Data on safety of intravaginal boric acid use in pregnant and non-pregnant women: a narrative review

Abstract

Intravaginal boric acid (IBA) represents one of the only options available to treat azole resistant vulvo-vaginal candidiasis (VVC), and is included as part of multiple national guidelines (including UK and US) for treatment of VVC or recurrent Bacterial Vaginosis (BV). Novel products utilizing IBA are under development for treatment and suppression of VVC and BV. Use of over the counter (OTC) or clinician-prescribed IBA in reproductive aged women is already widespread, and may increase further if drug resistance in VVC rises. Yet, IBA is not an FDA-approved drug and safety data are sparse. Given these factors, it is important to understand the currently available data on the safety of IBA use. Herein, we set out to synthesize human and animal data (converting, where appropriate, dose and serum values to standard units to facilitate comparison) to answer two key questions: 1. What are the data on the safety of IBA use for women? and 2. What are the data on the safety of IBA use in pregnancy? We find that, despite gaps, available data suggest IBA use is safe, at least when used in doses commonly described in the literature as being prescribed by clinicians. Information on harms in pregnancy is limited and data remain insufficient to change current guidelines which recommend IBA avoidance in pregnancy.

SUMMARY

A narrative review summarizing available data on the safety of intravaginal boric acid use for women and in the setting of pregnancy.

INTRODUCTION

Intravaginal boric acid (IBA) has been used for decades to treat vulvovaginal candidiasis (VVC), and bacterial vaginosis (BV) (1–3). IBA represents one of the few available options for azole-resistant Candida (4–6), is a treatment of last resort for nitroimidazole-resistant trichomoniasis (TV) (4–10% of TV isolates show some nitroimidazole resistance (6,7))(5,8–10), and is included as part of multiple national guidelines (Table 1) for infectious vaginitis. IBA is not an FDA approved drug; rather it can be prescribed from compounding pharmacies or is widely available over the counter(OTC) (often marketed as a “homeopathic” product). For example, a search on a large internet retailer reveals >20 brands of BA-containing vaginal suppositories, marketed for indications ranging from vaginal odor, discharge, and dryness to VVC, BV, and dyspareunia or “correcting” vaginal pH. Formulations vary; commonly BA powder is placed in a gelatin or vegetable based capsule for intravaginal use. Costs online range from $0.10 to $1.20 per capsule; costs at two compounding pharmacies were $0.19 to $2.33 per capsule.

Table 1.

National and International Guideline Recommended Use of Intravaginal BA

Guideline	Country	Indication for BA	Dosage/Duration BA	Comments on safety for patient	Comments on pregnancy	Other comments
Centers for Disease Control STD Treatment Guidelines, 2015 and 2021(5,6)	USA	Nonalbicans VVC (if recurrence after treatment with 7–14 days non-fluconazole azole regimen)	600mg IBA in gelatin capsule once daily for 2 weeks (3 weeks in 2021 guidelines)	None	None	Also mentions that clinical improvement for recalcitrant trichomonas has been reported with intravaginal BA.*
		Recurrent BV**	600mg IBA daily for 21 days (preceded by oral nitroimidazole, followed by metronidazole intravaginal gel twice weekly for 4–6 months)	None	None
ACOG 2020 “Vaginitis in Nonpregnant Patients”(107)	USA	Nonalbicans VVC	600mg IBA capsules daily for a minimum of 14 days. (No other options recommended except topical flucytosine)	“BA can be fatal if ingested orally and patients should be well counseled to use it only intravaginally, to place it out of the reach of children, and to use reliable contraception.”	“Patients should be counseled…to use reliable contraception”
Canadian Guidelines on Sexually Transmitted Infections 2010(108)	Canada	Recurrent VVC (4 or more episodes in 12mos). Recommend induction and maintenance therapy	Induction: 300–600mg IBA gelatin capsule daily for 14 days. Maintenance: 300mg IBA gelatin capsule 5 days each month on the first day of the menstrual cycle. Continue 6 months then observe. If recurrence occurs, treat the episode and re-introduce maintenance regimen	None	States “contraindicated in pregnancy”
		Non-albicans VVC	600mg IBA capsule daily for 14 days If symptoms recur retreat with 600mg daily for 14 days followed by “alternate day BA for several weeks”	Vaginal burning reported in <10%	None
BASHH guidelines vulvovaginal candidiasis 2019(109)	UK	Non-albicans Candida species and azole resistance-alternative regimen to nystatin	BA vaginal suppositories 600mg daily for 14 days	BA vaginal suppositories 600mg daily for 14 days are a “safe and effective alternative (Grade 1B).” If mucosal irritation occurs dose can be reduced to 300mg daily.	“There may be a teratogenic risk so BA should be avoided in pregnancy or risk of pregnancy”
Australian STI management guidelines for use in primary care(110)***** ††	Australia	C. glabrata	BA vaginal pessaries 600mg daily for 14 days	None	None

