Prevalence of Low Testosterone in Men with Cystic Fibrosis and CBAVD: a cross-sectional study using a large, multi-institutional database

Abstract

Objective:

To investigate the prevalence and treatment rates of low testosterone (T) in men with Cystic Fibrosis (CF). CF is a genetic disease with highly variable presentation that results from a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Phenotypic manifestations of CF include alterations in function of the lungs, liver, pancreas, and reproductive system. Despite the well-described association between CF and infertility secondary to congenital bilateral absence of the vas deferens (CBAVD), men with CF report further sexual and reproductive health concerns, many of which are often associated with low testosterone.

Methods:

We queried the TrinetX database for men over 18 years old with CF or CBAVD to assess what percentage of men had a T level measured, and if hypogonadal (below 300 ng/dL), what percentage received T therapy (TT). We hypothesized that low T would be under-evaluated in the CF population.

Results:

Serum T levels were measured in 10.1% of men with CF and 8.9% of men with CBAVD. Within each group, 464 men with CF (32.7%) and 132 with CBAVD (43.0%) demonstrated low T. The majority of men with T<300 ng/dL went on to appropriately receive TT: 59.3% of men with CF and 78% with CBAVD.

Conclusion:

Our data suggests that hypogonadism is highly prevalent in men with CF and CBAVD. Investigation and appropriate treatment of testosterone deficiency may significantly improve quality of life.

Introduction

Cystic Fibrosis (CF) is an inherited disorder caused by a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) gene located on chromosome 7.¹ CFTR is vital for normal function of multiple organ systems, including respiratory, gastrointestinal, and reproductive. Consequently, CF manifests with a heterogenous clinical phenotype that usually includes pulmonary disease, pancreatitis, and infertility.¹ While the most common cause of infertility in this population is obstructive azoospermia (OA) secondary to congenital bilateral absence of the vas deferens (CBAVD), men with CF present with numerous sexual and reproductive health concerns.²

Recent literature from a CF task force comprised of experts from CF centers across the county proposes that there is a lack of standardization of care provided to patients and gaps in knowledge of providers surrounding the sexual and reproductive health of these patients.^3–5 Importantly, sexual and reproductive health concerns are of high importance to men with CF, can have significant impacts on the mental health of patients and their families, and have become increasingly important as the life expectancy of CF patients increases.⁶ One important sexual and reproductive health concern in this population is low testosterone (T), estimated to be reported in 25–88% of men with CF compared to 2.1–12.8% of the general population.^7,8 Low T is of special interest in this population due to the important role of T in development and maintenance of muscle mass and bone health which are related to poorer long-term prognosis in patients with chronic diseases including CF.^9,10

In this study, we investigated the prevalence of low T in adult men with CF as well as those with CBAVD independent of the CF genotype. While AUA guidelines recommend that all men with infertility are at risk for TD¹¹, based on anecdotal evidence from CF patients in our high-volume male infertility clinic, we hypothesized that men with CF would be under evaluated and thus undertreated for low T.

Methods

Data Source and Study Design

The data used in this study was collected and analyzed in May 2023 from the TriNetX, LLC Research Network, which provided access to electronic medical records and insurance claims for approximately 122 million patients from 83 healthcare organizations. Data from TriNetX includes included information on demographics, diagnoses, procedures, prescriptions, and laboratory values. Diagnoses were recorded using International Classification of Disease (ICD) codes, procedures were recorded using Current Procedural Terminology (CPT) codes, and laboratory tests were identified via Logical Observation Identifiers Names and Codes (LOINC). Information on medications was obtained from prescriptions, orders, inpatient medication reconciliations, and charted medications and were identified in the database using Veterans Affairs (VA) Drug classification system. The data used in the study covered the period from January 2010 through May 2023.

The process by which the data was de-identified is attested to through a formal determination by a qualified expert as defined in Section §164.514(b)(1) of the HIPAA Privacy Rule. Because studies using TriNetX de-identified patient records do not involve the collection, use, or transmittal of individually identifiable data, the qualified expert has determined these studies are exempted from need of Institutional Review Board. Any patient counts less than 10 are obfuscated to ensure patient anonymity, and only aggregate patient counts and statistical summaries are provided.

Cohorts

To identify diagnoses and treatment patterns of testicular dysfunction and testosterone therapy (TT) in men with CF and CBAVD, we constructed a cohort of men 18 years and older who had not previously been treated with TT with a diagnosis of CF (ICD-10-E84). A second cohort of men 18 years and older was also created who had not previously been treated with TT with a diagnosis of CBAVD (ICD-10-Q55.4 or Q55.3).

