Blood Pressure Status in Adult Survivors of Childhood Cancer: A Report from the St. Jude Lifetime Cohort Study

Improvements in treatment and supportive care have dramatically increased survival after pediatric malignancies, with more than 83% of children now surviving at least 5 years and most becoming long-term survivors (1). Unfortunately, the growing population of childhood cancer survivors, currently exceeding 420,000 in the United States, experiences a wide range of adverse late health effects (2). Prominent among these late effects is cardiovascular disease, the leading noncancer cause of mortality and morbidity in survivors (3–8). Cardiovascular disease–related deaths are eight times more likely in childhood cancer survivors compared with the general population (9), and serious cardiac events are more than five times more likely in survivors compared with siblings (6).

Much of the increased risk of cardiovascular disease in survivors can be attributed to childhood exposure to chest-directed radiation and/or high-dose anthracyclines. However, aging survivors are increasingly susceptible to the same risk factors that impact the general population, including hypertension, obesity, and dyslipidemia (10). Cancer therapies, including exposure to specific chemotherapeutic agents (alkylating agents, antimetabolites, and heavy metals) and abdominal radiation, are considered risk factors for hypertension and indicators for active surveillance (11). Hypertension is a leading cause of cardiovascular disease in the general population (12), with recent evidence indicating it is particularly harmful in survivors of childhood cancer previously exposed to cardiotoxic treatments. A report from the Childhood Cancer Survivor Study (CCSS) found that although both cardiotoxic treatments and hypertension were independently associated with increased risk of coronary artery disease or heart failure, the combination of these factors resulted in a greater than additive increase in risk that yielded an 86-fold increased risk of heart failure in survivors exposed to both anthracyclines and hypertension compared with neither factor (10). This suggests that development of hypertension can exacerbate the damage caused by cardiotoxic cancer treatments (13, 14).

Given the importance of hypertension as an adverse modifying factor for cardiac disease in childhood cancer survivors, there is a clear need to better understand the blood pressure status of this unique population. Limitations of previous studies such as reliance on self-reported hypertension, small samples, inclusion of only specific diagnosis groups, and lack of age-specific rates have precluded a comprehensive understanding of blood pressure abnormalities in adult survivors of childhood cancer (15–27). The goal of this analysis was to examine clinically assessed blood pressures in a large cohort of well-characterized childhood cancer survivors with longitudinal follow-up.

Survivors of childhood cancer treated at St. Jude Children's Research Hospital (SJCRH, Memphis, TN) who were at least 18 years of age and at least 10 years from diagnosis were eligible to enroll in the St. Jude Lifetime Cohort Study (SJLIFE). The goal of SJLIFE is to provide ongoing medical assessments of survivors to advance knowledge of long-term health outcomes; details regarding recruitment and study design have been presented previously (28, 29). In brief, the SJLIFE study is a retrospective cohort with prospective follow-up in which participants returned to SJCRH to undergo comprehensive risk-based medical evaluations based on the Children's Oncology Group's Long Term Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers (30). Data collection included abstraction of medical records, completion of health questionnaires, neuromuscular functional assessment, and collection of biological samples. Additional retrospective clinical and diagnostic records were obtained from community providers and the After Completion of Therapy (ACT) Clinic at SJCRH and validated for presence of a prior hypertension diagnosis and treatment (28). The protocol was approved by the SJCRH Institutional Review Board, and all participants provided informed consent. Of 5,067 eligible survivors, 3,016 enrolled in SJLIFE and completed a baseline on-campus evaluation as of June 30, 2015, that included a standardized measurement of resting blood pressure in the St. Jude Human Performance Laboratory. Participants are brought back to SJCRH for follow-up assessments at least once every 5 years according to a research-driven schedule; at the time of this analysis, 1,185 participants had multiple visits resulting in longitudinal blood pressure measurements (Fig. 1).

Blood pressures were obtained by trained clinical staff following a 5-minute rest in a seated position. Three consecutive measurements, separated by a 1-minute rest, were obtained using a calibrated sphygmomanometer. At least two consecutive valid measurements were required for inclusion in this analysis, and blood pressure status was determined on the basis of the average of these measurements. Use of antihypertensive medications was determined by survivor report and review of medical records.

Blood pressure status was defined on the basis of the highest category of systolic or diastolic blood pressure (SBP/DBP) as normal (SBP <120 and DBP <80 mm Hg), prehypertension (SBP 120–139 or DBP 80–89) or hypertension (SBP ≥140 or DBP ≥90 mm Hg; ref. 31). Participants were considered to have hypertension if they had measured blood pressure in the hypertensive range or had been previously diagnosed and were taking antihypertensive medications. Date of onset was the start date for medications according to medical records or the date of hypertensive blood pressure measurement. Undiagnosed hypertensive blood pressure was defined as a measured blood pressure in the hypertensive range in a participant with no previously identified hypertension. Similarly, uncontrolled hypertension was defined as a measurement in the hypertensive range among participants with previously identified hypertension.

