Phenoxymethylpenicillin potassium

Developmental toxicity / teratogenicity

Administrative data

Endpoint:

developmental toxicity

Data waiving:

study scientifically not necessary / other information available

Justification for data waiving:

other:

Justification for type of information:

Penicillin V is a well known active pharmaceutical ingredient belonging to the β-lactam antibiotics and used since many decades in patients for the treatment of mild to moderately severe bacterial infections through the inhibition of the cell-wall synthesis. Since similar metabolic pathways do not exist in mammals, penicillins and related β-lactam antibiotics have a low toxicity profile for patients when used in therapeutic doses. The most often described effects in humans are gastrointestinal, demonstrated in mouse and rat. Penicillin V is generally well-tolerated in humans but may occasionally cause transient nausea, diarrhoea and allergic reactions. It is well-known that changes in the intestinal flora occur in all individuals treated orally with penicillin. The degree of alteration is related directly to the quantity administered. This effect is usually of no clinical significance and the normal microflora is re-established shortly after therapy is stopped.

The hazard characterisation of penicillin and derivatives in humans can be deduced from the available long-term studies in the literature. A review of the safety profile of long-term penicillin therapy revealed some cases of adverse effects such as transient nausea, diarrhea, and allergic reactions - albeit these adverse effects are consistent with the known adverse event profile of penicillin. There was no report of an increase in adverse drug reactions involving long-term administration of penicillin in these studies. Therefore, given that Penicillin has an extensive history of safe use, and the lack of any serious adverse events in these studies supports its status as a safe drug. Additionally, the European Medicines Agency (EMA) Committee for veterinary medicinal products (CVMP) has assessed Phenoxymethylpenicillin, and the CVMP concluded that phenoxymethylpenicillin is rapidly metabolised and excreted, and has minimal direct toxicity to humans or animals. They noted that the therapeutic index of penicillin is more than 100 and that any toxic effects of non-allergic nature have only been reported after extremely high doses.

The data from some large published clinical trials assessing the effectiveness of long-term administration of penicillin and derivatives are outlined below. There appeared to be no reports of uncharacteristic adverse events including serious adverse events or adverse events requiring treatment in these studies.

Basil et al (2000) was a cross-sectional study conducted to establish the profile of secondary prophylaxis among children (n=150) with rheumatic heart disease in Egypt. Prophylactic failure occurred in one-third of the patients, raising doubts about the efficacy of the brands of penicillin prescribed. Recurrence of rheumatic fever was recorded in 37.3% of the patients with semiurban or rural residence, with non-compliance with secondary prophylaxis the significant risk factors (Basil et al., 2000).

Lue et al (1994) investigated the long-term outcome of patients with rheumatic fever receiving benzathine penicillin G. The study compared the efficacy of injections of 1.2 million units of benzathine penicillin G given every 3 weeks versus every 4 weeks for secondary prevention of rheumatic fever, based on the long-term outcome of patients receiving such prophylaxis. 249 patients with rheumatic fever, randomly assigned to either a 3-week or a 4-week regimen, were examined every 3 to 6 months, and followed for 794 and 775 patient-years, respectively. There were no reports of adverse effects (Lue et al., 1994). 

In Massell et al (1979) study, orally administered clindamycin and penicillin were compared for effectiveness in preventing streptococcal infections in 202 randomly assigned patients with previous rheumatic fever (RF). Among 143 patients aged 21 years or younger observed for 537 patient-years, the number of streptococcal infections (and number per patient-year) was 23 (0.084) in the penicillin group, and 12 (0.045) in the clindamycin group (Massell et al., 1979).

Angelin et al (1984) investigated the effect of penicillin on nasopharyngeal colonization with Streptococcus pneumonia in children with sickle cell disease. They studied nasopharyngeal colonization with pneumococci in 34 patients with sickle cell anemia (aged 6 months to 5 years) receiving penicillin prophylaxis and in 63 age- and race-matched comparison patients. Their data suggested a mechanism of action for penicillin prophylaxis and provided evidence for its relative safety (Angelin et al., 1984).

Buchanan et al (1986) reviewed seven-year experience with a regimen of twice-daily oral penicillin V potassium prophylaxis in 88 affected children. The median age at the start of prophylaxis was 10 months, and the median duration of prophylaxis was 29 months (range, three months to seven years). The total period of observation of patients who were prescribed penicillin was 248 person-years. Most patients also received one or two doses of polyvalent pneumococcal vaccine (Buchanan et al., 1986).

