Safety and tolerability of Pau d' Arco (Tabebuia avellanedae) for primary dysmenorrhea: A single-arm, open-label trial on adults ages 18-45.

This study was a single arm, open-label trial conducted at the National University of Natural Medicine (NUNM; Portland, OR) following CONSORT guidelines. This clinical trial was registered on ClinicalTrials.gov ( NCT04245540 ) and approved by the Institutional Review Board (IRB# RB11617) at NUNM. A convenience sample of 12 generally healthy women aged 18–45 with PDM was recruited via electronic postings and flyers advertised on the NUNM campus, clinics, and website. A telephone screening determined initial eligibility in the study and blood tests confirming general health measures in the normal range allowed for enrollment in the study. Once enrolled, participants picked up study supplies from NUNM and participated in three study visits at Week 0 (first menstruation after enrollment), Week 4, and Week 8. Biologically female; ages 18–45; presence of PDM; had a ‘smart’ phone; willing to download and use the REDCap© app to complete study surveys and receive study reminders; willing to take a non-hormonal form of birth control throughout the trial period; lived in the Portland area; willing to consume bovine gelatin capsules ; able to speak, read and write English; had reliable transportation to clinic; willing to have four blood draws taken; pain scale rating of 6 or higher on the Visual Analog Scale (VAS) on the first day of menstruation; monthly pain that correlates with menstruation; regular menstrual cycle (21–35 day cycle). Exclusion criteria: Presence of secondary dysmenorrhea; general health measures outside of the normal range: blood pressure, aspartate aminotransferase (AST), alanine aminotransferase (AST), estimated glomerular filtration rate (eGFR), international normalized ratio (INR), red blood cell count , hemoglobin, hematocrit, white blood cell count, and/or platelet counts; women who were nursing, pregnant, or planning pregnancy in the next four months; difficulty swallowing or aversion to capsules, tablets, or pills; currently taking, and unwilling to discontinue NSAID medication, currently taking and unwilling to discontinue dietary supplements with any of the following ingredients in amounts over 250 mg/day: cinnamon, garlic, ginger, turmeric, and/or curcumin ; consumption of > 1 alcoholic drink per day during the study period; multi-System Adverse Event (AE) Form score(s) of 3 or more (indicative of pre-existing severe symptoms); currently taking any anticoagulation medications (e.g., Warfarin/Coumadin); currently taking, or had taken in the past two months, hormonal forms of birth control; dysmenorrhea reasonably suspected to be due to a copper IUD; unwillingness or inability, to take a monthly pregnancy test during the timeline of the study; currently enrolled, or planning to enroll, in another clinical trial; past or present medical history of any of the following: blood clotting disorders , liver disease, kidney disease, cardiac disease, anemia, inflammatory bowel disease , irritable bowel syndrome , endometriosis , obstructive endometrial polyps , chronic pelvic inflammatory disease , polycystic ovarian syndrome , adenomyosis , intrauterine or pelvic adhesions, congenital obstructive Mullerian malformations, cervical stenosis, use of an intrauterine contraceptive device that causes pain, pelvic congestion syndrome , reproductive cancer (uterine, ovarian, etc.), ovarian cysts , fibroids, and/or uteropelvic junction obstruction. TA was purchased from Santegra, USA, a Good Manufacturing Practice (GMP) certified facility. This product is commercially available and consists of TA (350 mg/capsule) and rutin (25 mg/capsule) in gelatin capsules with 3% naphthoquinone content in each capsule (batch number and quality analysis available on request from supplier). Participants were given 120 capsules (40 days of intervention) to take home at Week 0 and at Week 4 and asked to consume 3 capsules (1050 mg TA) per day with food. Compliance was measured by asking participants to bring any remaining study intervention capsules with them to each study visit and recording the remaining capsules. The dosage of TA was determined through allometric scaling from mouse models in which TA has been found to be anti-inflammatory, reduce edema, and to decrease pain responses [ 11 ], [ 12 ], [ 13 ], [ 16 ], [ 19 ], [ 20 ], [ 21 ], and because it was