Abstract
Background
Tamoxifen, clomiphene, and letrozole are primary pharmacological options for inducing ovulation in polycystic ovary syndrome (PCOS). Given the high clinical prevalence of PCOS, a comprehensive characterization of the adverse drug event (ADE) profiles associated with these treatments is essential.
Methods
We performed disproportionality analyses using FAERS data to detect ADE signals at the Preferred Term (PT) and System Organ Class (SOC) levels via four established algorithms (ROR, PRR, BCPNN, and EBGM). A drug-ADE network was constructed to visualize these associations.
Results
Of 21,730 identified PCOS-related reports, letrozole predominated (n=17,185), followed by tamoxifen (n=4465) and clomiphene (n=80). Most cases involved women aged 18–65 years (weight: 50–100 kg), primarily from the United States. Letrozole-related reports increased steadily, peaking at 2323 cases in 2021, while tamoxifen and clomiphene counts remained comparatively low. At the PT level, fatigue was a common signal for both tamoxifen and letrozole. Notably, malignant tumor progression, fatigue, and arthralgia emerged as shared signals across all three agents. At the SOC level, ADEs for tamoxifen and letrozole were frequently categorized under neoplasms; letrozole was specifically linked to hematological disorders (e.g, neutropenia). In contrast, clomiphene exhibited stronger associations with psychiatric and gastrointestinal events.
Conclusion
Distinct safety signals characterize these three primary PCOS treatments. These pharmacological variations underscore the necessity of personalized medicine; treatment selection must be tailored to the patient’s specific risk profile, with targeted clinical monitoring for relevant ADEs to optimize therapeutic outcomes.
Keywords: adverse events, food and drug administration adverse event reporting system, polycystic ovary syndrome, tamoxifen, clomiphene, letrozole
Introduction
PCOS is a prevalent endocrine and metabolic disorder1 affecting 4–21% of women globally.2 Its manifestations include menstrual irregularities, dermatological issues (acne, hirsutism), metabolic disturbances (hyperinsulinemia), reproductive dysfunction (infertility), and psychological symptoms (mood disorders).2 Per the 2003 Rotterdam criteria (ESHRE/ASRM), PCOS is diagnosed with at least two of: (1) oligo-/anovulation, (2) hyperandrogenism, (3) polycystic ovaries on ultrasound. While its exact etiology is unclear, growing evidence links genetic susceptibility and environmental factors to its development. PCOS treatments mainly include lifestyle interventions, pharmacotherapy, and surgery. Pharmacotherapy, the cornerstone of management, is often characterized by marked interindividual variability in efficacy, notable adverse drug events (ADEs), and limited indications, underscoring the clinical and scientific importance of investigating its safety profile.
International guidelines recommend letrozole and clomiphene as first-line ovulation-inducing agents for PCOS women,3 but their mechanisms, efficacy, and applications differ greatly. Clomiphene, a selective estrogen receptor modulator (SERM), blocks hypothalamic estrogen receptors to boost FSH/LH secretion; however, down-regulated receptors may cause adverse endometrial and cervical mucus changes, potentially impairing embryo implantation despite successful ovulation. Letrozole, an aromatase inhibitor, suppresses estrogen synthesis, thereby reducing negative feedback on the hypothalamic-pituitary axis and promoting endogenous FSH secretion to stimulate folliculogenesis. Due to its estrogen-lowering effects, it is mainly used as adjuvant therapy for hormone receptor-positive (HR+) breast cancer. While letrozole better improves ovulation and pregnancy rates, its ADEs and risks—such as frequent neutropenia during therapy, as noted in the Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) database—should not be ignored. Tamoxifen, a SERM with a mechanism similar to clomiphene,4 is primarily used to treat female breast cancer. In PCOS, it is indicated for patients who failed clomiphene-induced ovulation or had pregnancy failure due to clomiphene-impaired endometrial receptivity. The French Pharmacovigilance and Addictovigilance Database noted thrombosis cases linked to clomiphene in male hypogonadism; additionally, multiple reports showed acute cardiovascular complications in ovulatory dysfunction women after initial clomiphene exposure.5–8 Previous studies on tamoxifen’s adverse events focused mainly on female breast cancer patients, with limited data on toxicity in male breast cancer patients.9 However, a comprehensive comparison of their post-marketing safety profiles specifically in PCOS patients remains lacking. However, it is essential to recognize the interpretative boundaries of pharmacovigilance data. FAERS-based analyses identify safety signals—statistical associations that indicate an event is reported more frequently than expected—rather than confirmed adverse drug reactions. These signals are inherently subject to limitations such as reporting bias and confounding by indication, and they do not establish a definitive causal relationship. Acknowledging these constraints is crucial for a balanced interpretation of the safety profiles of ovulation-induction agents. This study aims to fill this gap by leveraging the FAERS database to systematically characterize and compare the adverse event profiles of tamoxifen, clomiphene, and letrozole in PCOS management.
This study used data from FAERS, the world’s largest pharmacovigilance database. As a post-marketing drug safety surveillance system, FAERS collects global spontaneous reports of ADEs since 2004, with twenty millions of cases, public access, and quarterly updates. It aggregates voluntary reports from healthcare professionals, consumers, and manufacturers (US and international) to continuously monitor the safety of FDA-approved drugs, boasting high reliability, multidimensional granularity, global coverage, and real-time updates—making it robust for pharmacovigilance research. Through analysis, it enables signal detection and association studies to identify unrecognized drug safety signals (including unlabeled adverse events), providing evidence for clinical medication safety to aid informed decisions.10–12 FAERS is indispensable for pharmacovigilance, especially evaluating post-marketing safety of tamoxifen, clomiphene, and letrozole in PCOS treatment. Without it, critical ADEs (notably rare or long-term ones) might go undetected, delaying interventions and risking patient safety. This study leverages FAERS to systematically identify and compare ADEs signals across the three drugs, filling a key gap in PCOS management. Findings enhance understanding of drug-specific risks, empowering clinicians to tailor therapies, minimize harm, and optimize outcomes for women with PCOS—thus advancing evidence-based, patient-centered care in the field.
Materials and Methods
Data Extraction
FAERS database (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html) collects ADEs reports voluntarily submitted by healthcare professionals, patients, and other sources. It is updated quarterly and publicly available for free. The data is stored in ASC or XML formats and is often used for mining ADEs signals of drugs that are already on the market. ADEs reports in the FAERS database regarding the treatment of PCOS with clomiphene, tamoxifen, and letrozole from 2004 up to 2024 were downloaded as raw data.
Using “tamoxifen”, “clomiphene”, and “letrozole” as the search terms, a search was conducted in the FAERS database. The ADEs in which tamoxifen, clomiphene, and letrozole were the drugs of primary suspect (PS) were included. The preferred terms (PT) and system organ classes (SOC) in the Medical Dictionary for Regulatory Activities (MedDRA) were used to describe and classify the relevant ADEs. The SOC is grouped according to the criteria of etiology (such as diseases of infection and infestation), site of onset (such as diseases of the gastrointestinal system), and purpose (such as various surgical and medical procedures). The PT is a specific term used to express a single medical concept, such as a certain symptom, sign, disease, diagnosis, indication, examination, surgical and medical procedure, medical history, social history, or family history. One PT corresponds to at least one SOC, and it can also correspond to multiple SOCs depending on the situation.
The downloaded American Standard Code for Information Interchange (ASCII) data files underwent data cleaning using the R package tidyr (v 1.3.1).13 The data exclusion criteria were as follows: The reasons for drug use were limited to PCOS, and cases with other clinical indications were excluded; Reports of irrelevant reactions (such as administration errors, quality complaints, and ineffective treatment) or reports containing other co-suspected products besides the drugs of interest were deleted; Duplicate reports involving age, weight, reported events, and the event occurrence date and country were deleted (Supplementary Table 1).
