SEARCH

SEARCH BY CITATION

Keywords:

  • birth rates;
  • cancer;
  • survivors;
  • female;
  • trends

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

BACKGROUND:

More women of fertile age are long-term survivors of cancer. However, population-based data on birth rates of female cancer survivors are rare.

METHODS:

A total of 42,691 women ≤ 45 years with a history of cancer were identified from the Swedish Multi-Generation Register and the Swedish Cancer Register, for whom relative birth rates were calculated as compared to the background population, ie, standardized birth ratios (SBRs). Independent factors associated with reduced birth rates among cancer survivors were estimated using Poisson modeling.

RESULTS:

Compared to the background population, cancer survivors were 27% less likely to give birth (SBR = 0.73, 95% confidence interval [CI] = 0.72-0.75). Large difference in SBRs existed by cancer site, with high SBRs for survivors of melanoma skin, thoracic, head and neck, and thyroid cancers, and low SBRs for reproductive, breast, brain and eye, and hematopoietic cancer survivors. Parity status at diagnosis affected fertility: women who already had a child at the time of diagnosis were less likely to give birth (SBR = 0.50, 95% CI = 0.48-0.53) than were nulliparous women (SBR = 0.87, 95% CI = 0.85-0.90). Multivariate analysis showed that cancer site (reproductive organs), age at onset of cancer (< 12 years), and parity status were all significant and independent predictors of a reduced probability of giving birth after diagnosis.

CONCLUSIONS:

Cancer survivors are less likely to give birth compared with the background population. Large variations in the likelihood to give birth after diagnosis were seen according to age at onset, cancer site, and parity status at diagnosis. Cancer 2013. © 2013 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

A substantial proportion of cancers are diagnosed prior to and during a woman's fertile years.1 As a result of increasing age at first birth, growing effectiveness of treatment, and advances in fertility preservation techniques, doctors are increasingly confronted with young female cancer survivors who wish to have a (or another) child. Fertility after cancer may be compromised by the cancer directly, through interference with the physiology of the reproductive organs, or by locoregional treatment. Systemic treatment, including cytotoxic and hormone therapy, may induce temporary or permanent ovarian failure.2-4 Some population-based studies have shown that cancer patients are less likely to give birth after their disease as compared with their cancer-free siblings.5, 6 A recent study by Stensheim et al, using a population-based matched cohort study, have demonstrated that overall, female cancer patients have decreased fertility after diagnosis and that fertility-preserving techniques for ovarian cancer patients have been successful in more recent times.7

Over the past decades, we have seen advances in medicine, which may have had a dual impact on fertility in cancer survivors. First, systemic treatment regimens, especially those for childhood leukemia and premenopausal breast cancer, have become more toxic and as such carry a higher risk of inducing permanent amenorrhea.8, 9 On the other hand, we have seen important advances in the development of fertility preservation techniques over the past decades, such as in vitro fertilization, ovary tissue transplantation, and oocyte harvesting.10-12

In this study, we evaluated trends in birth rates among female cancer survivors, and factors affecting these rates.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Study Design

We used a cohort of female residents in Sweden who have a unique national registration number, which allows linkage of different records of personal information from the Multi-Generation Register,13 the Swedish Cancer Register,14 the Cause of Death Register, and the Migration Register. Further linkages were made to the censuses of 1960, 1970, 1980, and 1990, which contain information on the socioeconomic status of each Swedish citizen. Socioeconomic status was estimated based on information about the highest level of employment in the household given in the censuses and was categorized into 5 groups: blue-collar workers, white-collar workers, self-employed workers, farmers, and unclassified. In total, this database contains more than 11 million individuals, belonging to around 3 million families and including more than a million individuals with cancer diagnosed between 1958 and 2001.13 The study was approved by the institutional review board at the Karolinska Institutet, Solna, Sweden.

