: 642–651.

10.

Yildirim

Gultekin

Kocyigit

, et al. Color Doppler analysis of pelvic arteries following bilateral internal iliac artery ligation for severe postpartum hemorrhage. Int J Gynaecol Obstet 2009; 104: 22–24.

11.

Domingo

Perales-Puchalt

Soler

, et al. Clinical outcome, fertility and uterine artery Doppler scans in women with obstetric bilateral internal iliac artery ligation or embolisation. J Obstet Gynaecol 2013; 33: 701–704.

12.

Raba

Baran

Hemodynamic parameters following bilateral internal iliac arteries ligation as a treatment of intrapartum hemorrhage. Ginekol Pol 2009; 80: 179–183.

13.

Takebe

Uchida

Teramoto

An experimental hemodynamic study of the pelvic collateral circulation. Acta Med Okayama 1994; 48: 31–38.

14.

Mamatha

Hemalatha

Vinodini

, et al. Anatomical study on the variations in the branching pattern of internal iliac artery. Indian J Surg 2015; 77: 248–252.

15.

Raba

Effect of internal iliac artery ligation on ovarian blood supply and ovarian reserve. Climacteric 2011; 14: 54–57.

16.

Khelifi

Amamou

Salem

, et al. Ligature therapeutique des arteres hypogastriques: surveillance par echo-Doppler couleur [Therapeutic ligature of hypogastric arteries: color Doppler follow-up]. J Radiol 2000; 81: 607–610.

17.

Chitrit

Guillaumin

Caubel

, et al. Absence of flow velocity waveform changes in uterine arteries after bilateral internal iliac artery ligation. Am J Obstet Gynecol 2000; 182: 727–728.

18.

Von Elm

Altman

Egger

; STROBE Initiativeet al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

19.

Kaya

Damarer

Daglar

, et al. Is there yet a role for internal iliac artery ligation in obstetric hemorrhage with the current gain in popularity of other uterus sparing techniques? J Matern Fetal Neonatal Med 2017; 30: 1325–1332.

20.

Gezginç

Yazici

Koyuncu

, et al. Bilateral uterine and ovarian artery ligation in addition to B-lynch suture may be an alternative to hysterectomy for uterine atonic hemorrhage. Clin Exp Obstet Gynecol 2012; 39: 168–170.

21.

Mathyk

Cetin

Atakul

, et al. Ovarian reserve after internal iliac artery ligation. J Obstet Gynaecol Res 2018; 44: 1761–1765.

22.

Çetin

Aydogan Mathyk

Atis Aydin

, et al. Comparing success rates of the Hayman compression suture and the Bakri balloon tamponade. J Matern Fetal Neonatal Med 2019; 32: 3034–3038.

23.

Şahin

Soylu Karapınar

Şahin

, et al. The effectiveness of the double B-lynch suture as a modification in the treatment of intractable postpartum haemorrhage. J Obstet Gynaecol 2018; 38: 796–799.

24.

Demirci

Ozdemir

Safak

, et al. Comparison of colour Doppler indices of pelvic arteries in women with bilateral hypogastric artery ligation and controls. J Obstet Gynaecol 2005; 25: 273–274.

25.

Thaler

Manor

Rottem

, et al. Hemodynamic evaluation of the female pelvic vessels using a high-frequency transvaginal image-directed Doppler system. J Clin Ultrasound 1990; 18: 364–369.

26.

Papp

Sztanyik

Szabó

, et al. Successful pregnancy after bilateral internal iliac artery ligation monitored by color Doppler imaging. Ultrasound Obstet Gynecol 1996; 7: 211–212.

27.

Evans

McShane

The efficacy of internal iliac artery ligation in obstetric hemorrhage. Surg Gynecol Obstet 1985; 160: 250–253.

28.

Verit

Çetin

Keskin

, et al. Does bilateral uterine artery ligation have negative effects on ovarian reserve markers and ovarian artery blood flow in women with postpartum hemorrhage? Clin Exp Reprod Med 2019; 46: 30–35.