Gatson et al (1986) conducted a multicenter, randomized, double-blind, placebo-controlled clinical trial to test whether the daily administration of oral penicillin would reduce the incidence of septicemia due to S. pneumonia in children with sickle cell anemia who were under the age of three years. The children were randomly assigned to receive either 125 mg of penicillin V potassium (105 children) or placebo (110 children) twice daily. The trial was terminated 8 months early, after an average of 15 months of follow-up. 84% reduction in the incidence of infection was observed in the group treated with penicillin, as compared with the group given placebo (13 of 110 patients vs. 2 of 105; P = 0.0025), with no deaths from pneumococcal septicemia occurring in the penicillin group but three deaths from the infection occurring in the placebo group. (Gatson et al., 1986).

John et al (1984) investigated the efficacy of prophylactic penicillin in preventing pneumococcal infection in homozygous sickle cell (SS) disease in 242 children aged 6 months to 3 years at entry. Penicillin was well tolerated, and no confirmed allergic reactions occurred in the study (John et al., 1984).

Cross-reference

Cross-reference 1

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

health surveillance data

Type of information:

other: Review

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Study type:

human medical data

Qualifier:

no guideline followed

Specific details on test material used for the study:

Beta-Lactams and Related Antibiotics - Penicillins

Type of population:

other: Pregnants

Details on study design:

During pregnancy, untreated sexually transmitted or urinary tract infections are associated with significant morbidity, including low birth weight, preterm birth, and spontaneous abortion. Approximately one in four women will be prescribed an antibiotic during pregnancy, accounting for nearly 80% of prescription medications in pregnant women. Antibiotic exposures during pregnancy have been associated with both short-term (e.g., congenital abnormalities) and long-term effects (e.g., changes in gut microbiome, asthma, atopic dermatitis) in the newborn. However, it is estimated that only 10% of medications have sufficient data related to safe and effective use in pregnancy. Antibiotics such as beta-lactams, vancomycin, nitrofurantoin, metronidazole, clindamycin, and fosfomycin are generally
considered safe and effective in pregnancy. Physiologic changes in pregnancy lead to an increase in glomerular filtration rate, increase in total body volume, and enhanced cardiac output. These changes may lead to pharmacokinetic alterations in antibiotics that require dose adjustment or careful monitoring and assessment.

Results:

Penicillins and their newer derivatives are the most widely prescribed antimicrobial class during pregnancy. Intravenous penicillin from the time of rupture of the placental membranes until delivery remains first-line prophylaxis if the patient is colonized with Group B Streptococcus, while ampicillin is recommended as a suitable alternative. Penicillins generally cross the placenta in high concentrations. Penicillins
with increased protein binding such as the anti-staphylococcal penicillins (except methicillin) produce lower fetal tissue concentrations compared with penicillins such as penicillin G or ampicillin that have low protein binding. Due to increased plasma volume and creatinine clearance in pregnant women, serum penicillin concentrations may be decreased by as much as 50%, which may require increased doses and/or
frequency.

Conclusions:

Penicillins have a long track record of safety, with the parent compound penicillin and the aminopenicillins (ampicillin and amoxicillin) having the most robust safety data. All penicillins and their derivatives, as well as penicillin combinations with beta-lactamase inhibitors such as clavulanate or sulbactam, have been assigned a Pregnancy Category B rating.

Executive summary:

The use of antibiotics in pregnancy requires careful assessment and a discussion of risk versus benefit to mother and fetus, both short and long term. In general, many antibiotics are considered safe in pregnancy, especially beta-lactams: Penicillins have a long track record of safety, with the parent compound penicillin and the aminopenicillins (ampicillin and amoxicillin) having the most robust safety data. All penicillins and their derivatives, as well as penicillin combinations with beta-lactamase inhibitors such as clavulanate or sulbactam, have been assigned a Pregnancy Category B rating. 

 

Cross-reference 2

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

epidemiological data

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

1998

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Study type:

case control study (retrospective)

Endpoint addressed:

developmental toxicity / teratogenicity

Qualifier:

no guideline followed

Type of population:

other: pregnant women

Ethical approval:

not applicable

Details on study design:

The large population-based dataset of the Hungarian Case–Control Surveillance of congenital Abnormalities (HCCSCA) makes it possible to check the teratogenic potential of antibiotics by comparison of 38.151 healthy infants with 23 677 cases with congenital abnormality born between 1980 and 1996. The aims of the international EU Biomed programme are
(1) to determine the occurrence and distribution of different antibiotic treatments during pregnancy,
(2) to evaluate their indications,
(3) to study the effects of antibiotics for intrauterine.

