Doxycycline use and adverse pregnancy or neonatal outcomes: A descriptive study using the United States Food and Drug Administration Adverse Event Reporting System database

Abstract Background and Aims Doxycycline is recommended for use in rickettsial diseases. The available evidence regarding its safety for rickettsial infection in pregnancy is limited. Our study aimed to describe the adverse events of doxycycline when used during pregnancy for any indication, in terms of adverse maternal and/or neonatal outcomes, using the United States Food and Drug Administration Adverse Event Reporting System (FAERS). Methods We used the OpenVigil software for extracting the safety reports from the United States submitted to the FAERS from 2004 to 2021. We manually reviewed reports of doxycycline use resulting in adverse pregnancy outcomes or congenital anomalies to describe the patient and safety event characteristics. Results From 2004 to 2021, 59 individual case safety reports containing preferred terms indicative of drug exposure during pregnancy or drug‐induced adverse fetal outcomes were identified in the FAERS database. Following deduplication and manual review, 20 relevant adverse event reports were obtained. Doxycycline was the suspect medication in 13/20 (65%) reports. The common adverse event terms reported were premature delivery/baby in 6 reports, spontaneous abortion in 6, intrauterine death in 2, and various congenital anomalies in the rest. Fifty percent of the safety reports contained other medications which could have potentially caused the outcome. Conclusions The number of reported events in the FAERS database of adverse pregnancy/neonatal outcomes following doxycycline use is small, similar to the numbers reported from large cohort or surveillance studies. Given the presence of concomitant medications that could have contributed to the outcome, there does not seem to be a strong signal of harm, although this needs to be confirmed by surveillance studies.


| BACKGROUND
Rickettsial infections during pregnancy can have an adverse impact on maternal and fetal outcomes. Rickettsiosis is an important differential diagnosis in cases of undifferentiated fever, especially in the South-Asian countries where it is one of the major causes of nonmalarial febrile illness. 1 Doxycycline, azithromycin, chloramphenicol, and rifamycins are useful in treating the disease to varying extents, with most studies supporting the use of doxycycline. 2 Doxycycline is generally contraindicated for use in pregnancy due to concerns of possible musculoskeletal and dental defects in the fetus and hepatotoxicity in the mother. 3,4 However, drugs other than doxycycline are not useful to treat all rickettsial diseases, such as anaplasmosis and ehrlichiosis, and the patient outcomes also may be comparatively inferior. 5 The Centers for Disease Control and Prevention recommends the use of doxycycline in pregnancy, although it does acknowledge that, based on the available evidence, a teratogenic risk cannot be completely ruled out. 2 The Indian Council of Medical Research guidelines on the diagnosis and treatment of rickettsial diseases recommends the use of azithromycin in pregnant women, 6 given the concerns regarding doxycycline use. The available literature evidence regarding doxycycline use for rickettsial infection in pregnancy and patient outcomes is limited. [7][8][9][10][11] Spontaneous reporting adverse event databases, such as the reporting on data regarding doxycycline use during pregnancy and the pregnancy and fetal outcomes using such spontaneous databases.
The objective of this study is to describe the adverse events of doxycycline when used during pregnancy for any indication, in terms of adverse maternal and/or neonatal outcomes, using the FAERS.