FAERS Data Mining
In this study, we employed disproportionality analysis to identify signals of adverse events (Supplementary Table 2). Its basic principle is to use a disproportionality fourfold table to compare the differences in the occurrence frequencies of the target drugs and target events with the background frequencies. The specific calculation formulas and thresholds of four methods, namely the Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN) method, and Empirical Bayesian Geometric Mean (EBGM), are provided (Supplementary Table 3).
The calculation formula for the ROR was: 
. The 95% confidence interval (CI) of the ROR value was calculated as: 
. When a ≥ 3 and the lower limit of the 95% CI of the ROR value was greater than 1, an effective ADEs signal was generated.
The calculation formulas for the PRR were as follows: 
, and 
. When a ≥ 3, PRR ≥ 2, and χ2 ≥ 4, an effective ADEs signal was generated.
The calculation formula for the EBGM was: 
. When EBGM05 > 2, it represented the generation of an effective ADEs signal.
The calculation formula for the BCPNN was: 
. When IC025 > 0, it represented the generation of an effective ADEs signal. While these disproportionality methods are standard for signal detection, it is important to note that the resulting signals indicate statistical associations rather than clinical causality. The analysis is subject to the inherent constraints of the FAERS database, such as reporting bias and potential confounding by indication. Furthermore, given the substantial imbalance in report numbers across the three drugs—particularly the limited cases for clomiphene (n=80)—direct comparisons between agents are exploratory and hypothesis-generating in nature.
Basic Characteristics, Temporal Trends and Gender Differences in the Treatment of PCOS
Based on the FAERS data from the establishment of the database up to now regarding the treatment of PCOS with 3 drugs, namely tamoxifen, clomiphene, and letrozole, the basic reporting information for the treatment of PCOS with these 3 drugs, such as age, gender, body weight, reporting personnel, and the countries where the reports originated, was statistically analyzed. The R package ggplot2 (v 3.4.4)14 was used to visualize the temporal trends of the reports, as well as the reporting proportions in different regions and among different genders.
Analysis of PCOS ADEs at PT Level
Based on the FAERS data from the establishment of the database up to now regarding the treatment of PCOS with clomiphene, tamoxifen, and letrozole, “ROR, PRR, BCPNN, and EBGM” were used for ADEs signal detection. The criteria for determining an effective ADEs signal were set as follows: the number of ADEs ≥ 3, the lower limit of the 95% CI of the ROR value > 1, or ADEs ≥ 3, PRR ≥ 2, and χ2 ≥ 4, or EBGM05 > 2, or IC025 > 0. All signals at the level of the PT were analyzed, and the top 30 most frequent signals were presented using the package ggplot2 (v 3.4.4).
ADEs Signal Detection at the SOC Level and the Distribution of Signals Within the SOC
Based on the FAERS data from the establishment of the database up to now, “ROR, PRR, BCPNN, and EBGM” were used for ADEs signal detection. Using MedDRA27.0, the ADEs signals of the 3 drugs were classified and analyzed according to the SOC of the affected organs and systems. The determination conditions for identifying valid ADEs signals were set as follows: the number of ADEs ≥ 3, the lower limit of the 95% CI of the ROR value > 1, or PRR ≥ 2 and χ2 ≥ 4, or EBGM05 > 2, or IC025 > 0. The number of ADEs at the SOC level for 3 drugs identified as having valid signals was visualized using the package ggplot2 (v 3.4.4). Moreover, the distribution of ADEs signals for these 3 drugs within each SOC would be displayed using the package ggraph (v 2.2.1).15
Signal Network Diagram of the Reports on Drugs and Adverse Reaction Events at the SOC Level
In order to evaluate the ADEs signals related to all 3 drugs, the top 10 ADEs signals at SOC level, which were sorted according to the number of reports, were selected, and the drug-ADEs signal network was constructed and visualized using the package ggraph (v 2.2.1).
Statistical Analysis
In this study, EXCEL 2022 was used to screen the mined data, and ADEs signals were statistically analyzed by R software (v 4.2.2). The table was read using the R package readxl (version 1.4.2).16
Results
Basic Characteristics, Temporal Trends and Gender Differences in the Treatment of PCOS
Letrozole was associated with the highest number of ADEs reports (n = 17,185), substantially outnumbering those for tamoxifen (n = 4465) and clomiphene (n = 80). The majority of reported cases involved female patients aged 18–65 years, with a body weight ranging from 50 to 100 kg. Analysis of reporter types revealed distinct patterns across the three drugs, with Medical Doctors (MDs) being the primary reporters for letrozole, consumers for tamoxifen, and other health professionals for clomiphene. Geographically, the United States was the predominant source of reports for tamoxifen, clomiphene and letrozole (Table 1).
Table 1.
Basic Characteristics Related to Adverse Drug Events (ADEs) of Three Drugs for Treating Polycystic Ovary Syndrome (PCOS)
| Letrozole (N=17185) | Tamoxifen (N=4465) | Clomifene (N=80) | |
|---|---|---|---|
| SEX | |||
| F | 15611 (90.8%) | 3751 (84.0%) | 56 (70.0%) |
| M | 205 (1.2%) | 196 (4.4%) | 18 (22.5%) |
| SEX_ Missing | 1369 (8.0%) | 518 (11.6%) | 6 (7.5%) |
| WT | |||
| <50 kg | 366 (2.1%) | 69 (1.5%) | 4 (5.0%) |
| >100 kg | 312 (1.8%) | 70 (1.6%) | – |
| 50~100 kg | 4698 (27.3%) | 1141 (25.6%) | 14 (17.5%) |
| WT_Missing | 11809 (68.7%) | 3185 (71.3%) | 62 (77.5%) |
| AGE | |||
| Under 18 years old | 58 (0.3%) | 19 (0.4%) | – |
| Older than 85 years old | 301 (1.8%) | 49 (1.1%) | – |
| Between 18 and 64.9 years old | 5302 (30.9%) | 2004 (44.9%) | 55 (68.8%) |
| Between 65 and 85 years old | 4595 (26.7%) | 830 (18.6%) | 2 (2.5%) |
| AGE_Missing | 6929 (40.3%) | 1563 (35.0%) | 23 (28.8%) |
| OCCP_COD | |||
| CN | 5004 (29.1%) | 1477 (33.1%) | 10 (12.5%) |
| HP | 2526 (14.7%) | 327 (7.3%) | 4 (5.0%) |
| LW | 3 (0.0%) | 22 (0.5%) | – |