Details regarding childhood cancer diagnosis and treatment were obtained from medical records, including site and dose of radiotherapy received, as well as specific chemotherapeutic agents and cumulative doses administered. Alkylating agents were collectively characterized according to the cyclophosphamide equivalent dose (CED; ref. 32). Participant characteristics were obtained from structured questionnaires in conjunction with medical/social history obtained during SJLIFE clinic visits. These included annual household income (≤$20,000, $20,000–$59,999, or ≥$60,000) and smoking status (never, former, or current). Height and weight were measured and used to categorize participants into body mass index (BMI; kg/m²) categories per the World Health Organization (underweight <18.5, normal 18.5–24.9, overweight 25.0–29.9, or obese ≥30.0 kg/m²; ref. 33).

Demographic, diagnosis, and treatment characteristics were compared using χ² tests for participants and nonparticipants in SJLIFE, as well as for participants with a single baseline SJLIFE assessment and those with at least one subsequent follow-up assessment. We incorporated age at first identification of hypertension to plot age-specific Kaplan–Meier cumulative prevalence curves along with associated 95% confidence intervals (CI) among SJLIFE participants, both overall and stratified by sex or race. Time at risk for each individual ended at the earliest of hypertension identification or most recent SJLIFE assessment. In addition, participants were categorized according to blood pressure status at their first SJLIFE assessment, and characteristics associated with prevalence of hypertension were identified using multivariable logistic regression models to estimate ORs and 95% CIs. The observed prevalence of hypertension in survivors was compared with the prevalence that would be expected based on age-, sex-, race/ethnicity-, and BMI category–specific rates from the National Health and Nutrition Examination Survey (2011–2012 NHANES), which used similar methods to identify and classify hypertension (34).

Participants without previously diagnosed hypertension were considered to be at risk for undiagnosed hypertensive blood pressure, and we calculated the proportion of those participants who were identified as having hypertensive blood pressure at their initial SJLIFE assessment. Conversely, participants with a prior hypertension diagnosis were considered to be at risk for uncontrolled hypertension and served as the denominator for calculating the proportion of participants with uncontrolled hypertension at their initial SJLIFE assessment. Univariate associations between demographic and treatment characteristics and prevalence of undiagnosed or uncontrolled hypertension were estimated by logistic regression, and only characteristics associated with P < 0.10 were included in multivariable models. Finally, within the subgroup with multiple SJLIFE assessments, we examined longitudinal changes in blood pressure status and associations between participant characteristics and changes in blood pressure category.

Of the 3,016 participants in this analysis, a slightly higher proportion were male (52%), the majority were non-Hispanic white (83%), and the mean age at initial SJLIFE assessment was 32 years (Table 1). Participants were more likely to be female compared with nonparticipants, and modest differences in the distribution of primary cancer diagnoses were reflected in higher proportions of participants being exposed to radiation or chemotherapy. Participants who returned for multiple SJLIFE assessments were older and more likely to be non-Hispanic white, and they had higher rates of exposure to radiation, alkylating agents, and epipodophyllotoxins compared with those who had not yet returned for a second assessment.

The age-specific cumulative prevalence of hypertension in childhood cancer survivors increased sharply with age; 13% had hypertension identified by age 30 years, with this proportion reaching 37% by age 40 and exceeding 70% by age 50 (Fig. 2). Although relatively few participants completed a SJLIFE assessment after the age of 55 (n = 20), our results suggest that as many as 80% of survivors may have hypertension by this age. The age-specific cumulative prevalence increased more steeply with age compared with what would be expected based on rates in the general population, with survivors generally having a prevalence similar to that expected in people about 10 years older. The prevalence of hypertension varied by childhood cancer diagnosis, with survivors of Wilms tumor generally having the highest age-specific prevalence, but survivors of all diagnosis groups developed hypertension substantially more often than would be expected in the general population (Fig. 3).