The first step of the procedure in the HCCSCA was the identification of cases referred to the Hungarian Congenital Abnormality Registry within the first quarter of postnatal life. The Hungarian Congenital Abnormality Registry is a national registry of malformed liveborn infants, malformed stillborn fetuses and selectively terminated malformed fetuses.
The second step was to acquire appropriate healthy controls for each case. Two newborn infants without congenital abnormalities were matched to every case according to gender, birth week, and district of parents’ residence from the national birth registry of the Central Statistical Office. The third step was to obtain exposure data. A reply-paid questionnaire, an explanatory letter and a list of drugs and diseases were mailed immediately after notification to the diseases were mailed immediately after notification to the requested information on drugs taken, maternal diseases, pregnancy complications during pregnancy according to gestational weeks. To standardize the answers, mothers were asked to read the enclosed lists of drugs and diseases before they replied. Furthermore, mothers were requested to send us their prenatal care logbook and every medical document concerning the pregnancy or child studied. In the group of cases, regional district nurses were asked to visit and question nonrespondent families. Thus, information was available for 81% (69% from reply, 12% from visit) of the cases. The response rate for controls was 65%. District nurses did not participate in the evaluation of nonrespondent healthy controls for ethical reason. The fourth step was to evaluate drug intake in seven different respects. (1) The source of information was divided into three groups: (a) only data from the prenatal care logbook (prenatal care obstetricians are obliged to record all prescribed drugs for women concerning pregnancy related complications and diseases in the logbook) or other medical documents, (b) only data for questionnaires (for drugs used for treatment of diseases unrelated to pregnancy are prescribed by general practitioners or other physicians, in addition drugs taken by personal choice of pregnant women) and (c) data concordant from both medical documents and questionnaire. (2) The use of the antibiotic studied alone and with other drugs were differentiated. For multiple use of different drugs, each drug was coded independently according to the internationally accepted ATC classification. (3) The route of administration, i.e., oral, parenteral and topical treatments were separated. (4) Gestational time was calculated from the first day of the last menstrual period and three time periods were considered. (a) The earliest was the first month of pregnancy, often as a continuation of preconception treatment. (b) The second and third months of gestation involve the most sensitive, critical period for major congenital abnormalities. (c) The third category was the fourth–ninth months of gestation. However, there were some pregnant women with unknown gestational time of drug use. Sometimes the same antibiotic was used two or more times in different periods of the study pregnancy, it explains that the number of pregnant women and treatments may be different. (5) The dose of drugs, the recommended dose of antibiotics was used in most pregnant women. (6) Potential confounding factors, such as maternal age, birth order, pregnancy complications, acute and chronic maternal fetuses. (7) Birth weight and gestation age were evaluated mainly in the group of controls for each case. Two newborn infants without controls: the data of infants born to mothers with or without antibiotic treatment were compared. The fifth step was statistical analysis of the data. First, the time trend of different antibiotic uses was analysed. Second, among potential confounders, mean maternal age and birth order, were evaluated using Student’s t-test, while the most frequent maternal infectious diseases as the main indications of treatments were compared in untreated and treated groups by X2 test. Finally mean birth weight and gestation age of newborn infants with or without the use of antibiotic studied were compared using Student t-tests.

Results:

The mean birth weight was higher in the treated than in the untreated group. This can be explained mainly by the higher mean birth weight after the maternal treatment of other macrolides, other cephalosporins, other aminopenicillins, oxytetracycline, penamecillin, erythromycin, doxycycline and ampicillin. However, the mean birth weight was lower in the group of aminoglycosides and cefalexin. The mean gestation age was similar in the untreated and treated subgroups. However, it was higher in the group of doxycycline and other cephalosporins while lower in the group of phenoxymethylpenicillin, other penicillins and chloramphenicol.

The mean gestation age was similar in the untreated and treated subgroups. However, it was higher in the group of doxycycline and other cephalosporins while lower in the group of phenoxymethylpenicillin, other penicillins and chloramphenicol. Many pregnant women were treated in parallel or successively by two or more antibiotics and their interaction may be important in the evaluation of their teratogenic potential.