| MD | 6486 (37.7%) | 1370 (30.7%) | 21 (26.3%) |
| OT | 1811 (10.5%) | 506 (11.3%) | 35 (43.8%) |
| PH | 1029 (6.0%) | 345 (7.7%) | 5 (6.3%) |
| RN | 3 (0.0%) | – | – |
| OCCP_COD_Missing | 323 (1.9%) | 418 (9.4%) | 5 (6.3%) |
| Reporter_Country | |||
| Albania | 1 (0.0%) | – | – |
| Argentina | 89 (0.5%) | 3 (0.1%) | – |
| Australia | 166 (1.0%) | 22 (0.5%) | 1 (1.3%) |
| Austria | 248 (1.4%) | 21 (0.5%) | – |
| Bangladesh | 22 (0.1%) | – | – |
| Belarus | 1 (0.0%) | – | – |
| Belgium | 101 (0.6%) | 51 (1.1%) | – |
| Bosnia and Herzegovina | 4 (0.0%) | 1 (0.0%) | – |
| Brazil | 158 (0.9%) | 19 (0.4%) | – |
| Brunei | 1 (0.0%) | – | – |
| Bulgaria | 11 (0.1%) | 1 (0.0%) | – |
| Canada | 1269 (7.4%) | 402 (9.0%) | 3 (3.8%) |
| Chile | 18 (0.1%) | – | – |
| China | 219 (1.3%) | 6 (0.1%) | 1 (1.3%) |
| Colombia | 42 (0.2%) | – | – |
| Costa rica | 2 (0.0%) | – | – |
| Country not specified | 827 (4.8%) | 25 (0.6%) | – |
| Croatia | 79 (0.5%) | 1 (0.0%) | 1 (1.3%) |
| Cyprus | 9 (0.1%) | – | – |
| Czechia | 95 (0.6%) | 4 (0.1%) | – |
| Denmark | 51 (0.3%) | 31 (0.7%) | 1 (1.3%) |
| Dominican Republic | 1 (0.0%) | – | – |
| Ecuador | 2 (0.0%) | – | – |
| Egypt | 36 (0.2%) | 1 (0.0%) | 1 (1.3%) |
| Entity 1 | 3 (0.0%) | 132 (3.0%) | – |
| Eswatini | 1 (0.0%) | – | – |
| Finland | 44 (0.3%) | 2 (0.0%) | – |
| France | 1650 (9.6%) | 544 (12.2%) | 6 (7.5%) |
| Georgia | 1 (0.0%) | – | – |
| Germany | 2805 (16.3%) | 148 (3.3%) | 3 (3.8%) |
| Greece | 96 (0.6%) | 8 (0.2%) | 1 (1.3%) |
| Guatemala | 3 (0.0%) | – | – |
| Hong Kong | 9 (0.1%) | 2 (0.0%) | – |
| Hungary | 146 (0.8%) | 2 (0.0%) | – |
| Iceland | 2 (0.0%) | – | – |
| India | 66 (0.4%) | 13 (0.3%) | 2 (2.5%) |
| Indonesia | 9 (0.1%) | – | – |
| Iran | 33 (0.2%) | 5 (0.1%) | – |
| Iran (Islamic Republic Of) | 1 (0.0%) | – | – |
| Ireland | 59 (0.3%) | 24 (0.5%) | – |
| Israel | 54 (0.3%) | 28 (0.6%) | 3 (3.8%) |
| Italy | 864 (5.0%) | 65 (1.5%) | 9 (11.3%) |
| Japan | 506 (2.9%) | 105 (2.4%) | 3 (3.8%) |
| Jordan | 7 (0.0%) | 1 (0.0%) | – |
| Korea, Republic of | 1 (0.0%) | 2 (0.0%) | – |
| Korea, South | 86 (0.5%) | 45 (1.0%) | – |
| Kuwait | 2 (0.0%) | – | – |
| Latvia | 3 (0.0%) | 1 (0.0%) | – |
| Lebanon | 7 (0.0%) | 4 (0.1%) | – |
| Lithuania | 3 (0.0%) | 1 (0.0%) | – |
| Luxembourg | 1 (0.0%) | – | – |
| Malaysia | 16 (0.1%) | 1 (0.0%) | – |
| Malta | 1 (0.0%) | – | – |
| Mexico | 71 (0.4%) | 4 (0.1%) | – |
| Moldova | 1 (0.0%) | – | – |
| Monaco | 1 (0.0%) | – | – |
| Montenegro | 1 (0.0%) | – | – |
| Morocco | 2 (0.0%) | 8 (0.2%) | – |
| Mozambique | 2 (0.0%) | – | – |
| Netherlands | 233 (1.4%) | 140 (3.1%) | 4 (5.0%) |
| New Zealand | 14 (0.1%) | 2 (0.0%) | – |
| Nigeria | 1 (0.0%) | – | 1 (1.3%) |
| Norway | 17 (0.1%) | 14 (0.3%) | – |
| Oman | 9 (0.1%) | – | 3 (3.8%) |
| Pakistan | 8 (0.0%) | – | – |
| Panama | 4 (0.0%) | – | – |
| Peru | 2 (0.0%) | 2 (0.0%) | – |
| Philippines | 16 (0.1%) | 2 (0.0%) | – |
| Poland | 181 (1.1%) | 5 (0.1%) | 1 (1.3%) |
| Portugal | 175 (1.0%) | 40 (0.9%) | – |
| Qatar | 1 (0.0%) | – | 1 (1.3%) |
| Romania | 55 (0.3%) | 1 (0.0%) | – |
| Russia | 42 (0.2%) | 1 (0.0%) | – |
| Russian Federation | 1 (0.0%) | – | – |
| Saudi Arabia | 7 (0.0%) | – | 1 (1.3%) |
| Singapore | 9 (0.1%) | 1 (0.0%) | – |
| Slovakia | 42 (0.2%) | 1 (0.0%) | – |
| Slovenia | 51 (0.3%) | – | – |
| South Africa | 7 (0.0%) | 1 (0.0%) | – |
| Spain | 515 (3.0%) | 60 (1.3%) | 1 (1.3%) |
| Sweden | 102 (0.6%) | 35 (0.8%) | – |
| Switzerland | 103 (0.6%) | 5 (0.1%) | – |
| Syrian Arab Republic | 1 (0.0%) | – | – |
| Taiwan | 63 (0.4%) | – | – |
| Thailand | 40 (0.2%) | 3 (0.1%) | – |
| Tunisia | 12 (0.1%) | 2 (0.0%) | – |
| Turkey | 49 (0.3%) | 39 (0.9%) | 7 (8.8%) |
| Ukraine | 6 (0.0%) | – | – |
| Um | 1 (0.0%) | – | – |
| United Arab Emirates | 3 (0.0%) | – | – |
| United Kingdom | 1621 (9.4%) | 684 (15.3%) | 7 (8.8%) |
| United States | 3754 (21.8%) | 1630 (36.5%) | 17 (21.3%) |
| Venezuela | 1 (0.0%) | – | – |
| Vietnam | 14 (0.1%) | – | – |
| Yemen | 12 (0.1%) | – | – |
| Czech Republic | – | 5 (0.1%) | – |
| Jamaica | – | 1 (0.0%) | – |
| Nepal | – | 2 (0.0%) | – |
| Puerto Rico | – | 2 (0.0%) | – |
| Serbia and Montenegro | – | 1 (0.0%) | – |
| Swaziland | – | 1 (0.0%) | – |
| Taiwan, Province of China | – | 4 (0.1%) | 1 (1.3%) |
| Reporter_Country_Missing | 4 (0.0%) | 48 (1.1%) | 1 (1.3%) |
Distinct temporal patterns were observed in the reporting of adverse events for the three therapeutic agents used in PCOS management (Figure 1a). Letrozole-associated reports demonstrated a two-phase trend: following a period of low reporting frequency from 2004 to 2010, a substantial increase was observed, culminating in a peak of 2323 reports in 2021. In contrast, tamoxifen-related reports maintained consistently lower levels throughout the study period, with modest elevations observed during 2006–2007 and 2014–2021, reaching a maximum of 488 reports in 2021. Clomiphene exhibited the most limited reporting profile, with annual case numbers remaining consistently low (range: 1–23 reports).
Figure 1.
Basic characteristics, temporal trends and gender differences in the treatment of PCOS. (a) Trend chart of reporting time changes in ADEs related to PCOS treatment. The different colored lines represented different drug treatments, and the size of each node represented the reported data. Larger nodes indicate a greater number of reports. (b) Number of regional reports on ADEs related to PCOS treatment. Different colors represented different regions. (c) Gender proportion of ADEs reports related to the treatment of PCOS. The pie charts represent the three drugs, with light blue represented females, dark blue represented males, and pink represented unknown. This figure is presented at the width of column 2.
Geographical analysis revealed that the United States was the predominant source of ADEs reports for all three drugs: tamoxifen, letrozole, and clomiphene (Figure 1b). For both tamoxifen and letrozole, the United States contributed the majority of reports, followed by various other countries. A more detailed examination showed that for clomiphene specifically, the United States was also the primary reporting country, with Italy ranking second.