At the time of their initial SJLIFE assessment, 44.0%, 33.6%, and 22.4% of survivors had a blood pressure status of normal, prehypertension, and hypertension, respectively (Table 2). The prevalence of hypertension was 2.6-fold (95% CI, 1.6–4.7) higher among childhood cancer survivors than expected, based on age-, sex-, race, and BMI-specific rates in the general population. Factors significantly associated with higher odds of having prevalent hypertension included male sex (OR, 1.38; 95% CI, 1.14–1.67), non-Hispanic black race (OR, 1.66; 95% CI, 1.28–2.16), older age at assessment (OR per one year of age, 1.10; 95% CI, 1.08–1.11), and being overweight (OR, 1.58; 95% CI, 1.21–2.07) or obese (OR, 3.02; 95% CI, 2.34–3.88). Exposures to radiotherapy or chemotherapy were not significantly associated with hypertension, with the exception of a CED >0 to ≤8,000 mg/m² (OR, 1.46; 95% CI, 1.10–1.93). Notably, higher doses of alkylating agents were not significantly associated with hypertension (OR, 1.15; 95% CI, 0.89–1.49). Survivors who underwent nephrectomy as part of their childhood cancer treatment were more likely to have developed hypertension (OR, 1.68; 95% CI, 1.11–2.53), but no significant association was found for exposure to abdominal or pelvic radiation (OR for ≥20 Gy compared with none, 1.23; 95% CI, 0.85–1.79).

Among 2,459 participants who attended their first SJLIFE assessment with no prior hypertension diagnosis, 207 (8.4%) were identified as having unrecognized hypertensive blood pressure (Table 3). Similar to prevalent hypertension overall, undiagnosed hypertensive blood pressure was more likely in males (OR, 2.92; 95% CI, 2.09–4.08), other race/ethnicity compared with non-Hispanic white (OR, 1.93; 95% CI, 1.34–2.78), older age at assessment (OR per one year of age, 1.08; 1.06–1.10), and overweight/obese compared with normal BMI (OR for overweight, 1.81; 95% CI, 1.16–2.81; OR for obese, 3.46; 95% CI, 2.31–5.18). A total of 675 participants were identified with hypertension based on our study criteria either before or during the first SJLIFE visit, but only 52% reported via questionnaire that they were aware of their hypertension.

Among the 557 participants who were diagnosed with hypertension prior to their first SJLIFE assessment, 124 (22.3%) were identified as having uncontrolled hypertensive blood pressure (Table 3). Significant risk factors included male sex (OR, 2.01; 95% CI, 1.26–3.21), race other than non-Hispanic white (OR, 3.80; 95% CI, 2.18–6.60), and older age (OR per year of age, 1.05; 95% CI, 1.02–1.08).

Among participants with normal or prehypertensive blood pressure status at the initial assessment, 4.5% or 21.2%, respectively, were identified as having hypertensive blood pressure at a subsequent assessment (mean time interval, 3.6 years; Table 4). Factors associated with subsequent development of hypertensive blood pressure included male sex (OR, 1.55; 95% CI, 1.02–2.38), older age at initial assessment (OR per year of age, 1.06; 95% CI, 1.03–1.09), and obesity (OR, 2.97; 95% CI, 1.77–4.97; Supplementary Table S1). Only 19.0% of participants with hypertension at the first assessment were found to have normal blood pressure at their subsequent visit, and most of those (65%) were on antihypertensive medications; 35.0% of those with hypertension at the first visit continued to have uncontrolled hypertensive blood pressure at the subsequent visit.

Hypertension is a leading contributor to cardiovascular morbidity and mortality in the general population (35, 36) and is even more deleterious among survivors of childhood cancer due to potentiation of cardiovascular damage caused by cancer treatments, including chest-directed radiotherapy and anthracycline chemotherapy. More than one in five survivors in SJLIFE were found to have this modifiable cardiovascular risk factor and 33% were identified with prehypertensive blood pressure, a risk factor for future hypertension (37). This study is the largest to date to examine measured blood pressure among survivors from different diagnosis groups, and the first to report prevalence by age and compare it with general population rates. Our results demonstrate that survivors have a higher prevalence of hypertension compared with individuals in the general population that is evident throughout the life course. Moreover, one in 12 survivors had undiagnosed hypertensive blood pressure, and more than 20% of those previously diagnosed with hypertension had persistent uncontrolled high blood pressure.

We used age-, sex-, race/ethnicity-, and BMI category–specific NHANES rates to calculate the expected prevalence of hypertensive blood pressure in the general population, so the increased prevalence in survivors is not readily explained by established hypertension risk factors. Childhood cancer treatments likely play a role, as supported by our findings that age-specific cumulative prevalence varied across diagnoses. However, it is notable that in a clinically assessed cohort of over 3,000 survivors with detailed treatment data, including radiation dosimetry and chemotherapy dosages, we identified only nephrectomy as a significant risk factor. Low-dose alkylating agents were also significantly associated with hypertension in the multivariable model, but the lack of any association at higher doses suggests this may not be a meaningful risk factor. We did not find a significant association between abdominal or pelvic radiation and hypertension, although the OR of 1.5 may be suggestive of a modest association. Previous findings regarding treatment factors and hypertension in childhood cancer survivors have been inconsistent, but the large CCSS (n = 8,599) and Dutch AMC/LATER (n = 1,442) studies reported significant associations between abdominal radiation and hypertension (OR, 1.9 and 2.5, respectively; refs. 21, 38). Survivors of Wilms tumor had the highest prevalence of hypertension in SJLIFE, supporting a role for renal etiology in a subgroup of survivors. However, our findings suggest that exposure to specific nephrotoxic treatments does not fully explain survivors' increased risk of hypertension.

Further research is needed to fully elucidate the multifactorial mechanisms underlying the increased risk of hypertension in childhood cancer survivors. In addition to direct tissue damage by radiation and specific chemotherapeutic agents, more general mechanisms related to cancer and its treatment may also contribute, such as chronic inflammation, endothelial dysfunction, altered immune function, and factors related to the accelerated aging phenotype that has been observed in some survivors (39–43). Therefore, our results support enhanced clinical attention to blood pressure in all survivors of childhood cancer.

Despite most SJLIFE participants having received ongoing survivorship care through the St. Jude After Completion of Therapy Clinic, 12.8% of those attending their first SJLIFE assessment were identified with undiagnosed hypertensive blood pressure. This underscores the importance of frequent blood pressure monitoring in this population, as rates of undiagnosed high blood pressure are likely higher in survivors not receiving routine follow-up. Data from NHANES during the same time period showed that 82% of adults with hypertensive blood pressure at examination reported being aware of their hypertension, which compares with 52% in SJLIFE (based on an identical questionnaire item). Although the prevalence of uncontrolled hypertensive blood pressure was lower in SJLIFE (22%) compared with NHANES (48% in 2011–2012), it remains unacceptably high for a population known to be at increased risk of cardiovascular complications. Even after the initial SJLIFE visit, at which hypertensive participants were informed of their increased risks and the need for medical management, 35% of those who returned for a subsequent assessment had uncontrolled high blood pressure. These data underscore the importance of improving awareness of the consequences of hypertension among childhood cancer survivors and their community providers.

Strengths of this analysis included the large sample size, direct clinical assessments, and the ability to examine a range of ages, diagnoses, treatments, and lifestyle factors. We used blood pressures measured during a single 3- to 4-day visit for determination of an individual's blood pressure status at that time, rather than separate measures over the course of multiple office visits (31), potentially resulting in misclassification of blood pressure status. Notably, NHANES estimates of U.S. hypertension rates are based on the same methodology. Blood pressure was measured in a relaxed nonoffice setting, potentially reducing the impact of a “white coat effect,” but measured blood pressures may not be representative of usual blood pressure in some participants (44, 45). Because our results are based on a selected population with frequent clinical follow-up, our estimates of undiagnosed and uncontrolled hypertension may be underestimates with respect to the larger population of childhood cancer survivors. Finally, our results could differ slightly using a different classification system to categorize blood pressure status, and existing guidelines do not address childhood cancer survivors (46).

Optimal blood pressure targets for childhood cancer survivors exposed to cardiotoxic therapies have not been established. Recently, the SPRINT trial demonstrated that among hypertensive adults ≥50 years of age with elevated cardiovascular risk, intensive treatment targeting a SBP <120 mm Hg reduced cardiovascular outcomes and all-cause mortality compared with standard treatment with an SBP target <140 mm Hg (47). It is reasonable to suspect that more aggressive targets could benefit childhood cancer survivors as well. Adoption of lower blood pressure targets would further increase the proportion of childhood cancer survivors potentially requiring intervention. Given that hypertension can potentiate the already increased risk of cardiovascular disease associated with cancer treatment, our study demonstrates the need for effective interventions targeting both prevention and control of hypertension in survivors exposed to cardiotoxic therapies.

No potential conflicts of interest were disclosed.

Conception and design: T.M. Gibson, G.T. Armstrong, D.A. Mulrooney, D. Srivastava, K.K. Ness, M.M. Hudson, L.L. Robison

Development of methodology: K.K. Ness, M.M. Hudson, L.L. Robison

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.M. Green, D.A. Mulrooney, N. Bhakta, K.K. Ness, M.M. Hudson, L.L. Robison

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.M. Gibson, Z. Li, G.T. Armstrong, D. Srivastava, K.K. Ness, M.M. Hudson, L.L. Robison

Writing, review, and/or revision of the manuscript: T.M. Gibson, D.M. Green, G.T. Armstrong, D.A. Mulrooney, D. Srivastava, N. Bhakta, K.K. Ness, M.M. Hudson, L.L. Robison

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.L. Robison

Study supervision: M.M. Hudson

This project was funded by the NCI (U01 CA195547, P30 CA021765) and the American Lebanese Syrian Associated Charities (ALSAC).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.