Objective
Cancer is diagnosed in approximately 1 per 1000 pregnant women. Lifesaving cancer therapy given to the mother during pregnancy appears in conflict with the interest of the developing fetus. Often, termination of pregnancy is suggested but has not been proven in any type of cancer to improve maternal prognosis, while very few studies have documented the long-term effects of in utero chemotherapy exposure on child outcome. To counsel patients about the risk of continuing a pregnancy while undergoing cancer treatment, we performed developmental testing to provide more detailed follow-up on children exposed in utero to chemotherapy.
Study Design
Mother-infant pairs, enrolled in the Cancer and Pregnancy Registry, were offered developmental testing for children who were ≥18 months of age. Based on age, the Bayley Scales of Infant Development–Third Edition, the Wechsler Preschool and Primary Scale of Intelligence-Revised, the Wechsler Intelligence Scale for Children, Third Edition, or the Wechsler Individual Achievement Test was administered. All parents or primary caregivers completed the Child Behavior Checklist, a parent questionnaire to assess behavior and emotional issues. Results of children exposed to chemotherapy before delivery were compared with children whose mothers were also diagnosed with cancer during pregnancy but did not receive chemotherapy before delivery.
Results
No significant differences were noted in cognitive skills, academic achievement, or behavioral competence between the chemotherapy-exposed group and the unexposed children. Of children, 95% scored within normal limits on cognitive assessments; 71% and 79% of children demonstrated at or above age equivalency in mathematics and reading scores, respectively; and 79% of children scored within normal limits on measures of behavior. Older children had significantly higher rates of internalizing behavior problems.
Conclusion
We could not demonstrate a significant difference in cognitive ability, school performance, or behavioral competence for children exposed to chemotherapy in utero compared with nonexposed controls. The majority of these children scored within normal limits on all developmental measures. Premature birth was more prevalent in the chemotherapy-exposed group yet did not predict developmental outcome. Older children in the sample demonstrated higher rates of internalizing behavior problems.
The diagnosis of cancer during pregnancy creates medical and moral dilemmas for physicians and patients. Pregnant women are hesitant to receive chemotherapy due to concerns about possible effects on the developing fetus, while physicians are hesitant to allow pregnant women to delay cancer treatment for the remainder of the pregnancy. Often, termination of the pregnancy is recommended. The Cancer and Pregnancy Registry, created in 1997 with approval of the Institutional Review Board at Cooper Medical School at Rowan University, follows up the pregnancies and long-term health of women diagnosed with cancer during pregnancy. Patients are enrolled at the time of cancer diagnosis, and the treatment during pregnancy is recorded. Several women were advised to terminate their pregnancies without information about receiving cancer treatment while pregnant. Absence of long-term follow-up data on children exposed to chemotherapy in utero influenced some patient and physician decisions regarding continuing a pregnancy complicated by cancer and/or cancer treatment during pregnancy.
A developmental psychologist performed standardized developmental testing for children born to mothers diagnosed with cancer during their pregnancies. The purpose of this study is to report developmental outcome for children exposed to chemotherapy in utero compared with a control group of unexposed children whose mothers were diagnosed with cancer while pregnant.
Approximately 1 in 1000 pregnancies is complicated by cancer. The number of births in 2010 was 3,999,386 (National Vital Statistics Report). In the past, providers have recommended termination of pregnancy but are now recognizing that pregnant patients are able to receive chemotherapy. As the majority of women do not receive chemotherapy during organogenesis, the malformation rate is not higher than the general population. The majority of organogenesis is completed by 12 weeks, yet the central nervous system continues to develop throughout gestation and after birth. A review summarized 340 fetal exposures to chemotherapy published to date and later expanded to 447 cases in 2008. The literature at that time provided details of the cancer diagnosis, treatment, and general pregnancy outcomes, but long-term follow-up on the children was limited. Avilés and Neri in 2001 reported long-term follow-up on 84 children born to mothers with hematological malignancies exposed in utero to chemotherapy, ranging in age from 6–29 years. All children were found to be normal physically and neurologically. School performances and standardized intelligence testing were within normal range and were not significantly different from controls (unrelated matched children and unexposed siblings). Eleven years later, a prospective study on the neurodevelopmental outcomes of 70 children aged 18 months to 18 years exposed to cancer treatment in utero was conducted, assessing health status, cognitive performance, and behavioral competence. This study reported that the majority of these children were doing well, and those children showing delays in development were concentrated in the group delivered preterm. This study lacked a control group of unexposed children and was unable to determine whether developmental delays were related to chemotherapy exposure or prematurity.