Demographic characteristics indicated a pronounced sex-based distribution (Figure 1c). Female patients accounted for the substantial majority of reports across all therapeutic agents, representing 90.8% of letrozole reports, 84.0% of tamoxifen reports, and 70.0% of clomiphene reports.
Analysis of PCOS ADEs at PT Level
Following data curation to exclude reports with errors, incomplete information, or non-relevant signals, 4078 ADEs signals at the Preferred Term (PT) level were identified for the three drugs in PCOS treatment. The top 30 most frequently reported PTs, along with their signal strengths (PRR, χ2, and ROR with 95% confidence interval), are summarized in Table 2. Analysis revealed distinct drug-specific profiles: for tamoxifen, the most common ADEs were tumor progression [246 cases, PRR=11.03, χ2=2241.45, ROR(95%Cl)=11.21 (9.88–12.72)], fatigue [244 cases, PRR=1.39, χ2=26.83, ROR(95%Cl)=1.4 (1.23–1.58)], and arthralgia [211 cases, PRR=2.26, χ2=149.37, ROR(95%Cl)=2.28 (1.99–2.61)] (Figure 2a); for clomiphene, the most common ADEs were vomiting [10 cases, PRR=2.91, χ2=12.59, ROR(95%Cl)=2.95 (1.58–5.52)], ovarian hyperstimulation syndrome [6 cases, PRR=348.78, χ2=2074.79, ROR(95%Cl)=353.51 (157.78–792.06)], and upper abdominal pain [5 cases, PRR=3.32, χ2=8.11, ROR(95%Cl)=3.34 (1.38–8.07)] (Figure 2b); and for letrozole, neutropenia [2245 cases, PRR=14.43, χ2=27,599.89, ROR(95%Cl)=14.87 (14.25–15.51)], tumor progression (1666 cases, PRR=14.79, χ2=21,055.35, ROR(95%Cl)=15.13 (14.4–15.89)], and fatigue [1591 cases, PRR=1.77, χ2=535.2, ROR(95%Cl)=1.78 (1.7–1.88)] were predominant (Figure 2c). Notably, fatigue emerged as a frequently reported event common to both tamoxifen and letrozole.
Table 2.
The Top 30 with the Largest Number of Adverse Drug Event (ADE) Signal Detections in the FAERS Database
| pt_name_Tamoxifen | Report Number Tamoxifen | PRR | χ2 | ROR(95%Cl) | pt_name_Clomifene | Report Number Clomifene | PRR | χ2 | ROR(95%Cl) | pt_name_Letrozole | Report Number Letrozole | PRR | χ2 | ROR(95%Cl) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Malignant neoplasm progression | 246 | 11.03 | 2241.45 | 11.21 (9.88–12.72) | Vomiting | 10 | 2.91 | 12.59 | 2.95 (1.58–5.52) | Neutropenia | 2245 | 14.43 | 27,599.89 | 14.87 (14.25–15.51) |
| Fatigue | 244 | 1.39 | 26.83 | 1.4 (1.23–1.58) | Ovarian hyperstimulation syndrome | 6 | 348.78 | 2074.79 | 353.51 (157.78–792.06) | Malignant neoplasm progression | 1666 | 14.79 | 21,055.35 | 15.13 (14.4–15.89) |
| Arthralgia | 211 | 2.26 | 149.37 | 2.28 (1.99–2.61) | Abdominal pain upper | 5 | 3.32 | 8.11 | 3.34 (1.38–8.07) | Fatigue | 1591 | 1.77 | 535.2 | 1.78 (1.7–1.88) |
| Hot flush | 200 | 12.28 | 2068.38 | 12.45 (10.82–14.32) | Abdominal pain | 5 | 2.91 | 6.28 | 2.93 (1.21–7.07) | Nausea | 1222 | 1.34 | 104.87 | 1.34 (1.27–1.42) |
| Disease progression | 136 | 5.12 | 451.33 | 5.16 (4.36–6.11) | Maternal exposure before pregnancy | 5 | 104.87 | 514 | 106.04 (43.9–256.13) | Arthralgia | 1198 | 2.51 | 1090.02 | 2.53 (2.39–2.68) |
| Pulmonary embolism | 123 | 5.54 | 458.26 | 5.59 (4.68–6.67) | Neonatal respiratory distress syndrome | 5 | 181.64 | 896.96 | 183.69 (76.03–443.8) | Diarrhoea | 942 | 1.27 | 55.25 | 1.28 (1.2–1.36) |
| Metastases to bone | 116 | 32.81 | 3548.54 | 33.08 (27.53–39.74) | Vision blurred | 4 | 3.99 | 8.98 | 4.02 (1.5–10.75) | Anaemia | 755 | 3.35 | 1242.59 | 3.37 (3.14–3.63) |
| Insomnia | 105 | 1.71 | 30.95 | 1.71 (1.41–2.08) | Abdominal distension | 4 | 5.26 | 13.82 | 5.3 (1.98–14.18) | Dyspnoea | 723 | 1.09 | 5.91 | 1.1 (1.02–1.18) |
| Myalgia | 101 | 2.6 | 99.82 | 2.61 (2.15–3.18) | Blood bilirubin increased | 4 | 19.41 | 69.87 | 19.58 (7.31–52.39) | Alopecia | 702 | 3.03 | 953.51 | 3.05 (2.83–3.29) |
| Depression | 100 | 1.88 | 41.02 | 1.88 (1.55–2.29) | Product use issue | 4 | 2.98 | 5.27 | 2.99 (1.12–8.01) | Breast cancer | 697 | 6.03 | 2906.89 | 6.08 (5.64–6.55) |
| Muscle spasms | 98 | 2.33 | 74.39 | 2.34 (1.92–2.85) | Visual acuity reduced | 3 | 11.43 | 28.56 | 11.5 (3.69–35.79) | Leukopenia | 695 | 12.08 | 6959.91 | 12.19 (11.31–13.14) |
| Weight increased | 96 | 1.92 | 42.51 | 1.93 (1.58–2.36) | Mania | 3 | 24.22 | 66.79 | 24.38 (7.83–75.88) | Metastases to bone | 685 | 39.34 | 24,358.64 | 39.71 (36.77–42.9) |
| Pain in extremity | 92 | 1.34 | 7.94 | 1.34 (1.09–1.65) | Alanine aminotransferase increased | 3 | 6.44 | 13.81 | 6.48 (2.08–20.17) | Breast cancer metastatic | 671 | 57.75 | 34,824.28 | 58.3 (53.88–63.08) |
| Alopecia | 92 | 2.03 | 48.26 | 2.04 (1.66–2.5) | Ascites | 3 | 13.59 | 34.99 | 13.67 (4.39–42.55) | Disease progression | 661 | 4.87 | 2026.01 | 4.91 (4.55–5.3) |
| Breast cancer | 91 | 4.01 | 206.09 | 4.03 (3.28–4.96) | Aspartate aminotransferase increased | 3 | 7.44 | 16.73 | 7.48 (2.4–23.28) | Asthenia | 588 | 1.34 | 49.95 | 1.34 (1.23–1.45) |
| Breast cancer recurrent | 80 | 117.61 | 8983.55 | 118.3 (94.65–147.85) | Exposure during pregnancy | 3 | 6.75 | 14.7 | 6.78 (2.18–21.12) | Thrombocytopenia | 553 | 4.28 | 1387.31 | 4.31 (3.96–4.69) |
| Deep vein thrombosis | 79 | 5.14 | 263.54 | 5.17 (4.14–6.45) | Foetal exposure during pregnancy | 3 | 5 | 9.61 | 5.03 (1.61–15.64) | Metastases to liver | 549 | 28.35 | 13,977.1 | 28.57 (26.23–31.12) |
| Bone pain | 76 | 5.68 | 293.17 | 5.71 (4.56–7.15) | Suicidal ideation | 3 | 4.37 | 7.81 | 4.39 (1.41–13.68) | Hot flush | 526 | 6.32 | 2341.1 | 6.36 (5.84–6.94) |
| Breast cancer metastatic | 76 | 31.49 | 2226.32 | 31.66 (25.25–39.7) | Mood altered | 3 | 14.92 | 38.96 | 15.01 (4.82–46.72) | Decreased appetite | 491 | 1.85 | 192.35 | 1.86 (1.7–2.03) |