well below the dose of lapachol previously shown to cause anticoagulation [ 24 ]. Rutin is a bioflavonoid found in many common foods; for example, apples and barley [ 25 ], [ 26 ], [ 27 ]. The dose of 75 mg/day of rutin in this product is considered unlikely to create an anti-inflammatory effect in this study because it is well-below the level associated with a decrease in inflammation (50 mg/kg/day) [ 28 ]. All study participants signed an informed consent form at screening and were screened for general health by assessing blood pressure and liver and kidney function via a complete metabolic panel (CMP) and confirming the absence of anemia via a complete blood count (CBC). Participants were also screened for normal coagulation parameters through an international normalized ratio (INR) test at screening, Week 4, and Week 8. All samples were processed by Quest Diagnostics. Adverse events (AEs) were assessed using a standardized, multi-system AE questionnaire at each study visit. Study personnel adhered to strict individual and overall study stopping guidelines in the case of certain adverse events or predetermined levels of general health measures outside of the normal range. Pregnancy tests (First Response ™ ) were completed each month on the first day they were determined to be accurate, based on the First Response ™ website calculator, beginning at the screening visit and continuing until completion or withdrawal from the study. Pregnancy tests were confirmed either by direct visualization by study staff or by a picture sent by the participant to study staff via Research Electronic Data Capture (REDCap©). REDCap© is a HIPAA-compliant web platform designed for data management, including the remote collection of data [ 29 ], [ 30 ]. Pregnancy at any point throughout the study would have resulted in immediate withdrawal. Pain intensity was measured by a validated Visual Analog Scale (VAS) [ 31 ], [ 32 ]. Assessments were taken during the telephone interview, at screening to assess baseline pain levels and via REDCap© at subsequent measurements. Subsequent measurements included: 4 h after taking the first dose of the study intervention on the first day of menstruation (Week 0), at Week 4, and at Week 8. Four hours were selected to allow for investigation of a potential mechanism of action of TA (PGE 2 inhibition) [ 2 ]. Quality of life and pain interference were measured through the validated Patient-Reported Outcomes Measurement Information System (PROMIS) 29 survey [ 33 ] and collected at baseline (Week 0) and at Week 8. Sexual function, a quality-of-life indicator that is negatively impacted by dysmenorrhea, was assessed through the validated PROMIS Sexual Function and Satisfaction: Interfering Factors [ 34 ] at baseline (Week 0) and at Week 8. All study surveys were administered and recorded via REDCap©. Blood samples for analysis of high sensitivity C-reactive protein (hs-CRP) were collected at Week 0 and Week 8 to assess for systemic inflammation. Power calculations were based on the secondary outcome measure of pain intensity to assess the clinical significance of any measured changes, which is a critical requisite to justify future clinical research, and because of the subjective nature of determining power based on a number of safety measures. A sample size of 12 was calculated to provide 80% power to detect a mean VAS score reduction of 11.9 points, using an unadjusted α = 0.05 and an assumed standard deviation of 17.3 and correlation r = 0.7 between pre- and post-measures [ 35 ], [ 36 ], [ 37 ]. A minimum 20-point reduction on the VAS scale has been associated with clinically significant outcomes in pain populations. [ 35 ], [ 38 ] All measures were scored using paired t-tests. For PROMIS scales, the sample size provided roughly ~50% power to detect a reasonable effect of d= 0.5. All necessary precautions were taken to maintain the safety and confidentiality of the participants. All participants signed informed consent forms before enrolling in the study. De-identified codes were used to protect confidentiality and all study staff received training in HIPAA regulations from NUNM. The first four authors took part in enrolling participants and assigned de-identified codes in sequential order from enrollment into the study. All participant information is stored in a locked file cabinet in a locked room within NUNM or in password-protected computers only accessible by study staff.

Participants were recruited between July 2019 and February 2020 and the study concluded in May 2020 ( Fig. 1 ). Thirty-three individuals were screened, resulting in 36.4% enrollment. Seventy-five percent of participants completed the study (n = 9/12); two (17%) participants voluntarily withdrew citing intolerability to the study intervention and one participant was withdrawn due to non-adherence to the study protocol. Mean age of participants was 31.7 ± 7.3 years, and the majority of participants were white. Race of participants was identified as white (n = 7), Asian (n = 2), Latino (n = 1), and 2 + races (n = 2). All participants had laboratory blood markers within the normal range at baseline ( Table 1 ). Five participants (n = 5/12; 41.7%) had abnormalities in laboratory values after study intervention at Week 4 or Week 8 including abnormal liver enzymes, potassium level, urea nitrogen, and markers of anemia ( Supplementary Table 3 ). No laboratory values exceeded predetermined levels for individual or collective study stopping criteria. Neither body mass index (BMI) nor blood pressure changed significantly during the study (p > .05) ( Supplementary Table 2 ). No serious adverse events occurred during the study. Nine of the twelve study participants (75%) reported an adverse event including bruising at venipuncture site (n = 1), discomfort with menstrual cup insertion (n = 1), momentary dizziness (n = 1), myokymia (n = 1), diarrhea (n = 2), nausea (n = 1), vomiting (n = 1), fever (n = 1), epistaxis (n = 1), and intermittent dyspareunia (n = 2) ( Supplementary Table 4 ). Almost all adverse events reported were mild and none were characterized as severe. There was no significant difference between pain levels reported during the telephone interview and pain levels reported during the screening visit, demonstrating reliability in baseline pain intensity levels. There was a statistically significant decrease (improvement) in pain intensity from baseline (screening, mean 74.17) at Week 0 (first menstruation during study timeline, mean 41.33; p < .01); Week 4 (mean 49.45; p < .01; 95% CI [10.1, 38.3]), and Week 8 (mean 43.88; p < .01; 95% CI [10.5, 46.8]) ( Fig. 2 ; Supplementary Table 5 ). At Week 0, eleven of the twelve (91.7%) study participants reported a decrease in pain from baseline pain levels. At Week 4, ten of the eleven remaining participants (90.9%) reported a decrease in pain from baseline pain levels. At Week 8, eight of the eight remaining participants (100%) reported a decrease in pain from baseline pain levels. There were no significant differences in pain intensity seen between Week 0 and Week 4 (p = .50; 95% CI [−34.0, 17.8]), between Week 0 and Week 8 (p = .84; 95% CI [−33.7, 28.2]), or between Week 4 and Week 8 (p = .39; 95% CI [−14.4, 32.9]). Five out of the twelve participants reported taking medications at some point throughout the trial for pain (NSAIDs, medical marijuana, essential oils, electrolyte powder, and ginger). None of these medications or supplements were reported as used for the first time ( Supplementary Table 6 ). Improvement in mean scores was observed in six of the seven domains on the PROMIS 29 with a slight decline seen in one domain over the course of this study ( Fig. 3 ; Supplementary Table 7 ). There were nonsignificant decreases (improvement) in the subscales anxiety/fear (p = .34; 95% CI [−6.4, 16.0]), depression/sadness (p = .22; 95% CI [−3.1, 11.3], fatigue (p = .16; 95% CI [−2.5, 12.5]), sleep disturbance (p = .51; 95% CI [−5.6, 10.6]), and pain interference (p = .42; 95% CI [−7.4, 15.9]); and increases (improvement) in the subscales ability to participate in social roles/activities (p = .11; 95% CI [−14.2, 1.9]). Physical function (p = .47; 95% CI [−9.6, 4.9]) decreased (worsened) from Week 0 to Week 8 ( Fig. 3 ; Supplementary Table 7 ). Sexual function and satisfaction scores improved nonsignificantly from Week 0 to Week 8 (p = .09; 95% CI [−14.1, 1.3]) ( Fig. 4 ; Supplementary Table 7 ). Systemic inflammation decreased (improved) nonsignificantly (p = .22; 95% CI [−3.8, 11.0]) between Week 0 and Week 8, with a mean decrease of 1.16 milligram per liter (mg/L; Supplementary Table 8 ).

This study is the first study assessing TA at a dosage of 1050 mg per day for eight weeks in women ages 18–45 with PDM, which was associated with general safety and moderate tolerability for most study participants, and a statistically significant improvement in pain intensity scores. The small size of this study, number of mild adverse events, and few clinically mild abnormalities warrant future investigation on the safety and tolerability of TA with larger sample size. A placebo-controlled, randomized clinical trial assessing the effects of TA on decreasing pain intensity, improving QOL, and improving pain interference is warranted to determine the efficacy of TA on improving PDM in women of reproductive age.

This is the first investigation of the safety and tolerability of 1050 mg per day of TA for women ages 18–45 with PDM. This study met recruitment goals but had moderate attrition (25%), due to perceived intolerability to the study intervention (2/12) and noncompliance (1/12). Participants (2) who reported intolerability to the study intervention cited diarrheal episodes (1) and muscle spasms (1) as the reasons for attrition from the study. It is unknown if these symptoms were related to the intervention and future studies should allow for reporting of these symptoms and assessment of any possible association. Several participants had abnormalities in general health laboratory values after study intervention of which most were mild and were confounded by other factors such as concurrent illness. Only abnormal laboratory markers consistent with anemia showed spontaneously in more than one participant after taking TA, suggesting a need to monitor for this potential effect in future clinical trials and in clinical management of those taking TA. A number of mild adverse events occurred during this study that were unlikely to be due to the study intervention and more likely due to study participation (e.g., bruising at the venipuncture site). A single participant had episodes of epistaxis but normal INR values at all study time points, thus reducing the clinical concern for anticoagulant activity. Diarrhea occurred in one participant almost 4 weeks after starting the study intervention and in a second participant concurrent with fever, nausea, and vomiting, suggesting the possibility of non-study related etiologies. Intermittent dyspareunia with sexual intercourse was seen in two participants which were not reported in previous clinical trials. This warrants the inclusion of future safety assessments with a focus on dyspareunia. A significant reduction in pain intensity was seen within four hours of administration of the first dose of TA, which is similar to the response time one would expect from NSAIDs [ 2 ], supporting the hypothesis that PGE 2 reduction may be at least one mechanism of action of TA on pain intensity. Nonsignificant improvements were noted in pain interference, quality of life, sexual function and satisfaction scores, and in measurement of systemic inflammation and should be investigated further in future studies. Several limitations were present in this study. Most notably, laboratory values were only able to be obtained for 50% of participants at Week 8, due to both withdrawals from the study and stay at home orders enacted at the onset of the COVID-19 pandemic. Adherence was not assessed for all participants due to an inability to conduct in-person study visits during the pandemic. However, survey measures were still assessed due to the remote method of data collection via REDCap©. Additionally, a product without the addition of rutin would have been preferable, but no product that included TA alone with appropriate standardization could be found at the time of this study. Other limitations included a small sample size, a lack of a placebo or standard of care arm, and use of NSAIDs and ginger by some study participants. Data from participants who use additional medications should be excluded from analysis in future studies with larger sample sizes. Strengths include assessment of safety data, quality of life, pain intensity, and pain interference for the first time with this botanical in this study population and continued study follow-up on remotely collected data measures despite a pandemic. Safety data collected in this trial that may be utilized to inform a larger placebo-controlled trial. Future studies should include pharmacokinetic studies and dose escalation studies to assess the most efficacious dose and dosing regimen of this botanical, preferably measuring pain at further time intervals to provide further understanding of the duration of effects of this intervention.

Primary dysmenorrhea (PDM) is defined as pelvic cramps experienced during the first days of the menstrual cycle [ 1 ], [ 2 ], [ 3 ], [ 4 ], [ 5 ], [ 6 ], [ 7 ], [ 8 ] and affects 41–90% of reproductive-age women, making it the most common gynecological disorder in this population [ 1 ], [ 2 ], [ 3 ], [ 4 ], [ 5 ], [ 6 ], [ 7 ]. PDM is caused by an increase in inflammatory prostaglandins , including prostaglandin E2 (PGE 2 ), produced by the enzyme cyclooxygenase-2 (COX-2), which cause uterine contractility and ischemia [ 1 ], [ 2 ], [ 3 ], [ 4 ], [ 5 ], [ 6 ], [ 9 ]. PDM accounts for 600 million lost working hours and $2 billion in lost productivity each year in the United States [ 3 ], [ 7 ], making it a public health priority. Tabebuia avellanedae (TA; pau d′ arco; taheebo bark) is a native Bignoniaceae tree located in Central and South America, the inner bark of which has been used medicinally for over 1000 years.[ 10 ], [ 11 ] TA and its active constituents, lapachol and β-lapachone, have been shown in vitro to inhibit the expression of PGE 2 and COX-2 in a concentration-dependent manner [ 10 ], [ 12 ], [ 13 ], [ 14 ], [ 15 ], [ 16 ], [ 17 ], [ 18 ]. In vivo , TA and its constituents have been shown to be anti-inflammatory, reduce edema, and to decrease pain responses in mice [ 11 ], [ 12 ], [ 13 ], [ 16 ], [ 19 ], [ 20 ], [ 21 ]. TA is of particular interest for use in PDM because in vitro and in vivo studies suggest that this botanical may have a similar mechanism of action as nonsteroidal anti-inflammatory drugs (NSAIDs), but without the gastrointestinal toxicity associated with these drugs [ 3 ], [ 4 ], [ 5 ], [ 6 ], [ 7 ], [ 9 ], [ 12 ], [ 21 ]. Human clinical trials on TA for cancer treatment have not been associated with major adverse effects [ 22 ], [ 23 ], yet this botanical has not been assessed for an anti-inflammatory effect in a clinical trial for PDM. The purpose of this study was to examine the safety and tolerability of TA in a sample of healthy reproductive aged women, and to conduct an exploratory analysis of the effects of TA for pain reduction in PDM.