^*Muzny et al(9) and Backus et al(8) 600mg intravaginal BID X 60 days, Aggarwal et al(10) (1 patient vaginal clotrimazole alternating every other night with 600mg BA intravaginally-6 weeks failed, but was used for 5 months which was successful; second patient vaginal clotrimazole in am with 600mg vaginal BA at night nightly for 3 weeks-failed, but then used 600mg intravaginal BA twice daily with weekly intravaginal gentian violet for 1 month, was successful)

^**“Limited data indicate that for women with multiple recurrences an oral nitroimidazole (metronidazole or tinidazole 500 mg twice daily for 7 days) followed by intravaginal BA 600 mg daily for 21 days and suppressive 0.75% metronidazole gel twice weekly for 4–6 months might be an option for women with recurrent BV.”

When described in the literature as prescribed by clinicians, IBA is most commonly used as a 600 mg suppository for 14 days for acute VVC; 600 mg twice weekly used even over years has been used for VVC or BV suppression(3), and 600 mg twice daily (BID) for up to 2 months for resistant TV(8–10) (Table 1, footnote). A protocol for a randomized IBA trial for BV treatment has been described(11), and a phase 2 trial of an IBA-containing product for treatment of BV and VVC was recently published(12). Several trials of IBA-based products are registered on clinicaltrials.gov, including twice weekly use for BV suppression(13).

Use of IBA is already widespread, and may increase further as drug resistance in VVC and TV rises, or if new trials(12,13) prove efficacious. The majority of women with infectious vaginitis are of reproductive age, and up to 50% of U.S. pregnancies are unplanned(14). Yet, safety data on IBA is relatively sparse(1). Given these factors, it is important to understand the available data on the safety of IBA use. Herein, we reviewed the literature to answer two key questions: 1. What are the data on the safety of IBA use for women? and 2. What are the data on the safety of IBA use in pregnancy?

MATERIALS AND METHODS

Searches were performed using keyword and controlled vocabulary terms. Abstracts were reviewed, reference lists searched and relevant papers abstracted (See Suppl. Table 1 and Suppl. Fig. 1). We considered data on borates and elemental boron: BA (H3BO3) is a borate compound resulting when boron combines with oxygen(15). For BA consumed orally or applied topically, absorbed boron can be measured in the blood. For selected papers, calculations were made to standardize measures for comparability (Suppl. Table 2).

RESULTS

What are the data on the safety of IBA use for women?

Toxicity from IBA use in humans

We were able to identify >1,100 individual patient uses based on published literature of IBA use (primarily for VVC)(3,8–10,12,16–53). Most described 600mg once daily or BID for 7–14 days(12,17,21–28,30,31,36,37,39,41,41,44–51), while others described BID for 1–2 months(10,9,8,53,43,16) or extended or suppressive regimens(3,18,23,24,26,27,30,35,40). Perhaps the greatest evidence for the safety of IBA is that despite this literature, we were not able to find published reports since the 1880’s of serious toxicities. In 1888 there was a report of toxicity in 3 women from IBA “packs”(52). The author described treating chronic leucorrhoea (heavy vaginal discharge) by packing 1–2 ounces of BA into the upper 1/3 of the vagina, tamponing it for 2–3 days “until liquefaction occurred”, then using a moderately hot douche to clean it out, 1–2 times per week. These women exhibited formication, depression, corrosion of vaginal mucous membrane, sunken eyes, weak voice, and skin swelling, charring, and skin exfoliation, which resolved on stopping the therapy, and were ascribed to BA toxicity. Of note, the amount of BA used daily in these cases would appear to be substantially larger (25–100 times) than the 600 mg daily or BID commonly prescribed or advised in guidelines. Since 1888, the only other documented adverse events have been sporadic and mild, primarily limited to local irritation or watery vaginal discharge(3,18,42,45).