Conclusions:

The evaluation of antibiotic use during pregnancy allows some conclusions to be drawn.

(1) Different antibiotics, including penicillins have different chemical structures and different indications for treatment, therefore it is not appropriate to study their teratogenic potential in general, i.e., in combined group, e.g., penicillins.

(2) The underlying maternal disorders which make the antibiotic treatments necessary (i.e., the specific indications) and other drugs may have an important interaction with drug studied, thus it is necessary to use appropriate controls and adjusted odds–risk ratios.

(3) The frequency and distribution of antibiotic treatments show significant differences in European and North-American countries. Antibiotics use and indications have different priorities in different countries and it is necessary to consider this in the evaluation of their teratogenic potential in different interactions of macrolides. Obviously, this may also apply to their use [10] Reese RE, Betts RF, Douglas RG. The tetracyclines. A practical during pregnancy, it would be useful to organize interna- approach to infectious diseases. Boston: Little Brown, 3rd ed. tional comparative studies.

(4) Recently, intrauterine in fections have been increasingly important in the origin of fetal and neonatal diseases, therefore the possible prophylactic antibiotic treatments may reduce these.

(5) Antibiotics are used frequently in Hungary, but so far their teratogenic hazards have not been detected.

Antibiotics provide a case in point where recommendations to avoid all drugs during pregnancy including the second–third months of gestation is unrealistic and may be dangerous.
About 8% of pregnant women need permanent drug treatment because of chronic disorders such as asthma, epilepsy, diabetes mellitus, hypertension, reduction of infant morbidity after preterm premature rupture of the etc.
In addition many more pregnant women have acute infectious diseases of respiratory system (e.g., flu), urinary tract (e.g., pyelocystitis) and genital organs (STDs) and these may cause serious hazards for the fetus due to high fever, endotoxins of Gram negative bacteria and amnionitis–intrauterine infections.
Thus, the anxiety and fear created by the notion that nearly all drugs cause congenital abnormalities may be more harmful than some proven human teratogenic drugs themselves. A better risk–benefit estimation of drug use, including antibiotics, during pregnancy is therefore an urgent and important task.

Executive summary:

Of 38 151 control pregnant women who delivered later newborn infants without congenital abnormality, 6554 (17.2%) were treated by antibiotics. Most women (14.5%) had penicillin, while 1.2% and 0.7% of pregnant women were treated by cephalosporins and tetracyclines, respectively. More than 100 pregnant women used the following antibiotics: ampicillin (6.9%), penamecillin (5.9%), cefalexin (1.0%), phenoxymethylpenicillin (0.6%), oxytetracycline (0.5%), erythromycin (0.45%), benzylpenicillin-procain (0.4%) and benzylpenicillin1benzylpenicillin-procain (0.3%). Different antibiotics had different indications for treatment. The mean birth weight was significantly lower in the treated group compared to the untreated group. Practical implications: Different antibiotics have different chemical structures and indications for treatment. Therefore it is not appropriate to evaluate their teratogenic potential of combined antibiotic groups. There may be many interactions between underlying maternal diseases, other drug uses, further confounding factors and
antibiotics studied, thus adequate controls are needed to estimate the adjusted teratogenic odds–risk ratios. European countries have different spectrum of antibiotic use. It would be necessary to know these baseline data of different populations. The anxiety and fear created by the notion that nearly all drugs cause congenital abnormalities may be more harmful than some proven human teratogenic
drugs themselves. Thus a better risk–benefit estimation for antibiotic uses during pregnancy is an urgent and important task.

Cross-reference 3

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

epidemiological data

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

2001

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Study type:

case control study (retrospective)

Endpoint addressed:

developmental toxicity / teratogenicity

Qualifier:

no guideline required

Type of population:

other: Pregnant Women

Details on study design:

Phenoxymethylpenicillin (penicillin V) remains the drug of choice for a number of Gram-positive
bacterial infections and for prophylactic treatment of carriers of group B b-hemolytic streptococci
during pregnancy, but data on this penicillin derivative with respect to safety during pregnancy
are few and inconsistent. According to on a population-based cohort study we report the association between maternal use of penicillin V during pregnancy and congenital abnormalities, preterm birth and low birth weight.