From the study population of 3,907,708 Swedish women, we selected all women aged 16 to 45 years, born after 1931, and who were alive in 1960. For these women, we ascertained number and dates of live childbirths, socioeconomic status, date of cancer diagnosis (if present), and followed them until death, emigration, or end of follow-up (December 31, 2002), whichever came first. End of follow-up was chosen as December 31, 2002, due to the availability of the numerous registers mentioned previously at the start of this study and updating the data would take several years to complete. For women diagnosed with cancer, we calculated the proportion and relative probability of giving birth after diagnosis. We excluded births occurring within 12 months after diagnosis. By choosing the 1-year cutoff level, we have included women who conceived at least 3 months after they got their cancer diagnosis. By that time, most patients would have received complete information on treatment plan and prognosis. Also, the decision to get pregnant was taken with the knowledge of the cancer diagnosis. For the background population, cancer-free women were matched to the female cancer survivors according to attained age and year of birth, and number of live childbirths of these women in the background population was then ascertained. Using these criteria, we identified 42,691 women with a history of cancer and 3,011,113 persons from the background population. Over the study period of 43 years, 5060 participants (0.1%) were lost to follow-up and 848,844 (21.7%) were excluded from the study population because they were out of the observed age range of 16 to 45 years old. Missing data is equally likely in both the female cancer survivors and the background population, which is why we chose to use standardized birth ratio to handle the nondifferential bias.

Statistical Analysis

Relative birth rates were expressed as a standardized birth ratio (SBR), ie, the ratio of observed births in our study group to the expected number of births, based on birth rates in the background population. Person-years at risk of giving birth were calculated from 1 year after the time of cancer diagnosis or 16th birthday, whichever came later, until age 45 years. We chose to start follow-up at age 16 years, because it is the legal age at which consensual intercourse is allowed, and we chose to censor follow-up at 45 years of age due to very few births occurring beyond that age. Background birth rates were derived using birth rates for all cancer-free women present in the Multi-Generation Register. The expected number of births was calculated by multiplying the 5-year age- and calendar-period–specific birth rates in the background population by correspondingly stratified person-years at risk and summing all the products.

We then compared relative birth rates by age at onset (childhood defined as below the age of 13 years, adolescence as 13 to 18 years, and adulthood as > 18 years), attained age, year and parity status at diagnosis (nulliparous and parous), cancer site, time since diagnosis, attained year, and socioeconomic status. Cancer site is classified into 12 main categories by International Classification of Diseases, revision 7 (ICD7): head and neck, digestive, thoracic, breast, reproductive (minus ovary), ovary, melanoma skin, brain and eye, thyroid, bone, hematopoietic, and all other cancer sites.15 We also estimated the relative birth rates of twinning among female cancer survivors as compared with the age- and period-specific rates of multiple births in the background population.

To identify associated predictors of giving birth after diagnosis, we calculated adjusted rate ratios by means of applied multivariate Poisson regression modeling using only data from cancer survivors adjusted for parity status at diagnosis, cancer site, age and year of diagnosis, time since diagnosis, calendar period, attained age, and socioeconomic status. These relative risks were expressed as birth rate ratios. Data preparation and analysis was done with SAS, version 9.2 (SAS Institute, Cary, NC).

Role of the Funding Source

The funders had no role in the study design, collection, analysis, data interpretation, or writing of the article. The corresponding author had full access to all data in the study and had final responsibility for the decision to submit for publication.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

In the study population of 3,907,708 Swedish women, 42,691 (1.1%) were diagnosed with cancer before the age of 45 years. The age at diagnosis for female cancer survivors had a median of 36 years and a range of 0 to 44 years over a median follow-up of 4.5 years. Most women were diagnosed with cancer during adulthood (92.7%) and the main cancer sites included breast (n = 12,139), reproductive organs (n = 6493), and melanoma skin (n = 4637) (Table 1). Of these cancer survivors, 4932 (11.6%) gave birth to 8080 children, 1 year or more after cancer whereas 1,864,511 of the background population (61.9%) gave birth to 3,786,420 children over the same calendar period and attained age.

Table 1. Women Diagnosed With Cancer in Sweden From 1960 to 2002, With Information on Subsequent Birth(s) by Age and Year of Diagnosis, Parity Status at Diagnosis, Cancer Site, and Socioeconomic Status
CharacteristicICD7 CodesaNo. of Women With Previous CancerNo. of Women With BirthsTotal No. of BirthsProportion of Women Who Gave Birth After Cancer Diagnosis
  • a

    ICD7 indicates International Classification of Diseases, 7th edition.