Demographic and comorbidity data including type 2 diabetes mellitus (ICD-10-E11), lung transplant status (ICD-10-Z94.2), hypertension (ICD-10-I10), osteoporosis (ICD-10-M81), pulmonary embolism (ICD-10-I26), other vascular thrombosis (ICD-10-I74) were collected for all patients. Within each initial cohort, we identified a cohort of patients with a documented serum total and free T level. We then defined a cohort of those patients who had at least one T level below 300 ng/dL. While a true diagnosis of low T requires two morning measurements of T, in this manuscript, due to inherent limitations of the TriNetX database, we defined a hormonal cohort of patients with low T as those with a single hormonal test and were unable to associate a multiple low T values or hypogonadal symptoms with these patients.

The primary outcome was prescription of any of the following forms of TT. These patients were identified by searching the initial cohorts for the medication codes associated with the following forms of TT: testosterone, testosterone 17-phenylpropionate, testosterone enanthate, testosterone cypionate, testosterone undecanoate. Secondarily, we investigated the prescription of other hormonal therapies including human chorionic gonadotropin, clomiphene citrate, and enclomiphene citrate.

Results

There were 14,033 men over age 18 years with a diagnosis of CF in the TriNetX database. The majority of men were white (71%) followed by men of unknown race (20%), African American (8%), and Asian (0%). Mean age was 44 +/− 20 years. Analysis of comorbidities is presented in Table 1. Of the initial CF population, 10.1% (n=1,417) had a documented serum T level. While mean serum T level was 337 +/− 124.5 ng/dL, 32.7% (n=464) were hypogonadal (based on a single lab measurement defined as a total T <300 ng/dL) and of which 59.3% (n=275) received TT (Figure 1). In the CF group an additional 23 received clomiphene citrate and 5 received human chorionic gonadotropin (hCG).

Table 1.

Demographics and comorbidites present in adult CF patients

Characteristics	Total (N = 14,033)
Age (median +/− std dev)	44 +/− 20 years
Race
White	9,963 (71%)
Black	1,123 (8%)
Asian	140 (1%)
American Indian or Pacific Islander	0 (0%)
Unknown	2,807 (20%)
Comorbidities
Type 2 Diabetes Mellitus	4,491 (32%)
Hypertension	4,911 (35%)
Osteoporosis	1,543 (11%)
Pulmonary embolism	1,123 (8%)
Other vascular thrombosis	561 (4%)
Lung transplant	1,123 (8%)
Total Testosterone (median +/− std dev)	337 +/− 124.5

Figure 1.

Adult men with CF who had a documented serum T level and T prescription.

There were 3,439 total men over age 18 with a diagnosis of CBAVD in the TriNetX database. The majority of men were white (69%) followed by men of unknown race (20%), Black (9%), and Asian (2%). Mean age was 40 +/− 19 years. Comorbidities are presented in Table 2. Of the CBAVD population, 8.9% (n=307) had a documented serum T level. Mean serum total T level in this group was 323 +/− 142 ng/dL. Over a third of men undergoing T testing (43%, n=132) were found to have a low total T, of which the majority (78%, n=103) received TT (Figure 2) with an additional 27 men receiving clomiphene citrate and 0 men receiving hCG.

Table 2.

Demographics and comorbidites present in adult men with CBAVD

Characteristics	Total (N = 14,033)
Age (median +/− std dev)	40 +/− 19 years
Race
White	2,373 (69%)
Black	308 (9%)
Asian	69 (2%)
American Indian or Pacific Islander	0 (0%)
Unknown	689 (20%)
Comorbidities
Type 2 Diabetes Mellitus	378 (11%)
Hypertension	791 (23%)
Osteoporosis	69 (2%)
Pulmonary embolism	69 (2%)
Other vascular thrombosis	34 (1%)
Total Testosterone (median +/− std dev)	492 +/− 182

Figure 2.

Adult men with CBAVD who had a documented serum T level and T prescription.

Discussion:

Low T can pose a significant burden on the quality of life including poor sexual function, alterations in mood and cognition, and elevated cardiovascular and metabolic disease risk.¹² We aimed to determine diagnosis and treatment rates of low T in men with CF and CBAVD using a large, multi-institutional, insurance claims database. Our work demonstrated a lower than anticipated frequency of documented serum T level in these populations. In both cohorts, we expected the rates of serum T measurements to be performed more often than our data showed (8.9% in men with CBAVD versus 10.1% in men with CF) since both populations of patients are frequently seen by a urologist for an infertility evaluation, which typically includes hormonal evaluation.^13–15

Our work also demonstrates a higher than anticipated prevalence of low T showing one third of men with CF and nearly half of men with CBAVD demonstrated low T which may be higher than an age-matched, healthy cohort would demonstrate. While reports of low T in the general population range from 10–40%, it remains difficult to estimate the prevalence in young, healthy men due to the well-known increase in hypogonadism with age and comorbidites.^16–19 Patients with CF and CBAVD are theorized to have an increased risk of low T for a multitude of reasons. Recurrent infections and chronic glucocorticoid use in men with CF can contribute to dysregulation of the hypothalamic-pituitary-gonadal axis, resulting in T deficiency.^20,21 Diminished CFTR expression in GnRH secreting cells is also thought to be contributory.²² Our data suggest that there is a high prevalence of low T in these patients. While there are no guidelines on T screening or management of low T in this patient population, CF patients with stable disease have been shown to have lower serum T levels compared to healthy controls.⁷ To date, there is a paucity literature examining the true prevalence of low T in this patient population. One study examined 40 adult, CF patients and noted that 28% of patients had T levels two standard deviations lower than age matched controls.² Meanwhile, another study demonstrated that 45% of their CF cohort had T levels below reference range.²³