| Metastases to liver | 74 | 19.01 | 1257.13 | 19.11 (15.2–24.03) | Abdominal pain lower | 3 | 15.81 | 41.62 | 15.91 (5.11–49.51) | White blood cell count decreased | 471 | 3.7 | 926.4 | 3.72 (3.4–4.07) |
| Back pain | 70 | 1.31 | 5.23 | 1.31 (1.04–1.66) | Multiple pregnancy | 3 | 2180.37 | 6419.81 | 2195.1 (698.18–6901.47) | Pyrexia | 468 | 1.15 | 8.76 | 1.15 (1.05–1.26) |
| Drug interaction | 65 | 1.79 | 22.52 | 1.79 (1.4–2.28) | Pancreatitis acute | 3 | 18.46 | 49.55 | 18.58 (5.97–57.83) | Bone pain | 466 | 6.84 | 2305.02 | 6.88 (6.28–7.54) |
| Thrombosis | 62 | 3.34 | 101.6 | 3.35 (2.61–4.3) | Retinal vein occlusion | 2 | 96.98 | 189.84 | 97.41 (24.27–390.94) | Neoplasm progression | 414 | 9.69 | 3186.53 | 9.74 (8.83–10.73) |
| Vaginal haemorrhage | 60 | 5.99 | 249.2 | 6.01 (4.66–7.75) | Grandiosity | 2 | 716.07 | 1419.78 | 719.28 (178.6–2896.82) | Pain in extremity | 410 | 1.16 | 9.52 | 1.17 (1.06–1.28) |
| Neoplasm progression | 57 | 6.77 | 280.1 | 6.8 (5.24–8.82) | Psychomotor hyperactivity | 2 | 15.67 | 27.47 | 15.73 (3.92–63.12) | Cough | 391 | 1.22 | 15.01 | 1.22 (1.1–1.34) |
| Maculopathy | 55 | 66.64 | 3497.68 | 66.91 (51.23–87.39) | Abortion spontaneous | 2 | 6.59 | 9.49 | 6.62 (1.65–26.55) | Myalgia | 386 | 1.94 | 176.07 | 1.95 (1.76–2.15) |
| Uterine polyp | 55 | 143.04 | 7487.63 | 143.61 (109.68–188.04) | Chromaturia | 2 | 11.79 | 19.75 | 11.83 (2.95–47.47) | General physical health deterioration | 385 | 3.02 | 518.03 | 3.03 (2.74–3.35) |
| Suicidal ideation | 54 | 2.57 | 51.9 | 2.58 (1.97–3.37) | Anger | 2 | 7.71 | 11.68 | 7.74 (1.93–31.03) | Constipation | 372 | 1.53 | 67.34 | 1.53 (1.38–1.69) |
| Neuropathy peripheral | 51 | 2.46 | 44.16 | 2.46 (1.87–3.24) | Persecutory delusion | 2 | 111.15 | 218.12 | 111.64 (27.82–448.08) | Back pain | 368 | 1.35 | 32.84 | 1.35 (1.22–1.49) |
| Metastases to lung | 51 | 20.11 | 921.51 | 20.18 (15.32–26.58) | Psychotic disorder | 2 | 9.05 | 14.33 | 9.09 (2.27–36.45) | Pleural effusion | 337 | 4.69 | 972.28 | 4.7 (4.23–5.24) |
Figure 2.
Analysis of PCOS ADEs at PT level. (a) The names and numbers of the top 30 ADEs with the highest number of tamoxifen signals detected in the FAERS database, along with their corresponding PTs. (b) The names and numbers of the top 30 ADEs with the highest number of clomiphene signals detected in the FAERS database, along with their corresponding PTs. (c) The names and numbers of the top 30 ADEs with the highest number of letrozole signals detected in the FAERS database, along with their corresponding PTs. The horizontal axis represented the number of ADEs reports, the vertical axis represented PT entries, and the legend corresponding to each entry is on the right side. This figure is presented at the width of column 2.
ADEs Signal Detection at the SOC Level and the Distribution Within the SOC
SOC analysis yielded 11 valid ADEs signals for letrozole in treating PCOS, 8 for tamoxifen, and 8 for clomiphene (Table 3). At the SOC level, the ADEs related to the treatment of PCOS with tamoxifen mostly involved organ systems such as neoplasms (benign, malignant, and unspecified tumors, including cysts and polyps) [1513 cases, PRR=4.16, χ2=3729.52, ROR(95%Cl)=4.55 (4.32–4.8)], musculoskeletal and connective tissue diseases [1080 cases, PRR=1.48, χ2=180.11, ROR(95%Cl)=1.53 (1.43–1.62)], and eye diseases [716 cases, PRR=2.59, χ2=710.58, ROR(95%Cl)=2.67 (2.48–2.88)] (Figure 3a and Supplementary Table 4). Specific ADEs within this SOC included malignant neoplasm progression (benign, malignant, and unspecified tumors, including cysts and polyps) (Figure 3b). The ADEs related to the treatment of PCOS with clomiphene mostly involved organ systems such as psychiatric disorders [63 cases, PRR=2.46, χ2=58.11, ROR(95%Cl)=2.7 (2.07–3.53)], gastrointestinal diseases [55 cases, PRR=1.43, χ2=7.75, ROR(95%Cl)=1.49 (1.12–1.97)], and investigations [45 cases, PRR=1.61, χ2=11.21, ROR(95%Cl)=1.68 (1.24–2.29)] (Figure 3c and Supplementary Table 5). Among them, ADEs like mania, suicidal ideation, and mood alteration were observed to be distributed within psychiatric disorders (Figure 3d). The ADEs related to the treatment of PCOS with letrozole mostly involved organ systems such as neoplasms (benign, malignant, and unspecified tumors, including cysts and polyps) [7502 cases, PRR=4.04, χ2=17,514.64, ROR(95%Cl)=4.4 (4.3–4.51)], investigations [6427 cases, PRR=1.47, χ2=1025.36, ROR(95%Cl)=1.52 (1.48–1.56)], and gastrointestinal diseases [6290 cases, PRR=1.04, χ2=11.12, ROR(95%Cl)=1.05 (1.02–1.07)] (Figure 3e and Supplementary Table 6). Among them, ADEs such as constipation, diarrhoea, and nausea were noted to be distributed within gastrointestinal diseases (Figure 3f).
Table 3.
The Distribution of Adverse Drug Event (ADE) Signals of the Three Drugs Related to the SOC Level
| soc_name_en | Report.number.Letrozole | Report.number.Tamoxifen | Report.number.Clomifene |
|---|---|---|---|
| Respiratory, Thoracic and Mediastinal Disorders | 3568 | – | – |
| Investigations | 6427 | – | 45 |
| Gastrointestinal Disorders | 6290 | – | 55 |
| Neoplasms Benign, Malignant and Unspecified (Incl Cysts and Polyps) | 7502 | 1513 | – |
| Vascular Disorders | 1681 | 616 | – |
| Reproductive System and Breast Disorders | 701 | 710 | 23 |
| Musculoskeletal and Connective Tissue Disorders | 5577 | 1080 | – |
| Blood and Lymphatic System Disorders | 5283 | – | – |
| Metabolism and Nutrition Disorders | 1700 | – | – |
| Congenital, Familial and Genetic Disorders | 247 | 106 | 8 |
| Hepatobiliary Disorders | 1474 | 234 | 10 |
| Eye Disorders | – | 716 | 31 |
| Ear and Labyrinth Disorders | – | 96 | – |
| Psychiatric Disorders | – | – | 63 |
| Pregnancy, Puerperium and Perinatal Conditions | – | – | 18 |
Figure 3.