Overall42,6914932808011.6%
Parity status at diagnosis     
 Nulliparous13,4373276598224.4%
 Parous29,254165620985.7%
Cancer site     
 Breast170.0-170.912,1391242141.0%
 Reproductive (minus ovary)171.0-174.9; 176.0-176.964933586485.5%
 Melanoma skin190.0-190.9463731520.7%
 Brain and eye192.0-193.937973424709.0%
 Thyroid194.0-195.936823074648.3%
 Hematopoietic201.0-208.9256626844810.4%
 Ovary175.0-175.92553958152137.5%
 Digestive tract150.0-158.92439639108426.2%
 Bone196.0-200.92109885143042.0%
 All other cancer sites180.0-181.9; 191.0-191.9; 209.0-209.9122637667730.7%
 Head and neck140.0-148.9; 160.0-161.967446577669.0%
 Thoracic162.0-164.937617929647.6%
Age at diagnosis (y)     
Childhood (younger than 13)152144086728.9%
 Adolescence (13-18)1618638128239.4%
 Adulthood (19 or older)39,552385459319.7%
Year of diagnosis     
 1958-19682333841156036.0%
 1969-197992191292224314.0%
 1980-199015,4431664277710.8%
 1991-200115,696113515007.2%
Time since diagnosis (y)     
 1-5.927313990
 6-10.911792123
 11-20.98051540
 21-40217427
Socioeconomic status     
 Blue-collar workers14,8001635268711.0%
 White-collar workers18,1982236372712.3%
 Self-employed1134991748.7%
 Farmers34431529.0%
 Unclassified82159311,44011.3%

Overall, cancer survivors were 27% less likely to give birth than the general population (SBR = 0.73, 95% CI = 0.72-0.75) (Table 2), a proportion that did not change over time. Survivors of thyroid, head and neck, thoracic, and melanoma skin cancer had practically similar relative birth rates as the background population (SBR = 0.98, 95% CI = 0.93-1.03; SBR = 1.00, 95% CI = 0.87-1.14; SBR = 1.01, 95% CI = 0.76-1.31; SBR = 1.04, 95% CI = 0.99-1.10, respectively), whereas women treated for cancer of the reproductive organs (minus ovary), breast, and ovary were significantly less likely to give birth than the background population (SBR = 0.32, 95% CI = 0.29-0.35; SBR = 0.52, 95% CI = 0.47-0.57; and SBR = 0.56, 95% CI = 0.51-0.61, respectively).

Table 2. Standardized Birth Ratios of Women With a History of Cancer, 1960 to 2002
CharacteristicStandardized Birth Ratio (95% Confidence Interval)Background Population Birth Rates per 10,000 Person-YearsSurvivor Population Birth Rates per 10,000 Person-Years
Overall0.73 (0.72-0.75)389285
Year of diagnosis   
 1958-19680.77 (0.73-0.81)518400
 1969-19790.72 (0.69-0.75)357256
 1980-19900.73 (0.71-0.76)370271
 1991-20010.72 (0.68-0.76)383276
Attained age (y)   
 16-20.90.76 (0.66-0.87)225172
 21-25.90.81 (0.77-0.85)921747
 26-30.90.78 (0.75-0.81)1180923
 31-25.90.71 (0.68-0.74)715509
 36-40.90.65 (0.61-0.68)243157
 41-45.90.51 (0.44-0.59)3618
Cancer site   
 Reproductive (minus ovary)0.32 (0.29-0.35)
 Breast0.52 (0.47-0.57)
 Ovary0.56 (0.51-0.61)
 Brain and eye0.67 (0.63-0.71)
 Hematopoietic0.68 (0.63-0.73)
 All other cancer sites0.82 (0.73-0.92)
 Bone0.83 (0.77-0.89)
 Digestive tract0.90 (0.83-0.97)
 Thyroid0.98 (0.93-1.03)
 Head and neck1.00 (0.87-1.14)
 Thoracic1.01 (0.76-1.31)
 Skin1.04 (0.99-1.10)
Age at diagnosis (y)   
 Childhood (younger than 13)0.72 (0.67-0.77)
  Diagnosis pre-19800.76 (0.70-0.81)
  Diagnosis 1980 onward0.57 (0.47-0.67)
 Adolescence (13-18)0.83 (0.79-0.88)
 Adulthood (19 or older)0.72 (0.70-0.74)
Time since diagnosis (y)   
 1-5.90.70 (0.68-0.72)
 6-10.90.78 (0.74-0.81)
 11-20.90.78 (0.74-0.82)
 21-400.73 (0.67-0.81)
Socioeconomic Status   
 Blue-collar Workers0.67 (0.65-0.70)
 White-collar Workers0.81 (0.78-0.83)
 Self-employed0.73 (0.62-0.84)
 Farmers0.85 (0.64-1.10)
 Unclassified0.69 (0.65-0.73)

For survivors of childhood cancer, those diagnosed before 1980 had higher SBRs than those diagnosed from 1980 onward (SBR = 0.76, 95% CI = 0.70-0.81; SBR = 0.57, 95% CI = 0.47-0.67, respectively). Comparing age at diagnosis, survivors of adolescent cancer had higher SBRs than survivors of childhood and adult cancer (SBR = 0.83, 95% CI = 0.79-0.88; SBR = 0.72, 95% CI = 0.67-0.77; SBR = 0.72, 95% CI = 0.70-0.74, respectively). Low SBRs were also observed from women who were recently diagnosed and women older than 36 years (Table 2).