Humans are routinely exposed to low levels of boron (B) in food and water; indeed some level of intake is expected and may even be important for health(54) — “average” levels are 34–95 ng/ml, (.04 ug/g WB)(55). However, exposure to large amounts of BA can cause toxicity and even death(56–62). Given limited data on IBA it might be useful to compare 1. the amount administered vaginally to “safe” or “toxic” exposures or 2. the blood level observed as a result of vaginal absorption to “safe” or “toxic” serum levels from the broader toxicity literature, accounting for differential absorption.

Is there a known “toxic” or “lethal” dose of BA based on amount administered or measured in blood?

Much of acute human toxicity data relates to case series of reported accidental or purposeful oral ingestion, often in children(56–58,61), of BA-containing insecticides or cleaning products(57,59,61). Orally ingested BA appears to be nearly 100% systemically absorbed(63,64), though quickly renally excreted. This was shown in one study in which 3 men drank 750 mg BA dissolved in water; 3 more ate 740–1473 mg BA(63). BA was rapidly excreted in the urine, showing 100% oral bioavailabilty. In another study, 562–611 mg BA was administered to human males intravenously (IV) over 20 minutes. (64). No volunteer in either study suffered ill effects.

A full review of all case reports of oral toxicity is outside of the scope of this paper (see Hadrup et al.(65) for additional details). However, in case series of large amounts of BA oral ingestion, nausea, vomiting, abdominal pain, and diarrhea were common(57,61). In more severe cases, patients exhibited blue-green emesis, CNS depression, fever, headache, an erythematous “boiled lobster” desquamating skin rash, weakness, cyanosis, and renal failure.(57–59,61,62) Yet, most (70–80%) of those with reported ingestions were asymptomatic(57,61). Case series have not reliably established a threshold blood boron level or a threshold BA ingestion amount consistently leading to toxicity or death (see Table 2). Reported toxic or fatal ingestion amounts are also variable. Litovitz noted “potentially lethal (oral) doses are…cited as 15 to 20 g for adults” but discuss that they did not observe fatalities or severe manifestations despite ingestions up to 88.8 g of BA(61). They do report on systemic symptoms resulting from ingestion of 0.2 g/kg BA (14,000 mg in a 70 kg adult) (57). A clear limitation of these studies is variable reporting on amount, concentration, and timing of ingestions and when after ingestion blood levels were drawn, as well as patient size (many children included).

Table 2.

Selected studies reporting acute BA exposure, measured blood levels and outcomes in nonpregnant humans and animals

Study	# of patients	Range of estimated daily Boron or BA exposure (in same units throughout)	Mode of exposure	Range of boron levels (whole blood)	Outcome
Rat oral exposure
Weir et al, 1972(66)	N/A	LD50: 710 mg B/kg in a single oral ingestion	Data reported as gB/kg. Female rats weighed 206– 248 g in this study	Would be 285,600 mg BA ingestion in a 70 kg human	Death in 50% of rats
Human oral exposure
Linden et al, 1986(57)	4 (aged 14 months – 65 years)	~10,000 mg- 297,000 mg BA (individual weights of patients not reported)	Oral	46 ug B/g WB – 2196 ug B/g WB	Toxicity, no deaths. 3 patients, including those with the lowest and highest blood levels, had GI symptoms; 1 with a blood level of 75.7 ug/g WB was asymptomatic
Ishii et al, 1993(58)	1 (77 years old)	~30,000 mg BA	Oral	35.7 ug B/g WB	Death
Human IV Exposure
Jansen et al, 1984(64)	8 (22–28 years)	570 – 620 mg BA	IV	9.5 ug B/g WB – 19 ug B/g WB*	No Toxicity
Human Vaginal Exposure
Van Slyke et al, 1981(44)	8	600–1200 mg IBA daily (105–210 mg B daily)	Intravaginal	0.40 ug B/g WB average in those using 600–1200mg IBA** 0.14 ug B/g WB at 1–18 hours in one individual after use of 600mg IBA	No Toxicity