Study population: the study has been conducted in North Jutland County, Denmark, which has approximately 490.000 inhabitants, based on all primiparous women who had a live single child or a stillbirth after the 28th gestational week (n¼35 950) for the period 1991–1998.
Exposure data: The North Jutland Pharmaco-Epidemiological Prescription Database The Danish National Health Service provides taxfinanced 50% reimbursement of the costs of penicillin V, which is purchased by prescription only.
The electronic accounting system, used primarily to secure reimbursement from the Health Service to the pharmacies, includes the type and amount of the prescribed drug according to the
Anatomical Therapeutical Classification Code (ATC), the defined daily dose, the date of issuing
the prescription, and the personal register number of the patient (CPR number) which is assigned to all citizens shortly after birth or at immigration, encoding sex and date of birth.
All data are transferred to the research Prescription Database, which retains key data on all prescribed, reimbursed drugs sold at pharmacies in the county since 1 January 1991. It is thus possible to construct a complete prescription history for each inhabitant in the county.

This registry contains information on all births in
Denmark since 1 January 1973, collected by midwives
and doctors attending the deliveries [13].
The main variables include maternal age, selfreported
smoking status, birth order, gestational
age, length and weight at birth, and CPR number
of both mother and child.

The County Hospital Discharge Registry
The County Hospital Discharge Registry was
established in 1977 and contains data on more
than 99% of the discharges from hospitals in the
county [14]. The main variables include the CPR
number of the patient, dates of admission and
discharge, surgical procedures, and up to 20 discharge
diagnoses, classified according to the 8th
revision of the Danish version of the International
Classification of Diseases (ICD) until the end of
1993, and according to the 10th revision thereafter.
The registry was used to obtain information
about congenital abnormalities based on the diagnoses
740–759 in ICD-8 and Q00-Q99 in ICD-10.
This registry has the advantage of accumulating
information about congenital abnormalities irrespective
of the actual age of the patient at the time
of diagnosing the condition. All children recorded
in the registry until December 1998 were considered.
We excluded the codes ‘dislocation of the
hip’ and ‘undescended testes’ (755.69 and 752.10–
752.19 in ICD 8, and Q53 and Q65 in ICD 10),
which we regard as invalid based on a previous
review of hospital records.
Linkage of data
All live births and stillborns after the 28th gestational
week were identified in the Birth Registry,
and, based on the reported gestational age at birth,
we calculated the time of conception. The pregnancies
were divided into period 1: the first trimester,
and period 2: the second and third
trimesters. The pregnancies in which the women
had purchased a prescription for penicillin V (ATC
code J01C EO2) were further categorized according
to use of other medications in the periods,
hereafter referred to as ‘penicillin V only’ and
‘penicillin V and other drugs’.
During the study period, prescriptions for children
were identified and reimbursed according to
the mother’s CPR numbers. To avoid the potential
misclassification resulting from antibiotics prescribed
for a child in the family, the analyses were
restricted to primiparous women.
Reference pregnancies
As a reference we selected all primiparous singleton
pregnancies (n¼9263) in which the mothers
had purchased no prescription of any kind of
reimbursed medicine during pregnancy.

Statistical methods:

It has been used logistic regression analyses to estimate the prevalence odds ratios (POR) of congenital abnormalities, preterm birth (gestational age <37 weeks), and low birth weight (<2500 g) associated with penicillin V exposure, adjusted for maternal age, birth order and smoking. Due to
the observation that 30% of congenital abnormalities were cardiovascular, we made a separate
analysis of the risk association in these cases.
Pregnancies exposed in period one were used to estimate the risk of congenital abnormalities, and
in periods one and two to estimate the risk of preterm birth and low birth weight (the latter analysis was restricted to full-term births). Regarding the risk of congenital abnormalities, we made separate analyses for ‘penicillin V only’ and ‘penicillin V and other drugs’.

Results:

1886 pregnant women who purchased at least one prescription for penicillin V
during pregnancy have been identified. Apart from a higher proportion of smokers among penicillin V and other drug users, we found no significant differences regarding major characteristics between
the cohorts.

Among 654 women exposed to penicillin V in the first trimester it has been found 4.6% (n¼30), compared with 3.6% in the reference cohort, giving an adjusted POR of 1.25 (95% CI: 0.84–1.86).