Women who were parous at diagnosis had a lower SBR (0.50, 95% CI = 0.48-0.53) than women who were nulliparous at diagnosis (SBR = 0.87, 95% CI = 0.85-0.90), a pattern that was observed for all cancer sites (Table 3).

Table 3. Standardized Birth Ratios of Women With a History of Cancer From 1960 to 2002, by Parity Status at diagnosis and cancer site
 ParousNulliparous
CharacteristicStandardized Birth Ratio (95% Confidence Interval)Standardized Birth Ratio (95% Confidence Interval)
Overall0.50 (0.48-0.53)0.87 (0.85-0.90)
Cancer site  
 Reproductive (minus ovary)0.20 (0.17-0.23)0.53 (0.47-0.59)
 Breast0.35 (0.31-0.40)0.91 (0.80-1.02)
 Hematopoietic0.38 (0.31-0.46)0.77 (0.71-0.83)
 Ovary0.46 (0.38-0.55)0.61 (0.54-0.67)
 Bone0.54 (0.44-0.65)0.92 (0.85-1.00)
 Brain and eye0.59 (0.51-0.67)0.69 (0.65-0.74)
 Digestive tract0.58 (0.48-0.69)1.03 (0.94-1.12)
 Thoracic0.71 (0.41-1.09)1.27 (0.89-1.73)
 All other cancer sites0.73 (0.58-0.90)0.86 (0.75-0.98)
 Head and neck0.76 (0.58-0.97)1.13 (0.96-1.31)
 Melanoma skin0.79 (0.72-0.87)1.20 (1.13-1.27)
 Thyroid0.81 (0.73-0.88)1.08 (1.02-1.15)
Age at diagnosis (y)  
 0-<190.71 (0.43-1.05)0.79 (0.75-0.82)
 19 or older0.50 (0.48-0.52)0.93 (0.90-0.96)
Attained age (y)  
 16-20.92.91 (0.92-6.01)0.75 (0.65-0.86)
 21-25.91.16 (0.99-1.35)0.78 (0.73-0.82)
 26-30.90.60 (0.55-0.65)0.85 (0.81-0.89)
 31-25.90.46 (0.43-0.49)0.95 (0.91-1.00)
 36-40.90.43 (0.40-0.47)1.03 (0.95-1.11)
 41-45.90.37 (0.31-0.45)0.92 (0.74-1.13)

Overall, the relative risk of twinning among the female cancer survivors was not larger than that of the background population, neither in the 1980s (SBR = 0.82, 95% CI = 0.54-1.17) nor in the 1990s (SBR = 0.92, 95% CI = 0.73-1.13).

With increasing age, survivors of adult cancer became less likely to give birth compared with women of similar age in the background population (Fig. 1). Survivors of childhood and adolescent cancer, on the other hand, showed a steep increase in relative fertility after the age of 35 years.

thumbnail image

Figure 1. Standardized birth ratio is shown for women with a history of cancer with increasing age by age at diagnosis.

Download figure to PowerPoint

Multivariate Poisson modeling, adjusting for attained age and year, year of birth, and period of diagnosis, showed that parity status, cancer site, age at onset, time since diagnosis, and socioeconomic status were all independently and significantly associated with the probability of giving birth after cancer diagnosis (Table 4). Cancer of the reproductive organs, recent cancer diagnosis, diagnosis during childhood, and being parous at diagnosis were significant and independent predictors of a reduced probability of giving birth after diagnosis.

Table 4. Poisson Regression Model of Relative Likelihood of Giving Birth for Women With a History of Cancer
CharacteristicUnadjusted Birth Rate Ratio (95% Confidenc Interval)Adjusted Birth Rate Ratioa (95% Confidence Interval)
  • a

    Birth rate ratios from multivariate Poisson model with the following covariates in the model: parity status at diagnosis, cancer site, age and year of diagnosis, time since diagnosis, calendar period, attained age, and socioeconomic status.