Of note, blood boron levels were detected by spectrophotometry in older studies, and by the more precise inductively coupled plasma mass or emission spectrometry beginning in the mid-1980s. B=boron, WB=whole blood. Across two large, retrospective case series of ingestion in children and adults(57),(61) the lowest blood boron level reported in an adult with symptoms was 20 ugB/g WB. However a patient with levels of 75.7 ugB/g WB was asymptomatic,(57) and one with levels of 2196 ugB/g WB only had vomiting and facial flushing. A separate case(58) reported a death from ingestion with a blood boron level of 35.7 ug/g WB.

^*Blood concentrations increased from <0.095–0.36 ug/g B to a peak of 9.5 – 19 ug/g B whole blood (WB) 25 minutes later. The estimated half-life was 21 hours.

^**Prior to the main study, they administered 600 mg IBA to 8 uninfected volunteers 1–2 times daily for 1–2 weeks and measured serum boron levels with a detection limit of 0.04 ugB/g WB. They reported serum levels <1.0 ugB/mL (<0.95 ugB/g whole blood) with a mean level 0.42 ugB/mL (0.40 ugB/g WB) during use, and undetectable levels before and 48 hours after use. A single (uninfected) woman had a level of 0.15 ug/ml (0.14 ugB/g WB) at 1–18 hours and 0.1 (0.095 ugB/g WB) at 24–36 hours after using one 600 mg IBA, from which they calculated approximately 6% absorption systemically from intravaginal use

In the animal literature, Weir established an LD50 (the amount of material causing death of 50% of test animals) of 3,450 mg BA/kg (603 mg B/kg) administered as an oral dose for male Sprague-Dawley rats, and 4,080 mg BA/kg (713 mg B/kg) (i.e. 285,600 mg BA in a 70 kg human) for females(66). No adverse effects were seen in animals fed “350 ppm” boron equivalent daily for 2 years, but conversion to B/kg is uncertain. Sprague-Dawley rats are the most commonly used animal model given their relative sensitivity to BA(67,68). (See reviews by Moore, Hadrup and Bolt (65,69,70)).

Toxicity from topical BA has been reported(65). In the late 1800s, BA was included in skin ointments used for diaper rash in infants(62). Unfortunately, this led to several cases of toxicity and even death(56,62). A number of controlled studies were performed to assess the degree of systemic absorption from BA application to skin(71–74). These studies found systemic absorption was minimal through intact skin, but application of BA containing acqueous jelly to severely damaged skin could lead to higher absorption.(73,74) No clear toxic dose was defined (Suppl. Table 3).

What is the systemic absorption from IBA use in women?

In 1974, Swate(42) treated 40 women with symptomatic VVC with 600 mg IBA twice daily, and 5% BA ointment to the vulva TID. Among 20 participants none had detectable serum BA levels after 14 days of IBA. However, contemporaries raised concern that the detection threshold of 10 mg/100 mL, (16.55 ug B/g WB) was not sensitive enough to be relevant(75). Van Slyke(44) performed a trial of IBA. They administered 600 mg IBA to 8 uninfected volunteers. Based on a single (uninfected) woman they calculated 6% absorption systemically from intravaginal use. Shinohara reported no detectable systemic levels in a woman who used IBA 600 mg BID for 10 days; no information on detection threshold or time frame was provided.(41)

Summary:

Available information suggests low systemic absorption of IBA based on 600 mg suppositories. However, data are limited, primarily to 8 uninfected volunteers, with an estimate of 6% absorption based on data from a single individual(44). Long clinical experience (in the context of VVC, where inflammation might be expected) would suggest little danger of toxicity at least when 600mg IBA is used daily or BID, however the exact amount of systemic absorption in women with inflamed vaginal mucosa or mucosal ulceration is uncertain. While there are reports of toxicity and increased absorption of BA from damaged skin, it is unknown how applicable this is to the vaginal mucosa. In contrast to IBA, BA ingested orally appears to be nearly 100% absorbed though it is quickly renally cleared. Doses of 15–20 g in adults are considered “potentially lethal” however studies have not established a clear oral dose limit in mg/kg or measured serum level resulting in either toxicity or death. The absolute amounts consumed orally in adults or the equivalent in animals associated with adverse effects appear much larger than the individual 600 mg IBA treatments used daily or BID, (even if those were to be 100% absorbed- IBA may not be), and indeed healthy volunteers who ate or had comparable doses of around 600 mg BA injected IV suffered ill effects. How this may relate to individuals with poor renal function or to extended IBA use is unclear. It appears that expected blood boron levels after use of IBA would be lower than those reported to be associated with toxicity, but data are limited (Table 2).

What are the data on the safety of IBA use in pregnancy?

Teratogenicity in human studies of IBA in pregnant women:

There are few studies of the impact of IBA use in human pregnancy. Acs (2006) reported data from the Hungarian Case Control Surveillance of Congenital Abnormalities of 1980–1996.(29) 22,843 infants with congenital abnormalities were compared to 38,151 controls. 43 mothers of children with abnormalities had used IBA during pregnancy, compared to 52 controls (prevalence OR 7.1 (95% CI: 1.7–29.5)). The authors report most women used 60 mg BA daily for 7 days (standard dosing is 600 mg). Two women treated with IBA at some point in pregnancy had children with skeletal abnormalities, compared to 0 controls (however authors note they used all population controls, not only matched control pairs for this calculation). Two women using IBA in their second and third months of pregnancy had children with neural tube defects, (prevalence OR 8.0 (95% CI: 1.7–37.8)). Overall, there was a 2.8 (95% CI: 1.1–7.1) fold increase in risk of congenital abnormalities when women were exposed to IBA in months 2–3 of gestation, and 1.6 (95% CI 1.0–2.4) for the entire pregnancy(29). This study was significantly limited by its retrospective nature (recall bias); also the absolute number of children with congenital abnormalities born to women who used BA was low, and the indication for IBA use was not well defined.

Heinonen reported on a case series (conducted 1958–65 in the United States) where >50,000 pregnancies were followed and women were queried about use of a large number of medications during pregnancy, including “topical” BA.(76) This study found 19 congenital abnormalities in the offspring of 253 mothers who used BA in the first 16 weeks. There were 13 major malformations (vs. 7.4 expected) yielding a standardized relative risk ratio (SRR) of 1.75, CI: 0.94–2.94. There were 5 minor malformations observed versus 4.6 expected, SRR: 1.10, CI: 0.36–2.53. For use of BA at any time during pregnancy, there were 15 abnormalities observed vs. 8.86 expected, SRR: 1.69, CI 0.95–2.76. None of these associations was statistically significant; authors noted the association between BA use in the first 4 months and major malformations might merit further investigation. Details were not available on dosage,duration or indication of topical BA, or if it was even intravaginal.(76)

Data from pregnant women exposed to Boron other than IBA:

Data from pregnant women exposed to boron in the water supply has been published(70). In a study based in Argentina and Chile, Harari reported no major differences in birth outcomes between women living in San Antonio de los Cobres with a higher average maternal plasma level of boron as compared to a group in Santiago(77). Igra reported that in 180 pregnant women living in the Andes,(78) there was a statistically significant decrease in birth length and weight when maternal blood boron levels increased.(44) One criticism, though the authors attempted to adjust for it, was that the water supply also contained high levels of lithium-which was independently associated with a decrease in birth weights.(79) Finally, Duydu(79) found no differences in birth outcomes in boron exposed women in Turkey comparing those with low, medium, and high blood boron levels (Suppl. Table 4).