A review of the hospital records confirmed all congenital abnormalities identified from the registry.
In the subgroup of 131 pregnancies exposed to penicillin V only, the adjusted POR of congenital
abnormalities was 1.35 (95% CI: 0.59–3.08), and for specific cardiovascular abnormalities (n¼9)
has been found an adjusted POR of 1.74 (95% CI: 0.83–3.65).

The proportion of preterm births was 5.5% in the users of penicillin V and 6.4% in the reference
group, giving an adjusted POR of 0.83 (95% CI:0.66–1.04).

When restricted to full-term births, the POR of low-birth-weight babies was 1.02 (95% CI: 0.71–1.47).

Conclusions:

The study study did not indicate that intrauterine exposure to penicillin V is associated
with any major additional risk to the fetus of congenital abnormalities, preterm birth, or low
birth weight.

Executive summary:

Objective To examine the risk of congenital abnormalities, preterm birth and low birth weight after exposure to phenoxymethylpenicillin in utero.
Methods Apopulation-based follow-up study in the County of North Jutland, Denmark.
Birth outcome for 1886 women, who redeemed prescriptions for phenoxymethylpenicillin during pregnancy was compared with the outcome for 9263 women who did not redeem any prescription during pregnancy.
Results The prevalence of congenital abnormalities in 654 users of phenoxymethylpenicillin with or without other drugs during the first trimester was 4.6% compared with 3.6% in the reference group, giving a prevalence odds ratio of 1.25 (95% CI: 0.84–1.86). The prevalence odds ratio was 1.35 (95% CI: 0.59–3.08) in 131 women who were exposed to phenoxymethylpenicillin only. Nine cardiovascular abnormalities were found, giving an adjusted prevalence odds ratio of 1.74 (95% CI: 0.83–3.65). The prevalence odds ratios of preterm birth and low birth weight were 0.83 (95% CI: 0.66–1.04) and 1.02 (95% CI: 0.71–
1.47), respectively.
Conclusion We found no significantly increased risk of congenital abnormalities, including cardiovascular abnormalities, preterm birth, or low birth weight in women who purchased phenoxymethylpenicillin during pregnancy.

Cross-reference 4

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

epidemiological data

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

December 1998

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Study type:

case control study (retrospective)

Qualifier:

no guideline followed

Type of population:

other: pregnant women

Details on study design:

The large population-based dataset of the Hungarian Case-Control Surveillance of Congenital Abnormalities between 1980 and 1996 was used in this study. Cases with isolated and multiple congenital abnormalities (CAs) were ascertained from the Hungarian Congenital Abnormality Registry. As population controls, two newborn infants without CAs were matched with every case according to sex, week of birth, and district of parents‘ residence from the national birth registry of the Central Statistical Office.
The exposure data were obtained prospectively through the antenatal care logbook and other medical documents, and retrospectively by maternal questionnaire. In addition district nurses were asked to visit and to question nonrespondent families. Thus, information was available for 81% of cases and 65% of controls. The study was focused on the second-third months of gestation as the most sensitive critical period for most major CAs. Maternal age, birth order, occurrence of acute and chronic maternal disorders, and other drug uses were evaluated as confounders.

Statistical methods:

The statistical analysis of data was performed by the STATA statistical software package. First, the use of PMP was compared between the study groups and crude odds ratios (OR) with 95% confidence interval (95% CI) were calculated. Second, the
source of information and potential confounders were evaluated using X^2 and Student t test, or ORs were calculated. Third, the McNemar test was used for the analysis of case-control pairs using conditional logistic regression model. At least one of the controls was available in 95% of cases.

Results:

There were 2,146,574 births in Hungary during the study period, 38,151 controls therefore represented 1.8% of the Hungarian births and 218 (0.6%) of them were born to mothers exposed to phenoxymethylpenicillin (PMP). The case group consisted of
22,865 malformed infants or fetuses and 173 (0.8%) had mothers with PMP treatment during pregnancy. The crude OR with 95% CI for PMP treatment during pregnancy was 1.3 (1.1–1.6).
The proportion of phenoxymethylpenicillin (PMP) treatments based on medical documents mainly antenatal logbook was higher in the control group (76: 35%) than in the case group (43: 25%) (c21=4.56; p=0.03). In general, confounders were similar in the two study groups and the occurrence of acute maternal diseases indicated that PMP was used mainly for the treatment of infectious diseases of respiratory system during pregnancy. The mean maternal age was 24.8±4.6 years in the case and 25.6±4.6 years in the control group (t=1.78; p=0.08). The birth order was the same (1.6±0.8 vs 1.6±0.7) in the two study groups (t=0.19; p=0.85).
The distribution of PMP treatments was similar according to gestational months in the case and control groups, with a maximum in the third-fourth months. Of 173 cases and 218 controls, 51 (29.5%) and 69 (31.7%) had treatment in the second-third months of gestation, respectively.
The McNemar analysis for case-control pairs including 14 CA-groups consisting of two or more cases indicated a higher maternal PMP treatment during the entire pregnancy in cleft lip±palate and rectal-anal atresia/stenosis. However, these CAs had not a significantly higher exposure to PMP during the second-third months of pregnancy. Adjusted OR for PMP treatment was significantly higher in hypospadias during the second-third months of pregnancy, but this CA exceptionally has a critical period during the third-fourth months of pregnancy and there was no significant difference in the use of PMP in this time period (7 cases vs 2 controls, OR: 3.5, 0.9–14.0). In addition at the evaluation of only medically documented PMP treatment during the entire pregnancy or the second-third months of pregnancy we did not find significant difference in any CA-group of case-control pairs.

Conclusions:

It has been examined the exposure to maternal PMP treatment during pregnancy in the group of cases with CAs and their matched controls without CA. Our findings did not support a causal association between this type of penicillin V and specific CAs.
The main strengths of our study are the large size (including 391 pregnant women with phenoxymethylpenicillin, PMP, treatment) and the population-based design thus we can avoid the usual bias in hospital-based studies with differential patient recruitment.
The weaknesses include the different response rate and possible recall bias. The latter was indicated by the higher crude OR for PMP treatment in the case group. However, the lack of association between PMP treatment and specific CA based on the medical documentation can restrict recall bias and makes a causal relationship unlikely. In addition, underlying infections may increase the risk for CA due to the specific agent or due to the unspecific fever and other maternal consequences of infections. Antibiotic treatment may be a surrogate measures of infections and/or antibiotics may prevent CAs caused by infections.
This study has resulted in two conclusions.

On the one hand, case-control studies based only on retrospective maternal questionnaires are vulnerable because of recall bias which was indicated in this study by the higher proportion of selfreported maternal data in the case group.
However, the use of only medically documented data,the evaluation of different gestational time periods (i.e., mainly major organ-forming period) and the analysis of
different CAs (i.e., the noxa-specificity) can restrict recall bias.

On the other hand, the type I error, i.e., the use of a very large number of statistical tests can produce a significant difference in every 20th calculation. Thus, an isolated association may be due to chance as a result of multiple comparisons. However, the inconsistency of the higher exposure studied in different approaches can reveal the chance-effect.

In conclusion, the analysis of the overall pattern of the dataset, based on the comparison of cases and controls, showed that PMP treatment during pregnancy presented very little if any human teratogenic risk to the fetus.

Executive summary:

The objective of the study was to examine the human teratogenic potential of oral penicillin V: phenoxymethylpenicillin treatment during pregnancy in the large population-based dataset of the Hungarian Case-Control Surveillance of Congenital Abnormalities, 1980–1996. The dataset included 22,865 pregnant women who had fetuses or newborns with congenital abnormalities and 38,151 matched pregnant women who had newborn infants without any congenital abnormality (population control group). Of 22,865 case pregnant women, 173 (0.8%) had phenoxymethylpenicillin treatments, while of 38,151 population controls, 218 (0.6)
were treated by this penicillin V (crude OR 1.3 with 95% CI: 1.1–1.6). This difference was explained mainly by recall bias and confounders because adjusted OR for medically documented phenoxymethylpenicillin treatments did not show difference during the second-third months of gestation, i.e. in the critical period for most major congenital abnormalities in case-matched control pairs. Thus, treatment with oral phenoxymethylpenicillin during pregnancy presents very little if any teratogenic risk to the fetus.

Materials and methods

Test material

Test material information

1

Reference substance name:

Phenoxymethylpenicillin potassium

EC Number:

205-086-5

EC Name:

Phenoxymethylpenicillin potassium

Cas Number:

132-98-9

Molecular formula:

C16H17KN2O5S

IUPAC Name:

potassium (2S,5R,6R)-3,3-dimethyl-7-oxo-6-[(2-phenoxyacetyl)amino]-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylate

Test material form:

solid

Test animals

Species:

other: test already availabe on humans