| METHODS
We performed a retrospective descriptive study of the individual case safety reports (ICSRs) in the United States FAERS database from the first quarter of 2004 to the third quarter of 2021.
OpenVigil version 2 (OpenVigil 2.1-MedDRA-v24) was used to extract the FAERS data. 12 OpenVigil uses a drug name mapping logic and any cases with errors in drug naming that cannot be resolved (incomplete cases) are excluded, thus providing a cleaned version of the FAERS data.
The ICSRs included in the study had to contain doxycycline reported as a suspect or concomitant medication and an adverse event term reporting at least one pregnancy and/or neonatal outcome and must have occurred in the United States. The adverse event terms in FAERS are coded using the Medical Dictionary for Regulatory Activities (MedDRA) terminology. However, the ICSRs do not have a data field to indicate whether an event occurred during pregnancy. Hence, to identify ICSRs reporting adverse pregnancy or neonatal outcomes due to drug use in pregnancy, we filtered the FAERS data using the standardized MedDRA query (SMQ) term "Pregnancy and neonatal topics" in OpenVigil; this includes the SMQ narrow terms congenital, familial and genetic disorders; fetal disorders; neonatal disorders; termination of pregnancy and risk of abortion; pregnancy, labor and delivery complications and risk factors (excl. abortions and stillbirth). 13,14 The ICSRs so obtained were further filtered using the following preferred terms: drug exposure during pregnancy; maternal drugs affecting fetus; maternal exposure during pregnancy; fetal exposure during pregnancy.
A major drawback of the FAERS database is the presence of duplicate case reports, occurring due to the reporting of the same event from different sources. We first identified multiple versions (follow-ups) of the same report by using the case ID, a unique number assigned to each report with the follow-up numbers added as a suffix, and then, retained only the latest version of the report. Then, the filtered list of ICSRs was reviewed manually to identify possible duplicates by comparing multiple data fields, such as the age, gender, date the report was received at FDA, drug name, event, and reporter country. 15 With respect to the pregnancy outcomes, only those reports indicative of premature labor or abortion were retained.
Reports containing adverse event terms unlikely to be drug-induced, such as "abortion induced" and "amniotic cavity infection" were excluded. This study was approved by the institutional ethics committee of Kasturba Medical College, Mangalore (IEC KMC MLR 04-2022/107).

| Statistical analysis
The ICSRs were downloaded from OpenVigil in Microsoft Excel file format (Version 2016) for deduplication and manual review. The list of ICSRs was analyzed to determine the various patient and adverse event characteristics such as age, gender, seriousness of the adverse event, time of onset of adverse event, and the reported event(s) in mother and fetus; the data are presented using descriptive statistics. Eleven ICSRs (47.83%) were reported in adult females; the rest were in newborns or where age was not reported. Table 2 shows the distribution of the adverse event terms as per the SMQ categories.

| RESULTS
The principle adverse event terms reported are listed in Table 3.
The adverse event outcomes reported were a congenital anomaly in 3 ICSRs (15%), death in 2 (10%), hospitalization in 2 (10%), and others in 13 ICSRs (65%). A brief description of each identified case is presented in Table 4. Five (25%) ICSRs were reported by consumers, and the rest by health professionals.
With regard to safety reports from other countries, 66 ICSRs were available (without manual review); 39 from Canada, United  show the presence of any congenital abnormalities in the children. 16 Studies using data from large registries, such as the Hungarian Case-Control Surveillance of Congenital Anomalies, which studied the data from 1980 to 1987, found that the use of teratogenic drugs during In contrast, data from the Quebec pregnancy cohort for the period 1998 to 2008 showed a twofold increased risk of circulatory system and cardiac malformations and a threefold increased risk of atrial or ventricular septal defect. 21 Despite the use of large databases in some of the above-mentioned studies, there were relatively few exposures to doxycycline in pregnant women, and only a small percentage of these had adverse outcomes. FAERS is one of the largest databases in terms of the number of adverse events reported every year; however, the number of reports identified in our study too was very low. It is also to be noted that the event statistic identified through these studies should be considered along with the background rates of birth defects of 3%−5%. 22 An alternative to doxycycline for rickettsiosis is azithromycin, which is generally considered safe and preferred for use in pregnancy. Interestingly, a multinational observational study of pregnancy outcomes among women who received macrolides during the first trimester of pregnancy compared with those who received known non-teratogenic drugs found a trend towards a higher rate of major congenital or cardiac malformations. 23  writingreview and editing. Both authors have read and approved the final version of the manuscript.

CONFLICT OF INTEREST
The authors declare no conflict of interest.

The lead authors Sammodavardhana Kaundinnyayana and Ashwin
Kamath affirm that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available in the