Boron toxicity in pregnant animals

Dosing pregnant animals with very large amounts of oral BA daily for an extended time can lead to significant fetal problems.(67,68,80–83) Initial studies by Heindel(68) in Sprague-Dawley rats were refined by Price(80) who found no adverse effects seen in the offspring of maternal rats (no-observable-adverse-effect level (NOAEL)) fed 55.1 mg BA/kg/day or less from gestational day 0 to 20 (Table 3).The Lowest Observed Adverse Effect Level (LOAEL)-i.e. the lowest dose at which an observed adverse effect level was demonstrated on the fetuses was 76 mgBA/kg/day or 13 mgB/kg/day. Increased skeletal abnormalities were observed in the offspring of pregnant rats treated with this amount or more. An NOAEL of 55.1 mg BA/kg/day, (10 mgB/kg/day), correlated with serum levels of 1.27 ugB/g whole blood in the maternal rats, though(67) these levels may be underestimated because there was a time lag of up to eight hours between BA administration and acquisition of blood samples. Subsequent studies reported fetal skeletal abnormalities, decreased weight and visceral abnormalities when pregnant animals were given large amounts of BA(81–83). The Code of Federal Regulations (21 CFR 170.22) for food additive procedural regulations typically necessitates a safety factor of 100 be applied to the NOAEL from animal data before extrapolating to the human data for determining a tolerable daily intake (TDI) in humans.(84) The U.S. Environmental Protection Agency has established an oral reference dose (RfD) of 0.2mgB/kg/day as the dose which is presumed to pose no risk of adverse effects during a lifetime of exposure.(85) The WHO has established a TDI of 0.17mgB/kg.(86,87)

Table 3:

Selected Studies NOAEL, and LOAEL for BA toxicity established in Sprague-Dawley rats

		Estimated daily Boron or BA intake in mother (in same units throughout)	Boron or BA levels (whole blood)	Reported Toxicity	Comments
Rat Oral Exposure
Price, CJ 1997(67)	NOAEL	10 mg B/kg/day orally	1.27 ug/g WB	No toxicity at this level	Reported in Rats fed BA 55mg BA/kg/day from gestational day 0–20 (3850mg in a 70 kg individual). Absorption expected to be 100%
Price, CJ 1997(67)	LOAEL	13 mg B/kg/day orally	1.53 ug/g WB	At this level, the main toxicity seen was decrease in fetal birth weight, in addition to some skeletal defects	Absorption expected to be 100%
Human Intravaginal Exposure
Van Slyke 1981(44)		600–1200 mg IBA daily (105–210 mg B intravaginally daily) 1.6–3 mg B/kg/day in a 70kg human	0.40 ug/g WB	N/A	Absorption expected to be less than 100%. Per Van Slyke absorption estimated at 6% which might be estimated at .09–0.18 mg B/kg/day.

NOAEL= no-observable-adverse-effect level (i.e. no adverse effects seen in the offspring of maternal rates fed boric acid). LOAEL=Lowest Observed Adverse Effect Level (i.e. the lowest dose at which an observed adverse effect level was demonstrated on the fetuses).

Summary:

Data from one(29) and less clearly a second(76) human study have raised concerns regarding the teratogenic potential of IBA. These studies were limited by methodologic concerns and small absolute numbers of birth defects observed. No data on other obstetric complications were available. Studies of birth outcomes in women consuming boron in drinking water are mixed, and may be limited by confounders; it is unclear how well these chronic daily exposures throughout pregnancy would relate to the short term or intermittent exposure resulting from IBA regimens. High doses of oral BA given daily in pregnancy clearly have teratogenic effects in animal models; data on other obstetric outcomes are scarce. There are no studies specifically assessing IBA in pregnant animals. The equivalent amounts of BA administered orally in animals appear greater than the typical IBA dose prescribed, even if IBA were 100% systemically absorbed, though safety factors must be applied. If estimates of absorption by Van Slyke are accurate at 6%, the expected daily exposure from IBA (0.096–0.18 mgB/kg/day) might be similar to established RfD and TDI levels, but given the limitations of the data no definitive conclusions can be drawn.

DISCUSSION

Herein, we attempted to summarize relevant data on the safety of IBA for women and in pregnancy. There were several limitations to our paper. This was a narrative rather than a systematic review, only conducted through 2019. We abstracted data from all available papers regarding use of IBA. However, since others have comprehensively reviewed extensive literature relating to oral ingestion and non-vaginal topical or environmental exposure to BA/boron (65,69,70), we summarized only key papers on these topics. We did not find any literature regarding effects on fertility of IBA use. Data on effects of oral boric acid or boron on female fertility(70) in humans(88) or animals(66,89) were scant, difficult to disentangle from effects on male fertility, and were not extensively reviewed(69,70). Finally, we relied on published literature; it is possible that adverse events relating to IBA use have occurred but have not been reported.