Documenting the long-term follow-up on children exposed to chemotherapy in utero could provide women and their physicians the information necessary to make informed decisions during diagnosis of cancer during pregnancy. In the present study, we report on cognitive and behavioral outcomes for children born to mothers with cancer diagnoses during pregnancy with comparisons of those children exposed in utero to chemotherapy with a control group of nonexposed children of mothers diagnosed during pregnancy.
Materials and Methods
Study sample
A cohort of women diagnosed with cancer during pregnancy was enrolled and followed up in the international Cancer and Pregnancy Registry. Establishment and conduction of follow-up for the registry was approved by the Institutional Review Board of Cooper Medical School at Rowan University. Since collecting cases in 1997, the registry is compiled of 338 pregnant women diagnosed with various types of cancer. Pregnant women, enrolled at the time of their cancer diagnosis, provide information, verified by medical records, of their diagnosis and treatment during pregnancy. Treatment course was determined by the oncologist caring for the patient based on stage of cancer and gestational age at diagnosis.
Women participating in ongoing data collection for the Cancer and Pregnancy Registry were offered standardized developmental testing if their child was at least 18 months of age. Letters were sent to 149 eligible participants regardless of cancer type or treatment during pregnancy. Patients were not paid for their participation, but travel expenses and parking were covered by grants from the Cooper Cancer Institute. Parents were allowed to be present during testing. In all, 53 women diagnosed with cancer during their pregnancy provided consent for their children to undergo developmental and behavioral assessments, and in 2 cases primary caregivers in families in which the mothers were deceased provided consent. Study participants were separated into 2 groups depending on type of cancer treatment (eg, chemotherapy vs nonchemotherapy) received before delivery. In the chemotherapy exposure group (n = 35), 26 women were diagnosed with breast cancer during pregnancy. Other cancer types in this exposed group included ovarian (4), Hodgkin disease (4), and acute leukemia (1). The control group, those who did not receive chemotherapy during pregnancy (n = 22), included the following cancer types: breast (6); melanoma (6); 2 cases each of cervical, bladder, central nervous system, and Hodgkin disease; and 1 case each of rectal and parotid gland tumors.
Participants
Including 2 sets of twins, 57 children of 55 mothers diagnosed with cancer while pregnant served as the study subjects and were assessed on age-appropriate developmental tests. In all, 35 children exposed in utero to chemotherapy comprised the exposure group and 22 children of women diagnosed with cancer who did not undergo chemotherapy during pregnancy comprised the control group. During pregnancy women diagnosed with breast cancer received doxorubicin/cyclophosphamide (22), doxorubicin/cyclophosphamide with 5 fluorouracil (3), or doxorubicin/cyclophosphamide followed by paclitaxel (1). Patients diagnosed with ovarian cancer were treated with cisplatin/paclitaxel (2), etoposide/cisplatin/bleomycin (1), or carboplatin/paclitaxel (1). Four patients with Hodgkin disease received doxorubicin/bleomycin/vinblastine/dacarbazine. The sole participant diagnosed with acute leukemia received cyclophosphamide, daunorubicin, vincristine, L-asparaginase, cytarabine, 6-mercaptopurine, and intrathecal methotrexate. No one received chemotherapy until after 12 completed weeks of pregnancy. Women requiring chemotherapy postpartum were advised by their physicians not to breast-feed their infant.