Advances in the multidisciplinary treatment have led to men with CF living longer, healthier lives. Thus, they are increasingly facing greater sexual and reproductive health concerns.²⁴ Hypogonadism is a known risk factor for sexual dysfunction and infertility, thus CF providers should be aware of this data and consider earlier T testing in this population. Of note, while we advocate for testing and evaluation of the hormonal status of male CF patients, before starting treatment with TT providers must discuss the long-term impacts of exogenous T on spermatogenesis in accordance with AUA guidelines.¹¹ Even men who have CBAVD (and resultant OA) may produce sperm and desire current or future fertility, therefore shared decision making between the patient and physician must include the potential long-term risk of nonobstructive azoospermia (NOA) as a result of TT.²⁵ For some patients, therapy with selective estrogen receptor modulators (SERM) may be a viable alternative for alleviation of hypogonadal symptoms and preservation of spermatogenesis.²⁶ Moreover, these findings represent an opportunity for quality improvement in the care of men with CF as the presence of low T is a risk factor for metabolic derangements such as muscle wasting, poor growth kinetics and decreased bone mineral density which are especially important considerations for long-term CF survivors.²⁷ While there is a paucity of data examining the effect of modern CF treatments on T level, it has been hypothesized that low T could lead to pathologic fractures, deconditioning, and compromised respiratory function in this vulnerable patient population.²²

Recent work has identified patient demonstrated need for early counseling and intervention to optimize male sexual and reproductive health especially in those with chronic medical conditions.^24,28 For instance, past work has demonstrated that the minority of CF patients receive sexual health checkups and that many patients lack understanding about the effect of CF on fertility and sexual function.^28,29 Given the potential harm of T deficiency in this vulnerable patient population, future work implementing T screening in CF patients could help improve patient’s hormonal and metabolic profile. This could be adopted as a recommendation from The Cystic Fibrosis Foundation to accompany their existing recommendation of obtaining routine dual-energy X-ray absorptiometry testing (DEXA) starting at age 8.³⁰

Our study is not without limitations most of which are inherent to the TriNetX database. While this is a large, comprehensive database, it does not allow us to access information about patients who received care or medications at an institution that does not report to TriNetX. If these individuals received clinical workup and/or TT prescriptions from an organization that does not report to the database or compounding pharmacies, we may underestimate the true rate of treatment with TT in this group. Additionally, the database does not allow us to gather information regarding an individual patient’s overall health status or symptoms; therefore we cannot investigate the clinical diagnosis of T deficiency which is defined by the AUA as low T with symptoms.¹¹ For example, if an individual had a medical condition that could explain their low T (or even older age), by not querying the database for that specific condition, we do not obtain the relevant data. Additionally, we are unable to obtain information regarding individuals who were trying to conceive or interested in fertility who may not be prescribed TT due to well-described relationship between T and azoospermia. While we attempted to capture these patients through querying these populations for prescription of human chorionic gonadotropin and/or clomiphene citrate (often used as a treatment for hypogonadal symptoms in those desiring fertility^14,15,31), the number of patients who received these treatments is low, and the reasoning behind these alternative prescriptions remains unknown. Alternatively, if patients are chronically ill and undergoing lung transplant evaluation or requiring frequent hospitalizations, they may be unable to undergo hypogonadism evaluation and treatment. Finally, the TriNetX database does not provide information about patient symptoms, therefore even men who demonstrate low T may not be provided treatment if asymptomatic. Despite these limitations, our study represents the first of its kind to objectively describe the rates of low T and subsequent TT prescriptions in these populations. Importantly, our data suggest that we are likely under-evaluating and may be undertreating these populations. Continued work into the role of T in men with CF is necessary.

Conclusion

This study represents the first and largest epidemiological study of hypogonadism in men with CF and CBAVD, and we found that men with a diagnosis of CF and CBAVD are under evaluated and may be undertreated for low T. Our data showed that, while men with CF and CBAVD appear to be provided treatment once identified as having low T, the majority of men in both groups do not undergo hormonal evaluation, despite the risk of infertility. Given that CF mutations are present in varied tissues, the effects are seen throughout the body which may include hormonal impacts. Further work in the hormone status of these populations is necessary to corroborate our findings and elucidate the potential pathophysiologic mechanisms by which CF mutations lead to low T production in the testes.