Overall, lack of reported toxicities despite use of IBA for inflammatory conditions such as VVC is reassuring. Though limited, oral toxicity studies suggest that the amount of IBA required to induce significant toxicity in women would likely be more than the intravaginal doses commonly described in the literature as being prescribed by clinicians to healthy adults with normal renal function. Less data is available on the safety of longer term or suppressive regimens prescribed though significant toxicities have not been reported(3,8,9). Of note, the 600mg dose of IBA is largely arbitrary-no dose-range studies to evaluate safety or efficacy have ever been performed, mechanisms of action are uncertain, though under investigation(90–92), and there are no accepted standardized susceptibility tests for vaginal pathogens such as yeast or trichomonas.

Several guidelines advise (ACOG, Canada, BASSH) that IBA should be avoided in pregnancy. Human data raises some concerns regarding teratogenic effects of IBA, however it is sparse and flawed. High level oral exposure to BA in animals is teratogenic. The level of exposure from IBA is likely much lower, however trying to relate animal data to the human scenario is limited in part by uncertainty regarding systemic absorption of IBA under conditions of inflammation. While pregnant and non-pregnant women(93) and animals(94) appear to clear boron at similar rates, how else the physiology of pregnancy might impact IBA effects is unknown. We were not able to find data addressing maternal absorption of IBA, what amount of boron might cross the placenta with IBA use, or the impact of IBA on obstetrical outcomes other than teratogenicity. Therefore it is not possible to recommend, based on existing data, changes to guidelines advising against IBA use in pregnancy.

45% of pregnancies in the US are unplanned(14); half are diagnosed after the first six weeks.(95) Thus, women of reproductive potential are ideally informed about possible adverse pregnancy related effects of medications taken. Isotretinoin(96) is associated with birth defects in any amount-prescribers and patients are held to special requirements to ensure pregnancy avoidance. In comparison, while providers may routinely counsel women on risks of NSAIDS, acetaminophen or fluconazole (which have more robust evidence for adverse pregnancy effects than IBA),(97–106) there are no similar requirements to isotretinoin that all women using these medications have antecedent pregnancy testing, repeated testing during use, or sign a commitment to use effective contraception. Such stringent measures would not seem appropriate for IBA given available data. It may be reasonable to discuss the sparsity of data on pregnancy and the drug threshold of embryopathy in IBA to facilitate a joint informed-consent process to consider pregnancy avoidance while in use or defer if pregnancy is planned. Also, women should be counselled to use the medication only intravaginally and to keep it out of reach of children to avoid accidental ingestions.

Our findings and analysis are focused on 600mg IBA regimens commonly described in the literature as prescribed by clinicians. However, given the ease of obtaining IBA OTC, women may be self-medicating with non-recommended regimens, with products of uncertain dosage and purity, and may inadvertently use IBA in early pregnancy. Toxicity and safety issues aside, given widespread availability (and direct to consumer marketing for a variety of indications (some nonspecific or unsupported by studies)), IBA may potentially be overused by healthy women for perceived problems. Alternatively in those with true urogenital infections or other pathology, IBA self-treatment may mask disease manifestations, delaying needed medical care or making accurate clinician diagnosis more difficult. Similarly, uninformed clinicians may misdiagnose patients or overprescribe IBA for unsupported indications or regimens.

Conclusions:

Despite limitations, available evidence suggests IBA use at dosages commonly described in the literature is likely safe for women, at least those with normal renal function. However we feel IBA should be prescribed by a provider; given uncertainties, we can not recommend unregulated use of OTC products. Clinicians should initiate a discussion for a joint informed-consent process with patients on pregnancy avoidance with IBA use. Additional public and clinician education regarding IBA is needed. In the long term, additional regulation (e.g of OTC availability, marketing claims, and product quality) may be needed.