The children undergoing developmental testing ranged in age from 18 months to 10.4 years. The psychologist performing the evaluations was blinded to the maternal treatment during pregnancy. Two women in the study were deceased. One child (exposed group) lost his mother at age 4.6 years and was tested at age 7 years, the other child (control group) lost her mother when she was 9 months old and underwent testing at age 4.2 years.
Measures
Cognitive assessment
Bayley Scales of Infant Development–Third Edition (BSID-III) was administered to the 29 children aged 18-42 months. The BSID-III assesses development across 5 domains, with 3 used for the present study. Composite scores were generated for cognitive, language, and motor abilities (mean, 100; SD, 15). Scores <85 were considered abnormal.
The Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R) was administered to the 19 children aged 3-7 years. This battery of tests assesses intellectual functioning and provides composite scores that represent cognitive functioning in verbal (verbal intelligence quotient [VIQ]) and performance (performance intelligence quotient [PIQ]) domains, as well as a composite score (full-scale intelligence quotient [FSIQ]) reflective of a child’s overall intellectual ability (mean, 100; SD, 15). Scores <85 were considered abnormal.
The Wechsler Intelligence Scale for Children, Third Edition (WISC-III) was administered to the 9 children aged ≥7 years. Similar to the WPPSI-R, it provides composite scores of cognitive functioning (VIQ, PIQ, FSIQ) for the older children in the sample (mean, 100; SD, 15). Scores <85 were considered abnormal.
School performance
Wechsler Individual Achievement Test was administered to the 14 school-aged children ≥7 years of age. This provides a comprehensive test of reading, writing, language, and mathematics. For the present study, only scores for reading and mathematics were considered (mean, 100; SD, 15). Standard scores <25th percentile (<90), were below age and/or grade expectancy and considered abnormal.
Behavioral competence
The Child Behavior Checklist (CBCL) is a parent-report questionnaire on which the child is rated on various behavioral and emotional problems. The CBCL assesses internalizing (ie, anxious, depressive, and overcontrolled) and externalizing (ie, aggressive, hyperactive, noncompliant, and undercontrolled) behaviors. Scales are age-specific and generate internalizing scores, externalizing scores, and total-problem behavior scores. The CBCL ranks children’s behavior according to severity compared with other children of the same age. The CBCL scores are categorized by established clinical cutoffs to represent behavior problems according to risk level. Borderline clinical scores (T ≥65) were considered abnormal for this report. Similar questions are grouped into a number of syndrome scales (ie, attention, somatic symptoms) to produce a score for each syndrome. In all, 55 respondents (53 mothers, 1 father from a deceased mother family, and 1 grandmother from the other deceased mother family) completed the 118-item questionnaire for the 57 children.
Results
Demographic comparisons
Comparisons of the demographic variables can be found in Table 1 . Only mean gestational age at birth was significantly different between the 2 groups (36.7 vs 38.2 weeks; P = .04) with the chemotherapy-exposed group demonstrating higher levels of prematurity. Although the mean gestational age was significantly earlier in the chemotherapy-exposed group, the incidence of premature birth <36.9 weeks was not statistically different between the 2 groups (51.4% vs 38.1%; P = .33). One child in this cohort weighed within the <10% percentile at birth and was in the exposed group. Two children in the exposed group were diagnosed with congenital anomalies, each resolved by the time of developmental testing. One child was born with plagiocephaly, corrected with a helmet, and the other child was born with syndactyly of 2 fingers on 1 hand, surgically corrected soon after birth. The mean age at developmental evaluation was 4.5 ± 3.1 years in the exposed group and 4.9 ± 2.6 years in the control group ( P = .59).