Parity status at diagnosis  
 Nulliparous1.001.00
 Parous0.36 (0.34-0.38)0.62 (0.58-0.66)
Cancer site  
 Melanoma skin1.001.00
 Breast0.24 (0.21-0.26)0.52 (0.47-0.58)
 Reproductive (minus ovary)0.34 (0.31-0.38)0.40 (0.36-0.45)
 Brain and eye0.83 (0.74-0.92)0.68 (0.61-0.75)
 Ovary0.84 (0.78-0.91)0.70 (0.64-0.76)
 Hematopoietic0.93 (0.85-1.01)0.72 (0.66-0.78)
 Thyroid1.01 (0.94-1.08)0.97 (0.90-1.04)
 Bone1.16 (1.06-1.27)0.93 (0.85-1.02)
 All other cancer sites1.23 (1.08-1.39)1.05 (0.92-1.19)
 Digestive tract1.26 (1.15-1.38)1.03 (0.94-1.13)
 Head and neck1.63 (1.41-1.88)1.33 (1.15-1.54)
 Thoracic3.04 (2.30-4.00)1.98 (1.50-2.62)
Age at diagnosis (years)  
 Childhood (younger than 13)1.001.00
 Adolescence (13-18)1.31 (1.21-1.43)1.29 (1.15-1.45)
 Adulthood (19 or older)0.80 (0.74-0.85)1.53 (1.34-1.75)
Year of diagnosis  
 1958-19681.001.00
 1969-19790.67 (0.63-0.71)0.82 (0.75-0.90)
 1980-19900.60 (0.56-0.64)0.75 (0.66-0.85)
 1991-20010.54 (0.50-0.58)0.70 (0.60-0.81)
Time since diagnosis (y)  
 1-5.91.001.00
 6-10.91.52 (1.44-1.62)1.04 (0.98-1.11)
 11-20.91.89 (1.71-2.08)1.12 (1.02-1.23)
 21-401.36 (1.29-1.44)1.27 (1.06-1.52)
Calendar period  
 1960-19691.001.00
 1970-19790.64 (0.58-0.71)0.88 (0.78-0.99)
 1980-19890.48 (0.44-0.53)0.95 (0.83-1.10)
 1990-20020.58 (0.53-0.64)1.17 (0.99-1.39)
Attained age (y)  
 16-20.91.001.00
 21-25.94.16 (3.59-4.83)3.57 (3.06-4.17)
 26-30.95.21 (4.51-6.01)4.53 (3.86-5.32)
 31-25.93.16 (2.73-3.64)3.11 (2.62-3.69)
 36-40.91.28 (1.11-1.49)1.46 (1.22-1.76)
 41-45.90.37 (0.30-0.45)0.50 (0.40-0.63)
Socioeconomic status  
 Blue-collar workers1.001.00
 White-collar workers1.02 (0.97-1.08)1.02 (0.97-1.08)
 Self-employed2.36 (2.03-2.75)1.94 (1.66-2.26)
 Farmers3.94 (2.99-5.18)2.28 (1.72-3.02)
 Unclassfied1.13 (1.06-1.21)0.97 (0.91-1.04)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

More women of fertile age are long-term survivors of childhood, adolescent, and adult cancer. However, population-based data on birth rates of female cancer survivors are rare.

This study shows that female cancer survivors are 27% less likely to give birth as compared with the background population, and that there is large variation in birth rates depending on age at diagnosis, attained age, cancer site, time since diagnosis, and parity status at diagnosis. Cancer survivors with lowest fertility were those with previous childbirth, recent diagnosis, cancer of the reproductive organs and breast, those diagnosed during childhood, and those older than 35 years.

According to Hewitt et al, more than 75% of girls and women who are diagnosed with cancer before the age of 45 years survive at least 5 years, and about 1 in every 1000 adults is a survivor of childhood cancer.16 More recently, 8 of every 10 children diagnosed with cancer will survive for 5 years or more.17 Reasons for low birth rates among cancer survivors are multifactorial and may include the impact of the cancer process itself, the impact of locoregional and systemic treatment, negative advice from doctors in getting pregnant after cancer, and psychosocial factors.