Detection of ADEs signals at the SOC level and distribution of signals within the SOC, along with network diagrams of drug-adverse reaction event reports at the SOC level. (a) Tamoxifen ADEs report and ROR signal detection results (95% CI). The horizontal axis in the figure represented the number of ADEs reports, the vertical axis represented SOC entries, and the legend corresponding to each entry is on the right side. (b) Distribution network diagram of tamoxifen ADEs signal in each SOC. The root node represents the number of tamoxifen and ADEs signals, with SOC in the inner loop and ADEs signal names in the outer loop. (c) clomiphene ADEs report and ROR signal detection results (95% CI). (d) Distribution network diagram of clomiphene ADEs signal in each SOC. (e) Letrozole ADEs report and ROR signal detection results (95% CI). (f) Distribution network diagram of letrozole ADEs signal in each SOC. (g) Network diagram of ADEs signals for tamoxifen, clomiphene, and letrozole in the FAERS database. The root node represented the number of drugs and ADEs signals, with drugs in the inner loop and ADEs signal names in the outer loop. This figure is presented at the width of column 2.
In summary, analysis at the System Organ Class (SOC) level revealed that neoplasms (benign, malignant, and unspecified, including cysts and polyps) represented the most frequently reported category for ADEs associated with both tamoxifen and letrozole in the treatment of PCOS. Additionally, gastrointestinal disorders emerged as another commonly affected organ system, being prominently associated with both clomiphene and letrozole therapy in PCOS patients.
Signal Network Diagram of the Reports on Drugs and Adverse Reaction Events at the SOC Level
Analysis of the top 10 ADEs signals revealed three common signals shared between tamoxifen and letrozole: malignant neoplasm progression, fatigue, and arthralgia (Figure 3g). These findings indicate that these concomitant adverse events—particularly malignant tumor progression, fatigue, and arthralgia—merit particular attention in the clinical management of PCOS patients receiving these therapeutic agents.
Discussion
PCOS pathogenesis is primarily attributed to dysregulation of the hypothalamic-pituitary-ovarian axis (HPOA).17 As a systemic endocrine and metabolic disturbance, PCOS exerts long-term effects on female reproductive health and overall well-being throughout the lifespan. PCOS is additionally associated with increased risks of diabetes mellitus, cardio/cerebrovascular disorders, endometrial hyperplasia/carcinoma, breast cancer, sleep disturbances, and psychological comorbidities including depression and anxiety.18
Clomiphene and serve as first-line therapies for ovulation induction in PCOS, for addressing infertility associated with ovulatory dysfunction.19 Patients undergoing clomiphene treatment tend to develop a higher number of smaller follicles, which may reduce the likelihood of proper follicular maturation and rupture, while potentially increasing the risks of ovarian hyperstimulation and multiple pregnancies. Some studies have reported that patients treated with letrozole exhibit significantly fewer mature follicles per cycle compared to other regimens. However, other research found no statistically significant difference in the standardized mean difference (SMD) between letrozole and clomiphene treatment groups, although the letrozole group still demonstrated a lower number of mature follicles.20,21 In letrozole-treated patients with PCOS, endometrial thickness was significantly increased, and resistance index (RI) was significantly lower in the subendometrial arteries—a critical determinant of clinical pregnancy rates. Letrozole had been established as the optimal first-line agent for ovulation induction in women with PCOS. However, the management of its associated ADEs remains a clinical challenge, warranting further investigation. Using the FAERS database, we conducted what is, to our knowledge, the first systematic comparison of adverse event signals associated with three pharmacological treatments for PCOS, analyzing both baseline characteristics and signal strengths at the PT and SOC levels. This study provides a meaningful real-world contribution by highlighting differential adverse event reporting patterns across these therapies. It is important to emphasize that these results, while providing novel insights into the safety profiles of PCOS management, should be interpreted as preliminary signal detection rather than established causal evidence. ADEs reports for tamoxifen, clomiphene, and letrozole were exclusively derived from women aged 18–65 years with body weights ranging from 50 to 100 kg. Letrozole-and clomiphene-related ADEs were predominantly reported by healthcare professionals. Geographically, three medicines’ ADEs were primarily reported in the United States. It is noteworthy that the number of reports for clomiphene was substantially lower than for the other two drugs (n=80). As the FAERS database is a spontaneous reporting system reliant on voluntary submissions from the public with diverse ages, backgrounds, and professions, it inherently suffers from limitations such as under-reporting and geographic clustering of reports, which can introduce reporting bias. Furthermore, the annual number of adverse event reports for clomiphene has exhibited a declining trend in recent years, a phenomenon potentially associated with the increased clinical use of letrozole.22 Fatigue emerged as a common ADEs for both tamoxifen and letrozole in PCOS treatment. Tamoxifen and letrozole ADEs reports frequently involved benign, malignant, and unspecified neoplasms (including cysts and polyps). Clomiphene and letrozole ADEs were often associated with gastrointestinal disorders. Signal Detection: Strong disproportionality signals were identified for malignant neoplasm progression, fatigue, and arthralgia with tamoxifen and letrozole. This study provides a systematic and large-scale assessment of the safety profiles of these PCOS therapies, elucidating their differential impacts across multiple organ systems.
Our analysis of ADR reports for these three agents primarily focused on the predominant age and weight ranges, which may reflect the higher prevalence of PCOS in adult women. This demographic pattern likely contributes to the more frequent clinical use of tamoxifen, clomiphene citrate, and letrozole in female populations. Notably, racial and ethnic disparities exist in seeking infertility care. For instance, African American and Asian American women—particularly those of Chinese descent—express greater concern regarding the social stigma associated with infertility compared to their white counterparts. These findings highlight a potential “culture of silence”, in which some women may avoid seeking medical care for conditions such as infertility and PCOS.23 Our findings are consistent with previous observations that the majority of adverse events for tamoxifen and letrozole originated from the United States.
Common symptoms of PCOS include menstrual irregularities (81.8%), fatigue (71.6%), and hyperandrogenism (77.6%), as well as anxiety and obesity.24–26 Our study demonstrates that fatigue is a frequently reported shared adverse event (AE) in ADEs reports associated with tamoxifen and letrozole treatment for PCOS, with a notably high number of cases. Furthermore, two additional comparative studies evaluating the ovulation induction efficacy of letrozole versus clomiphene in women with PCOS also demonstrated a higher incidence of hot flashes in the clomiphene group, whereas a significantly increased frequency of dizziness and fatigue was observed in the letrozole group.27,28 However, these adverse events were well tolerated by the patients, and no severe adverse reactions were observed.
It was also found that malignant neoplasm progression and arthralgia were signals shared between tamoxifen and letrozole. Neoplasms (benign, malignant, and unspecified, including cysts and polyps) were the most frequently reported SOC for both tamoxifen and letrozole, while gastrointestinal disorders were another commonly reported SOC shared by clomiphene and letrozole. Previous studies have also indicated that tamoxifen is associated with various adverse effects, including an increased risk of endometrial cancer, venous thromboembolic events, bone loss, and arthralgia, all of which significantly impair quality of life.29 PCOS is characterized by anovulation, progesterone deficiency, hyperandrogenism, and insulin resistance. These factors disrupt endometrial homeostasis in PCOS patients, potentially leading to chronic low-grade endometrial inflammation, infertility, endometrial hyperplasia, polyps, and even endometrial carcinoma. Notably, women with PCOS face a 2- to 6-fold higher risk of endometrial cancer compared to non-PCOS individuals.30 Clinically significant is the recognition that PCOS constitutes an established risk factor for endometrial carcinoma, while tamoxifen’s potential to stimulate endometrial proliferation further amplifies this oncogenic risk. Consequently, tamoxifen is not considered a first-line agent for ovulation induction in PCOS and is primarily reserved for cases that have failed prior stimulation with clomiphene or letrozole. Beyond their use in PCOS ovulation induction, both letrozole and tamoxifen are also employed in the treatment of estrogen-sensitive breast cancer. This broader usage in oncology may also be linked to the signal of malignant neoplasm progression observed in our analysis. Estrogen plays a pivotal role in maintaining normal bone mass in the human body by balancing the dynamic process of bone resorption by osteoclasts and bone formation by osteoblasts, with estrogen being the primary steroid hormone involved in bone resorption.31 Under estrogen-lowering conditions, the estrogen deprivation that occurs in both gonadal and extragonadal organs is often abrupt. This reduces the beneficial effects of estrogen on skeletal health and shifts the bone remodeling balance towards increased osteoclast-mediated bone resorption, resulting in net bone loss and reduced bone mass. Most researchers have found that aromatase is expressed in the synovial cells and chondrocytes of articular cartilage. Consequently, one proposed explanation is that the estrogen deficiency induced by aromatase inhibitors promotes the production of pro-inflammatory cytokines, such as interleukin-6 (IL-6) and IL-1, in articular chondrocytes. This process compromises the chondroprotective effects of estrogen, ultimately leading to joint pain and swelling.32,33
The adverse effects of letrozole on the cardiovascular system have been thoroughly investigated. Comparative studies on the cardiotoxicity of aromatase inhibitors, including letrozole and tamoxifen, have demonstrated that patients receiving tamoxifen exhibit better cardiovascular health outcomes due to its cardioprotective properties.34 Moreover, increasing doses of letrozole can induce various congenital abnormalities and even lead to fetal death. In rabbits, letrozole at 0.002 mg/kg causes embryotoxicity, while a dose of 0.02 mg/kg results in fetotoxicity.35 However, these risks typically manifest only with prolonged administration or elevated doses. Moreover, in a rat model of PCOS induced by letrozole, the gut microbiota exhibited significant alterations compared to healthy controls, including reduced abundances of Lactobacillus, Ruminococcus, and Clostridium, alongside enriched Incertae sedis.36 PCOS patients display decreased gut microbial diversity, compositional dysbiosis, and impaired intestinal mucosal barrier function relative to healthy subjects.37 The potential link between letrozole, PCOS, and gut microbiota dysbiosis warrants further investigation, as it may open new avenues for adjunctive therapies, such as probiotic supplementation, to mitigate gastrointestinal adverse events. In summary, the adverse event patterns identified in this study align with our objective of characterizing the pharmacovigilance signals for PCOS therapies. These findings contribute a meaningful real-world context for understanding post-marketing safety considerations. However, the broader interpretation of these results must remain cautious. As these findings reflect reporting signals derived from spontaneous data rather than confirmed clinical risks, they should be viewed as preliminary indicators that warrant further clinical investigation.
Conclusions
The widespread use of tamoxifen, clomiphene, and letrozole in PCOS underscores the importance of identifying potential safety signals, particularly concerning potential oncogenic risks. In this study, we conducted a signal-detection analysis of adverse reaction signals associated with tamoxifen, clomiphene, and letrozole in the treatment of PCOS using the FAERS database. This analysis provides crucial real-world evidence on the distinct safety profiles of tamoxifen, clomiphene, and letrozole in PCOS, offering valuable insights for informing clinical decision-making. However, as these signals are derived from spontaneous reports, they represent statistical associations rather than established causal relationships. These findings may be subject to confounding factors, including variation in patient history, dosage regimens, and reporting thresholds. This study has limitations inherent to spontaneous reporting systems. These constraints include: (1) incomplete data entries, (2) absence of comprehensive demographic and health information, (3) reporting biases, and (4) potential confounding factors such as concomitant medication use. While bioinformatics analyses provide valuable preliminary findings, their clinical translation requires rigorous validation through prospective clinical trials or large-scale, controlled longitudinal studies to establish causality. Furthermore, we emphasize the necessity for continued investigation into the underlying mechanisms of these observed associations.
Acknowledgments
We would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research.
Funding Statement
The research reported in this project was generously supported by Tongji University under grant agreement number [2024DR021]. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Abbreviations
PCOS, Polycystic ovary syndrome; ADEs, adverse drug events; FAERS, Food and Drug Administration Adverse Event Reporting System; PRR, proportional reporting ratio; ROR, reporting odds ratio; BCPNN, Bayesian confidence propagation neural network; EBGM, empirical Bayesian geometric mean; SERM, selective estrogen receptor modulator; FDA, Food and Drug Administration; PS, primary suspect; PT, preferred terms; SOC, system organ classes; MedDRA, Medical Dictionary for Regulatory Activities; ASCII, American Standard Code for Information Interchange; MD, Medical Doctors; CN, Consumers; OT, Other health professionals; HPOA, hypothalamic, pituitary, ovarian axis; AE, adverse event.
Data Statement
The datasets analysed during the current study are available in the FAERS database (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html).
Ethics Approval and Consent to Participate
This study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Shanghai Tenth People’s Hospital (approval No. SHSY-IEC-6.0/26K504/P01). This study was based exclusively on publicly available, de-identified data from the US Food and Drug Administration Adverse Event Reporting System (FAERS) and involved no direct contact with human participants or access to identifiable personal information. Individual informed consent was therefore not applicable to this study.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors declared that there were no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
- 1.Favaloro EJ, Curnow J, Pasalic L, et al. Laboratory assessment of Factor VIII inhibitors: when is it required? A Perspective informed by local practice. J Clin Med. 2024;14(1):13. doi: 10.3390/jcm14010013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Osibogun O, Ogunmoroti O, Michos ED. Polycystic ovary syndrome and cardiometabolic risk: opportunities for cardiovascular disease prevention. Trends Cardiovasc Med. 2020;30:399–19. doi: 10.1016/j.tcm.2019.08.010 [DOI] [PubMed] [Google Scholar]
- 3.Sarkar S, Gupta N, Joseph T, Yadav B, Kunjummen AT, Kamath MS. Comparison of letrozole versus combination letrozole and clomiphene citrate (CC) for ovulation induction in sub fertile women with polycystic ovarian syndrome (PCOS)-an open label randomized control trial. Reprod Sci. 2024;31:3834–3842. doi: 10.1007/s43032-024-01743-0 [DOI] [PubMed] [Google Scholar]
- 4.Ignatov A, Ortmann O. Endocrine risk factors of endometrial cancer: polycystic ovary syndrome, oral contraceptives, infertility, tamoxifen. Cancers. 2020;12. doi: 10.3390/cancers12071766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Abuzeyad FH, Ibnaouf ES, Farras MA. Clomiphene associated inferior STEMI in a Young female due to right coronary artery dissection. Case Rep Emerg Med. 2017;2017:4747831. doi: 10.1155/2017/4747831 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Avşar Ş, Öz A, Köken Avşar A, Kaya A, Börklü EB. Acute myocardial infarction associated with clomiphene citrate in a young woman. Turk Kardiyol Dern Ars. 2018;46:401–405. doi: 10.5543/tkda.2017.55237 [DOI] [PubMed] [Google Scholar]
- 7.Duran JR 3rd, Raja ML. Myocardial infarction in pregnancy associated with clomiphene citrate for ovulation induction: a case report. J Reprod Med. 2007;52:1059–1062. [PubMed] [Google Scholar]
- 8.Hardegree EL, Tweet MS, Hayes SN, Gulati R, Kane GC. Multivessel spontaneous coronary artery dissection associated with hormonal infertility therapy in a 39-year-old female. J Cardiol Cases. 2012;5:e69–e72. doi: 10.1016/j.jccase.2012.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ke C, Chen M, Lin L, Huang Y. Postmarketing adverse events of tamoxifen in male and female patients with breast cancer. Int J Cancer. 2025;156:734–743. doi: 10.1002/ijc.35193 [DOI] [PubMed] [Google Scholar]
- 10.Chen X, Jiang Y, Zhu H, Tian M. Mining and evaluation of adverse event signals for capmatinib based on the FAERS database. Front Pharmacol. 2024;15:1417661. doi: 10.3389/fphar.2024.1417661 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Li J, Xie J, Han Y, Zhang W, Wang Y, Jiang Z. A real-world disproportionality analysis of tirzepatide-related adverse events based on the FDA adverse event reporting system (FAERS) database. Endocr J. 2025;72:273–283. doi: 10.1507/endocrj.EJ24-0286 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wu J, Wu J, Tang B, et al. Suspected adverse drug reactions of rivaroxaban reported in the United States food and drug administration adverse event reporting system database: a pharmacovigilance study. Front Pharmacol. 2024;15:1399172. doi: 10.3389/fphar.2024.1399172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mangiola S, Doyle MA, Papenfuss AT. Interfacing Seurat with the R tidy universe. Bioinformatics. 2021;37:4100–4107. doi: 10.1093/bioinformatics/btab404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics. 2022;38:3844–3846. doi: 10.1093/bioinformatics/btac409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gentleman R, Whalen E, Huber W, Falcon S. Graph: graph: a package to handle graph data structures. 2023. doi: 10.18129/B9.bioc.graph [DOI]
- 16.Wickham HBJ. readxl: Read Excel Files. R package version 1.4.3. 2023.
- 17.Wang R, Li W, Bordewijk EM, et al. First-line ovulation induction for polycystic ovary syndrome: an individual participant data meta-analysis. Hum Reprod Update. 2019;25:717–732. doi: 10.1093/humupd/dmz029 [DOI] [PubMed] [Google Scholar]
- 18.Dandekar MP, Tadas M, Satthi S, et al. Oxyberberine revokes letrozole-induced polycystic ovarian syndrome and depression-like behavior in female Sprague-Dawley rats. Eur J Pharmacol. 2025;997:177613. doi: 10.1016/j.ejphar.2025.177613 [DOI] [PubMed] [Google Scholar]
- 19.Salaheldin AbdelHamid AM, Rateb AM, Ismail Madkour WA. Is clomiphene citrate stair-step protocol a good alternative to gonadotrophins in clomiphene-resistant PCO patients? Prospective Study J Obstet Gynaecol Res. 2016;42:547–553. doi: 10.1111/jog.12940 [DOI] [PubMed] [Google Scholar]
- 20.Akinoso-Imran AQ, Adetunji H. Systematic review and meta-analysis of letrozole and clomiphene citrate in polycystic ovary syndrome. Middle East Fertil Soc J. 2018;23:163–170. doi: 10.1016/j.mefs.2018.03.008 [DOI] [Google Scholar]
- 21.He D, Jiang F. Meta-analysis of letrozole versus clomiphene citrate in polycystic ovary syndrome. Reprod Biomed Online. 2011;23:91–96. doi: 10.1016/j.rbmo.2011.03.024 [DOI] [PubMed] [Google Scholar]
- 22.Liu Z, Geng Y, Huang Y, et al. Letrozole compared with clomiphene citrate for polycystic ovarian syndrome: a systematic review and meta-analysis. Obstet Gynecol. 2023;141:523–534. doi: 10.1097/aog.0000000000005070 [DOI] [PubMed] [Google Scholar]
- 23.VanHise K, Wang ET, Norris K, Azziz R, Pisarska MD, Chan JL. Racial and ethnic disparities in polycystic ovary syndrome. Fertil Steril. 2023;119:348–354. doi: 10.1016/j.fertnstert.2023.01.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Safiri S, Noori M, Nejadghaderi SA, et al. Prevalence, incidence and years lived with disability due to polycystic ovary syndrome in 204 countries and territories, 1990-2019. Hum Reprod. 2022;37:1919–1931. doi: 10.1093/humrep/deac091 [DOI] [PubMed] [Google Scholar]
- 25.Awwad O, Saleh M, Bardaweel S, et al. Use of complementary and alternative medicine among females with polycystic ovary syndrome in Jordan: a cross-sectional study. BMJ Open. 2024;14:e088945. doi: 10.1136/bmjopen-2024-088945 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Dumesic DA, Oberfield SE, Stener-Victorin E, Marshall JC, Laven JS, Legro RS. Scientific statement on the diagnostic criteria, epidemiology, pathophysiology, and molecular genetics of polycystic ovary syndrome. Endocr Rev. 2015;36:487–525. doi: 10.1210/er.2015-1018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wasim T, Nasrin T, Zunair J, Irshad S. Efficacy of Letrozole vs Clomiphene Citrate for induction of ovulation in women with polycystic ovarian syndrome. Pak J Med Sci. 2024;40:78–83. doi: 10.12669/pjms.40.1.7971 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Legro RS, Brzyski RG, Diamond MP, et al. Letrozole versus clomiphene for infertility in the polycystic ovary syndrome. N Engl J Med. 2014;371:119–129. doi: 10.1056/NEJMoa1313517 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhang Z, Park JW, Ahn IS, et al. Estrogen receptor alpha in the brain mediates tamoxifen-induced changes in physiology in mice. Elife. 2021;10. doi: 10.7554/eLife.63333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lu L, Luo J, Deng J, Huang C, Li C. Polycystic ovary syndrome is associated with a higher risk of premalignant and malignant endometrial polyps in premenopausal women: a retrospective study in a tertiary teaching hospital. BMC Womens Health. 2023;23:127. doi: 10.1186/s12905-023-02269-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Brufsky AM. Managing bone loss in women with early-stage breast cancer receiving aromatase inhibitors. Clin Breast Cancer. 2007;8 Suppl 1:S22–34. doi: 10.3816/cbc.2007.s.009 [DOI] [PubMed] [Google Scholar]
- 32.Le Bail J, Liagre B, Vergne P, Bertin P, Beneytout J, Habrioux G. Aromatase in synovial cells from postmenopausal women. Steroids. 2001;66:749–757. doi: 10.1016/s0039-128x(01)00104-0 [DOI] [PubMed] [Google Scholar]
- 33.Schicht M, Ernst J, Nielitz A, et al. Articular cartilage chondrocytes express aromatase and use enzymes involved in estrogen metabolism. Arthritis Res Ther. 2014;16:R93. doi: 10.1186/ar4539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ewer MS, Glück S. A woman’s heart: the impact of adjuvant endocrine therapy on cardiovascular health. Cancer. 2009;115:1813–1826. doi: 10.1002/cncr.24219 [DOI] [PubMed] [Google Scholar]
- 35.Tiboni GM, Ponzano A. Fetal safety profile of aromatase inhibitors: animal data. Reprod Toxicol. 2016;66:84–92. doi: 10.1016/j.reprotox.2016.09.016 [DOI] [PubMed] [Google Scholar]
- 36.Zhao H, Chen R, Zheng D, et al. Modified banxia xiexin decoction ameliorates polycystic ovarian syndrome with insulin resistance by regulating intestinal microbiota. Front Cell Infect Microbiol. 2022;12:854796. doi: 10.3389/fcimb.2022.854796 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Sun Y, Gao S, Ye C, Zhao W. Gut microbiota dysbiosis in polycystic ovary syndrome: mechanisms of progression and clinical applications. Front Cell Infect Microbiol. 2023;13:1142041. doi: 10.3389/fcimb.2023.1142041 [DOI] [PMC free article] [PubMed] [Google Scholar]