| Variable | Exposure | n | Mean | SD | n (%) | P value |
|---|---|---|---|---|---|---|
| GA birth, wk | Chemo | 35 | 36.7 | ±2.5 | .04 a,b | |
| Controls | 22 | 38.2 | ±2.7 | |||
| Male sex | Chemo | 35 | 24 (68.6) | .16 c | ||
| Controls | 22 | 11 (50) | ||||
| Children with mother alive | Chemo | 35 | 34 (97.1) | 1.0 d | ||
| Controls | 22 | 21 (95.5) | ||||
| Children with mother undergoing treatment for recurrence | Chemo | 34 | 4 (11.8) | .46 d | ||
| Controls | 21 | 4 (19.0) | ||||
| Age at evaluation, y | Chemo | 35 | 4.5 | ±3.1 | .59 a | |
| Controls | 22 | 4.9 | ±2.6 | |||
| GA at first chemo treatment, wk (range) | 35 | 22.0 (11.7–31.3) | ±4.9 |
Cognitive assessment
Across the entire sample, 54 of 57 children (95%) scored within normal limits on age-appropriate cognitive assessments ( Table 2 ). Only 1 child of 35 children exposed to chemotherapy in utero had abnormal cognitive test results compared with 2 children (of 22) in the control group who tested abnormally. This difference between groups scoring within normal limits on assessments of cognitive ability was not statistically significant ( P = .55).
| Assessment | Exposure (n) | n (%WNL) | P value | Mean+ | SD | P value |
|---|---|---|---|---|---|---|
| Cognitive | n (%WNL) | |||||
| WNL | ||||||
| 54/57 | Chemo (35) | 34 (97) | .55 a | |||
| (95%) | Controls (22) | 20 (91) | ||||
| School performance math | ||||||
| WNL | Chemo (8) | 6 (75) | 1.0 a | 99.1 | 27 | .30 b |
| 10/14 | Controls (6) | 4 (67) | ||||
| (71%) | 113.8 | 22.3 | ||||
| School performance reading | ||||||
| WNL | Chemo (8) | 6 (75) | 1.0 a | 101.1 | 26.1 | .93 b |
| 11/14 | Controls (6) | 5 (83) | ||||
| (79%) | 102.2 | 17.3 | ||||
| Behavior | ||||||
| WNL | ||||||
| 45/57 | Chemo (35) | 27 (77) | .75 c | |||
| (79%) | Controls (22) | 18 (82) | ||||
| Internalizing behaviors | Chemo (35) | 46.7 | 8.7 | .32 b | ||
| Controls (22) | 44.1 | 10.4 | ||||
| Externalizing behaviors | Chemo (35) | 45.4 | 10 | .88 b | ||
| Controls (22) | 44.9 | 11.5 | ||||
| Total problem behaviors | Chemo (35) | 46.3 | 10 | .68 b | ||
| Controls (22) | 45.1 | 11.8 |
Analysis of group differences of cognitive ability across age cohorts showed similar results. No significant differences were noted in cognitive skills for the younger cohort on the BSID-III (cognitive, language scores) or for the older cohort on the WPPSI-R/WISC-III (VIQ, PIQ, or FSIQ scores) between exposed and nonexposed groups ( Table 3 ).
| Cognitive scale | Exposure | n | Mean | SD | P value |
|---|---|---|---|---|---|
| BSID-III, n = 29 | |||||
| Cognitive | .23 | ||||
| Chemo | 21 | 111.43 | 10.97 | ||
| Controls | 8 | 118.13 | 18.11 | ||
| Language | .15 | ||||
| Chemo | 21 | 105.43 | 10.86 | ||
| Controls | 8 | 112.63 | 13.50 | ||
| WPPSI-R/WISC-III, n = 28 | |||||
| VIQ | .94 | ||||
| Chemo | 14 | 113.00 | 18.50 | ||
| Controls | 14 | 113.57 | 20.70 | ||
| PIQ | .66 | ||||
| Chemo | 14 | 107.43 | 17.99 | ||
| Controls | 14 | 110.57 | 19.14 | ||
| FSIQ | .92 | ||||
| Chemo | 14 | 111.71 | 17.56 | ||
| Controls | 14 | 112.43 | 20.85 |
School performance
On tests of academic achievement for all children, 71% demonstrated mathematic ability at or above age equivalency, with 79% demonstrating at or above age expectancy in reading levels ( Table 2 ). Two children (25%) in the chemotherapy-exposed group and 2 children (33%) in the nonexposed group showed academic deficits in mathematics, and 2 children (25%) in the chemotherapy-exposed group and 1 child (17%) in the control group showed deficits in reading levels. Academic achievement scores did not differ between groups (for math, P = .30; for reading, P = .93).
Behavioral competence
For the entire sample, 79% of children scored in the normal range on maternal reports of problem behaviors ( Table 2 ). Twelve of the 57 children (21%) scored in the borderline clinical range on the maternal-reported behavior scales. Eight children in the exposed group (23%), demonstrated behavior problems in the clinical range, compared with 4 children (18%) in the control group showing similar behavioral issues ( P = .75). Mean scores were not statistically different for chemotherapy-exposed vs control children on measures of internalizing ( P = .32), externalizing ( P = .88), or total behavior problem ( P = .68) scales ( Table 2 ).
Demographic predictors of developmental outcome
Cognitive ability
Cognitive ability did not appear to be influenced by demographic variables ( Table 4 ). Mean gestational age at birth did not differ for children with normal (37.2 ± 2.6 weeks) and abnormal (37.0 ± 4.1 weeks) test results ( P = .87) and sex was not statistically different between these 2 groups ( P = .28). However, all 3 children scoring in the abnormal category for cognitive ability were male. On the BSID-III, males scored statistically lower than females on both the cognitive scales ( P = .031) and on the language scales ( P = .011). These significant sex differences were not found in the older children on the WPPSI-R/WISC-III.
| Variable | Cognitive category | Mean | SD | Percent | P value |
|---|---|---|---|---|---|
| GA birth, wk (n = 57) | WNL (54) | 37.2 | 2.6 | ||
| Abnormal (3) | 37.0 | 4.1 | .87 a | ||
| Male sex (n = 35) | WNL (32) | 59.3 | |||
| Abnormal (3) | 100 | .28 b | |||
| BSID-III, n = 29 | |||||
| Cognitive | |||||
| Male (n = 18) | 109.2 | 10.2 | |||
| Female (n = 11) | 120.0 | 15.5 | .03 | ||
| Language | |||||
| Male (n = 18) | 103.2 | 11.4 | .01 | ||
| Female (n = 11) | 114.4 | 9.2 | |||
| WPPSI-R/WISC-III, n = 28 | |||||
| VIQ | |||||
| Male (n = 17) | 115.5 | 22.5 | |||
| Female (n = 11) | 109.8 | 13.0 | .45 | ||
| PIQ | |||||
| Male (n = 17) | 106.8 | 20.5 | |||
| Female (n = 11) | 112.4 | 14.5 | .44 | ||
| FSIQ | |||||
| Male (n = 17) | 112 | 21.9 | |||
| Female (n = 11) | 112.2 | 14.1 | .98 | ||
| Mother alive (n = 55) | WNL (52) | ||||
| Abnormal (3) | 94.5 | ||||
| Mother deceased (n = 2) | WNL (2) | 100 | 1.0 b | ||
| Abnormal (0) | |||||
| Mother undergoing treatment (n = 8) | WNL (7) | 87.5 | |||
| Abnormal (1) | |||||
| .27 b | |||||
| Mother cancer free (n = 47) | WNL (45) | 95.7 | |||
| Abnormal (2) | |||||
| Age at evaluation, y (n = 57) | WNL (54) | .08 | |||
| Abnormal (3) | 4.5 | ||||
| 7.5 | |||||
| GA at first chemotherapy, wk (n = 35) | WNL (34) | ||||
| Abnormal (1) | 21.9 ± 5.2 | ||||
| 23.3 c |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