Most cancer types have no direct effect on a woman's oocytes and hormone levels, except for some brain tumors (pituitary gland) and uterine, cervical, and genital cancers.18 Cancer treatment on the other hand, may result in mechanical, hormonal, and toxic damage to ovaries and reproductive organs, which may render survivors physiologically unable to give birth. For example, surgery and radiotherapy of colorectal and reproductive cancers may impair the ability to conceive and complete pregnancy. Cytotoxic treatment often results in ovarian damage or failure, mainly of the maturing follicles, and the extent of damage depends on the type of cytotoxic regimens received, the dosage, and the patient's age.19-22 Until recently, pregnancy was thought to have a detrimental impact on the outcome of hormone-dependent cancers, particularly in breast cancer, which often resulted in negative doctor's advice to survivors with a wish to conceive. However, recent studies have now convincingly shown that pregnancy is unlikely to impair the outcome of breast cancer.23, 24

Another reason for low birth rates may be related to the fact that cancer survivors suffer more often from psychological conditions such as depression and anxiety disorders25, 26 and often have less stable incomes, job prospects, financial security, or fewer stable relationships.27-30 These factors, in addition to fear of heritability of the disease, and uncertainty about potential disease recurrence, may contribute to the survivors' hesitation toward starting or expanding a family.31, 32

We observed very different fertility patterns for women with and without a child at diagnosis. Women who were parous at diagnosis were substantially less likely to give birth than those who were nulliparous at diagnosis even after adjustment for cancer site and age at diagnosis (Table 4). Nulliparous women had increasing fertility with increasing age and achieved similar fertility as the background population by age 31 onward (Table 3). This very likely represents a selection process, where nulliparous women are more eager to conceive, rather than biological ability, because there is no obvious reason for disease extent and treatment to differ by parity status at diagnosis.

Fertility patterns by increasing age vary by age at diagnosis (Fig. 1). The reduction in relative fertility with increasing age, starting at age 30 years in women diagnosed as adults, suggests that cancer survivors may have less ovarian reserve and may be more prone to accelerated oocyte atresia and decreased oocyte quality.33, 34 In contrast, for women with childhood or adolescent cancer, there are marked increases in relative fertility after the age of 35 years, reaching virtually the same fertility as the background population. The increasing fertility by increasing age may represent an increased willingness of older cancer survivors to have a (or another) child, and it is plausible that the increase in birth rates among women aged more than 35 years who were diagnosed before adulthood could be due to improvements in and increasing use of fertility preservation techniques, such as in vitro fertilization and embryo oocyte or ovary/ovarian tissue cryopreservation.3, 10, 12

Cancer site, age, and parity status at diagnosis, and time since diagnosis were all independently associated with fertility after diagnosis. It is not surprising to see large differences in fertility by cancer site, which are due to differences in disease-associated prognosis and toxicity of treatment regimens. Notably, we see decreased birth rates among recent childhood cancer survivors as compared with those diagnosed before 1980, which likely reflects more toxic treatments (Table 2).The independent effect of increasing fertility by time since diagnosis may represent an increased ovarian capacity following cancer treatment.

Some strengths of our study are the population-based design and an unbiased ascertainment of cancers, deaths, and births. By calculating SBRs, we have taken into account societal changes in reproductive behavior over time. We acknowledge that our study suffers from some limitations, including the absence of information on spontaneous and induced abortions, as well as the lack of treatment details and use of fertility preservation techniques which would have allowed for a deeper understanding of the observed trends. Our findings of an overall decreased fertility for female cancer patients as well as an association of parity status at diagnosis to postdiagnostic fertility has also been demonstrated in other similar populations.5, 7

The birth rates presented in this study may be somewhat different in other populations due to social and medical reasons. In Sweden, the standard of medicine is high, and as a result, all women have access to complete treatment and have relatively high survival probabilities. Facilities for assisted reproduction are common.35 Also, Swedish women make independent decisions regarding abortion and whether to conceive. These factors, which may be different in other regions of the world, all affect the probability of giving birth, both of the cancer survivors and the background populations.

In summary, a cancer diagnosis has an important impact on a woman's likelihood to give birth after diagnosis. Birth patterns vary with cancer site, attained age, and with age of cancer diagnosis. Large differences in SBRs by parity status at diagnosis are suggestive of differential willingness among cancer survivors to pursue a (or another) child and that the true infertility following cancer is only marginally less than what is observed in cancer-free women. Identifying women at increased risk of infertility after diagnosis will allow for possible family planning and would still give women the choice of parenthood after a cancer diagnosis.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

This study was funded by the Swedish Research Council (SIMSAM grant number 80748301), National Medical Research Council (NMRC/1180/2008), and National University of Singapore Start-up Fund DPRT.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosure.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES