Evaluation of breast calcif ication, calcif ication characteristics, and BI-RADS categories in patients with primary hyperparathyroidism.

INTRODUCTION
Primary hyperparathyroidism (PHPT) is a disease characterized by the autonomous
secretion of parathyroid hormone (PTH) from the parathyroid glands due to adenoma,
hyperplasia, or carcinoma, despite normal or elevated serum calcium levels. The
clinical manifestations of this disease primarily include
nephrolithiasis-nephrocalcinosis, and osteoporosis. Additionally, disrupted calcium
metabolism in PHPT (1) can lead to
ectopic calcification in several organs, such as the skin (2), gallbladder (3), sclerochoroid (4-7), eyes (8),
heart valves (9,10), myocardium (11), abdominal aorta (12), brain (13-15), cerebral gyrus
(16), joint (11,17), lungs (18-20), pancreas
(21), and stomach
(22). The precise
mechanism of this type of calcification is not fully understood; however, potential
causes include increased levels of the calcium-phosphate product and the local
secretion of free hydrogen ions, which result in an alkaline environment (18,22).
Breast calcif ication, commonly observed in mammography, can be present in both
breast cancer and benign lesions. The morphology and distribution of calcifications
are instrumental in evaluating whether breast lesions are benign or suspicious.
Although studies (23-26), case series (27), and case reports (28-31) have documented breast calcification in patients with
secondary hyperparathyroidism (SHPT), to our knowledge, there is no study examining
its prevalence and characteristics in patients with PHPT. The incidence of breast
cancer has been reported to be higher in patients with PHPT compared to those
without PHPT. Hypothetically, this finding could be attributed to increased calcium
deposition in the breasts and more frequent breast examinations in these patients
(32). Given the above, we
aimed to determine the frequency and characteristics of breast calcification,
analyze the distribution of Breast Imaging-Reporting and Data System (BI-RADS)
scores, and explore the relationship between calcification and biochemical or
clinical features in patients with PHPT.

SUBJECTS AND METHODS
Female patients aged ≥ 40 years old with PHPT and age- and body mass index
(BMI)-matched healthy women were recruited. The healthy participants were selected
from women referred to the radiology clinic for routine mammographic screening by
physicians in family medicine, general surgery, and obstetrics and gynecology. These
participants were matched for age, BMI, smoking status, menopausal status, and
breast disease history. According to the American Cancer Society’s 2015 guideline
update, women aged ≥ 40 years are recommended to undergo mammography
(33).
PHPT was diagnosed based on either the presence of hypercalcemia along with elevated
or inappropriately normal serum PTH levels, indicating overt PHPT, or normal serum
calcium and 25-hydroxyvitamin (OH) D levels with elevated serum PTH and increased
24-hour urinary calcium/fractional excretion of calcium levels, indicating
normocalcemic PHPT (9). Exclusion
criteria included a PHPT duration of < 6 months, chronic kidney or liver disease,
malignancy, pregnancy, lactation, history of breast surgery, history of chest
radiotherapy, and use of medications affecting blood calcium levels (e.g., calcium
supplements, thiazide diuretics, lithium, vitamin A, teriparatide, bisphosphonates,
corticosteroids, rifampin, phenobarbital, phenytoin, calcitonin, chloroquine,
plicamycin, etc.). Smoking habits, alcohol consumption, breast disease history, and
family history of breast cancer were also recorded. Weight (kilograms) was divided
by the square of height (meters) to determine the BMI.
Serum levels of calcium (normal: 8.5–10.4 mg/dL), albumin (normal: 3.2–4.8 g/dL),
phosphate (normal: 2.4–5.1 mg/dL), magnesium (normal: 1.3–2.7 mg/dL), intact PTH
(normal: 18.4–80.4 ng/L), 25 (OH) D (normal: 30–150 ng/mL), and alkaline phosphatase
(normal: 42–98 U/L) were measured in the morning after an 8-hour fast and recorded.
The 24-hour urinary calcium and fractional excretion of calcium were similarly
documented. Corrected serum calcium was calculated using the formula: Corrected
calcium (mg/dL) = [0.8 × (4 - serum albumin (g/dL))] + measured calcium
(mg/dL). The highest recorded calcium and PTH levels were noted from the Ankara
Bilkent City Hospital database or the national patient health database. Calcium
× phosphate product levels and the highest such levels in the databases were
also recorded. Hypercalcemia was classified as mild (10.5–11.9 mg/dL), moderate
(12–13.9 mg/dL), or severe (≥ 14 mg/dL). PHPT duration was defined as the
time between the first diagnosis and the date of mammography. The earliest available
high serum calcium value in the hospital or national database was recorded as the
time of initial diagnosis.
All participants underwent mammography using the Senographe Pristina Mammography
system (GE Healthcare, Fairfield, Buc, France). Mammographic features were analyzed
and reported according to the American College of Radiology’s BI-RADS by our
radiologist (A.O.), who was blinded to the patients’ diagnoses. Calcification
morphology was classified as benign or suspicious. Benign calcifications included
skin, vascular, popcorn-like, large rod-like, round, rim, dystrophic,
milk-of-calcium, and suture calcifications. Suspicious calcifications were described
as amorphous, coarse heterogeneous, fine pleomorphic, or fine linear/linear
branching. Additionally, the distribution of calcifications was classified as benign
if diffuse, and as suspicious if linear or segmental (34).
Calcifications were recorded as unilateral or bilateral, indicating presence in one
breast or both, respectively. Breast density was classified as A, B, C, or D,
reflecting breast composition from almost entirely fatty to extremely dense,
according to BI-RADS. Scores ranged from 0 to 6, where 0 required further imaging, 1
was negative, 2 was benign, 3 was probably benign, 4 was suspicious (with
subcategories 4A, 4B, and 4C for low, intermediate, and high suspicion of
malignancy, respectively), 5 was highly suggestive of malignancy, and 6 was a
biopsy-proven malignancy (34).
Participants were asked to return for further diagnostic evaluation when findings
such as suspicious breast calcifications, asymmetric densities, or a mass were
identified. If malignancy was suspected, a biopsy followed by surgery, if necessary,
was suggested.
Dual energy X-ray absorptiometry was used to measure bone mineral density (BMD) in
the femur, radius, and lumbar spine (L1–L4) with an anterior-posterior projection in
both the patient and control groups (Lunar iDXA, GE Healthcare, USA). Postmenopausal
subjects with a BMD T score of ≤ -2.5 were classified as having osteoporosis,
those with a T score between -2.5 and -1 were classified as having osteopenia, and
those with a T score ≥ -1 were categorized as normal. In premenopausal women,
a BMD Z score of ≤ -2 indicated a bone density below the expected range for
age, and > -2 was within the expected range. Premenopausal subjects with scores
below the expected range were included in the osteoporosis group, whereas those
within the expected range were considered to have normal BMD. Nephrolithiasis was
defined as either a history of the condition or diagnosis by imaging modalities. If
any, histopathological findings of parathyroidectomy with/without thyroidectomy as
well as breast cytopathology were noted.
Patients and controls were compared with regard to demographic and laboratory data,
as well as mammographic results, including calcification presence and types and
BI-RADS scores, were compared between patients and controls. PHPT patients were
further divided into groups based on disease duration of less than or more than 48
months and analyzed with respect to the same parameters. Additionally, patients with
and without breast calcifications were compared.
This study was approved by the institution’s ethical committee (Date: January 18,
2021, no: 26379996/21, Chairman: Kara H.), in compliance with the Declaration of
Helsinki. Informed consent was obtained from all patients and controls.
SPSS 24 package (IBM Corp., Armonk, NY, USA) was used for the statistical analysis.
Descriptive statistics were given as mean ± standard deviation when normally
distributed, and median (interquartile range 25%-75%) when not normally distributed.
Nominal variables were given as percentage and number of cases. Chi-squared or
Fisher exact test was used when categorical variables were compared. Student
t test was used for parametric variables and Kruskal-Wallis
test was used for nonparametric variables. Spearman’s or Pearson’s tests were used
to assess correlation of breast calcification and laboratory or clinical variables.
A p value < 0.05 was considered significant.
The sample size was determined for a comparison between two independent groups with
PHPT and healthy controls. The calculation assumed a significance level (α)
of 0.05, a statistical power (1–β) of 0.80, and an expected effect size
(Cohen’s d) of 0.35, defined as the smallest clinically meaningful difference
anticipated in the study. Under the assumptions of equal group sizes, approximate
normal distribution, and homogeneous variances, the required sample size was 102
participants per group (204 in total). The Student’s t-test will be
used for the primary comparison, while the Mann–Whitney U test will be applied if
distributional assumptions are not met. All calculations were performed using
G*Power version 3.1.

RESULTS
Overall, 104 patients and 107 controls were included. Demographic and clinical
findings, laboratory results, and imaging results are summarized in Tables 1–3. Age distribution and BMI were similar between both groups
(p = 0.920 and p = 0.391, respectively). The
median duration of PHPT was 48 months (min-max: 6–180 months). The frequency of
current smoking was comparable between the patient and control groups. Regular
alcohol consumption was observed in 3 patients and 1 control. The rates of previous
benign breast disease and family breast cancer history were similar in both groups
(p = 0.416 and p = 0.367, respectively).
Menopausal status was also comparable (p = 0.066).
Serum levels of calcium, corrected calcium, highest calcium, magnesium, alkaline
phosphatase, and PTH were higher, while phosphate, albumin, calcium x phosphate, and
highest calcium x phosphate levels were lower in the PHPT group than the control
group (p < 0.001 for all). Similarly, the 25 (OH) vitamin D
level was lower in the patient group (p = 0.001). The majority of
PHPT patients (63.5%) had mild hypercalcemia, 9.6% had moderate hypercalcemia, and
the remaining 26.9% had normocalcemic PHPT. The rates of nephrolithiasis and
osteoporosis were higher in the PHPT group than in the control group
(p < 0.001 for both).
The BI-RADS score did not differ, with BI-RADS 0 being the most prevalent score in
both groups. Calcifications were observed in 63 (60.6%) patients with PHPT and 72
(67.3%) individuals in the control group. In the PHPT group, 1 patient had
suspicious calcifications, 1 patient had both benign and suspicious calcifications
(Figure 1), and the rest
had benign calcifications on mammography. All breast calcifications were benign in
the control group. Calcifications were mostly bilateral in both groups
(p = 0.392).
Parathyroidectomy was performed on 63 patients (60.6%). Except for one patient with
atypical parathyroid adenoma, all patients had histopathologically confirmed
parathyroid adenoma. Concomitant thyroidectomy was performed on 19 (18.3%) patients,
revealing papillary thyroid carcinoma in 12 patients, noninvasive follicular thyroid
neoplasm with papillary-like nuclear features (NIFT-P) in 1 patient, concurrent
papillary thyroid carcinoma and NIFT-P in 2 patients, and benign findings in 4
patients.
PHPT patients were categorized based on the presence (n = 63, 60.6%) or absence (n =
41, 39.4%) of breast calcification. PTH and urinary calcium levels were lower, and
age was higher in PHPT patients with breast calcification than those without
calcification (p = 0.026, p = 0.048, and
p = 0.026, respectively). Other clinical and laboratory
variables were similar in both groups. A weak correlation was found between breast
calcification and age (r = 0.219, p = 0.025), PTH (r = -0.220,
p = 0.025), and 24-hour urinary calcium levels (r = -0.195,
p = 0.048) in PHPT patients. There was no cut-off for disease
duration that could predict the presence of calcification. The area under the ROC
curve was 0.534 (95% CI: 0.456–0.612, p = 0.400), indicating no
significance.
PHPT patients were divided based on a disease duration of less than or greater than
48 months, the median duration. Both groups were similar regarding the frequency and
localization of calcification, BI-RADS score, breast density, history of benign
breast disease, and family breast cancer history (p > 0.05 for
all). Similarly, age, BMI, current smoking status, menopausal status, and BMD,=,
along with laboratory parameters except for the highest calcium level, were
comparable between the two groups (p > 0.05 for all). The
highest calcium level and nephrolithiasis frequency were more prominent in the
longer duration group than the shorter duration one (p < 0.001
and p = 0.008, respectively). All patients with calcium levels of
> 12 mg/dL were in the longer duration group. One patient with suspicious
calcification was in the shorter duration group, and one with both suspicious and
benign calcification was in the longer duration group; others had benign
calcifications.
Biopsy referral rates were 1.9% (2/104) in PHPT patients and 3.7% (4/107) in the
control group. Two PHPT patients with suspicious calcifications underwent excisional
breast biopsy, resulting in one invasive breast cancer and one atypical ductal
hyperplasia diagnosis. Three controls underwent tru-cut biopsy, and one underwent
excisional biopsy, with all 4 controls having benign cytology.

DISCUSSION
The rate and type of breast calcification, as well as the distribution of BI-RADS
scores, were similar in female patients with and without PHPT. Patients with disease
durations both shorter and longer than 48 months also exhibited similar mammographic
findings. These results may suggest that the presence of calcification is not
associated with an increased incidence of breast cancer in PHPT patients. In the
patient group, breast calcification was positively correlated with age and
negatively correlated with PTH and 24-hour urine calcium levels.
Several studies have explored the relationship between PHPT and breast cancer. Both
conditions exhibit similar sex and age associations (35) and have shown a steadily increased incidence
over the past 40 years (32).
Breast cancer is the most frequent cancer in women and is more common in women than
in men, similarly to PHPT (36,37). The frequency of the association
between breast cancer and PHPT was found to be similar to that of the general
population in one study (38).
Conversely, in a large-scale cohort study comparing the incidence of subsequent
malignancy in patients who underwent surgery for PHPT with those who did not, PHPT
was associated with an increased risk of developing cancer (e.g., kidney, colon, and
squamous cell skin cancer), particularly breast cancer. This risk remains consistent
over time and does not diminish even 15 years post-operation. Therefore, the
increased risk of malignancy is not due to the biochemical changes in PHPT, but
likely results from shared genetic and environmental etiological factors (39). These common factors are
considered to be ionizing radiation and hypercalcemia. Another reported risk factor
associated with a higher incidence of both diseases is being overweight (40).
Although there is no literature examining hypercalcemia-specific breast tissue, it is
hypothesized that calcium accumulation in the breast may be more pronounced in PHPT,
leading to an increased diagnosis of breast cancer due to more frequent
breast-focused examinations (32).
To determine whether calcium accumulation in breast tissue increases in PHPT, we
designed this study. We observed similar frequencies and morphologies of breast
calcification, as well as BI-RADS scores in PHPT patients and controls. Therefore,
we conclude that the higher prevalence of breast cancer among PHPT patients, as
suggested in the literature, may not be attributed to calcification presence. On the
other hand, we identified one malignant and one atypical ductal hyperplasia case,
both in the PHPT group, while all four cases in the control group were benign.
Although there is no statistically significant difference between patient and
control groups, we believe that PHPT patients should be carefully examined for
breast cancer as both malignant and atypical cytological findings were present in
this group.
Examining the inverse relationship, the frequency of PHPT in breast cancer patients
was higher than its incidence in the healthy female population (35,41). In non-aggressive breast cancer patients without evidence
of PHPT, significantly higher calcium and PTH levels were noted, regardless of
clinical stage and antitumor treatment, compared to the healthy controls (41). The prognosis of breast cancer
patients, irrespective of previous PHPT operations, was found to be similar
(40).
Although no study exists in the literature examining breast tissue calcification in
PHPT patients, there are three studies concerning patients with SHPT due to chronic
kidney disease or end-stage renal disease (23-25). An increase of
up to 94% in breast calcification rates has been noted in patients with SHPT
associated with chronic renal failure, specifically among those on long-term
hemodialysis (25). The increased
calcification in SHPT compared to the general population is thought to relate to
decreased phosphate excretion and rising levels of serum calcium-phosphate product,
though it likely results from multiple factors, including the duration of renal
failure, dialysate calcium concentration, tissue pH, calcium, magnesium and fluoride
levels, metabolic acidosis, hyperoxalemia, hyperlipidemia, and corticosteroid
treatment. Calcifications occurred in both breasts in 98% of SHPT patients
(23).
Correlation between calcification and either serum PTH levels or the
calcium-phosphate product SHPT has been inconsistent across studies. Breast
calcifications associated with SHPT are benign, not linked with breast masses, and
should not be confused with suspicious calcifications indicative of breast cancer,
according to earlier studies (23,25). These studies, however, did not
include data on BI-RADS, callback rates for additional diagnostic workups, biopsy
recommendation rates, clinical outcomes related to increasing calcifications, or the
incidence of suspicious calcifications. A more recent study by Castellanos and cols.
(24) found that while
increased breast calcification is generally benign, the incidence of
malignancy-associated calcifications is slightly higher in SHPT patients compared to
a control group. Consequently, these patients have a higher likelihood of being
referred for a biopsy if recalled for further diagnostic evaluation (24). None of the three studies
reported breast cytology results, leaving a gap in the literature on whether women
with SHPT have a higher incidence of breast cancer.
Contrary to these findings in SHPT patients, our study did not detect an increased
rate of breast calcification in PHPT patients. Similar to findings in SHPT studies,
most calcifications in PHPT were benign and bilateral. Only two participants, both
with PHPT, had suspicious calcifications. One potential explanation for the
differing calcification rates between PHPT and SHPT could be the comparatively lower
phosphate and calcium-phosphate product levels in PHPT patients, as opposed to SHPT
patients. Despite a reduction in phosphate levels, the calcium-phosphate product may
still be high in PHPT due to elevated calcium levels from increased bone resorption,
particularly in severe hypercalcemia. The role of the calcium-phosphate product in
PHPT remains unclear, despite its widespread application in chronic renal disease
management. Few studies have explored its level in PHPT patients, with some linking
it to renal dysfunction but not to aortic valve calcification (9,42). Lower product level may be related to the reduced phosphate
levels and the modestly elevated or even normal calcium levels observed in most of
our patients. Another possibility regarding the differences in PHPT and SHPT
calcification rates is that the microenvironment of breast tissue in PHPT patients
differs from systemic levels in a manner that promotes calcification.
In our study, we found a negative correlation between calcification in PHPT patients
and PTH levels, in contrast to the study of Sommer and cols. (23) involving patients with renal
hyperparathyroidism, which showed a positive correlation. We additionally observed
positive correlations of breast calcification with age and a negative correlation
with urinary calcium levels. The frequency of breast calcification is known to
increase with age, and microcalcification is more common in older patients
(43). However, our
evaluation of PHPT patients concerning breast calcification and its related factors
is the first in the literature. Further research is needed to elucidate whether
renal dysfunction, more severe hypercalcemia/hypophosphatemia, higher PTH levels, or
other factors might affect calcification rates in PHPT patients.
Several studies have compared laboratory and demographic characteristics in PHPT
patients with and without calcifications in various tissues or organs. The severity
of abdominal aortic calcification in PHPT and breast calcification frequencies in
SHPT correlate with PTH levels (12,23). Patients with mild PHPT have
shown increased calcification area in the aortic valve, positively and independently
associated with PTH levels (9).
Pepe and cols. (12) reported that
older age, years since menopause, and time since diagnosis as distinguishing factors
between PHPT patients with and without abdominal aortic calcification. In another
study, age was the sole predictor of gallstone development in PHPT patients
(3). Normocalcemic PHPT
patients with renal calcifications had increased serum PTH, 1.25(OH)2
vitamin D, and urinary calcium compared to those without calcifications (44). Contrary to this study, Perez
and cols. (45) reported that PHPT
patients with nephrolithiasis had similar serum calcium and PTH levels compared to
patients without nephrolithiasis, although urinary calcium levels were increased.
Ejlsmark-Svensson (46) found that
renal calcification correlated with the severity of PHPT as measured by PTH levels,
24-hour urine calcium, and the degree of hypercalcemia (46).
It is suggested that increased serum calcium levels due to the oversecretion of PTH
may lead to calcium deposits in breast ducts (47). In our study, we did not find an increased rate of
calcification in PHPT patients compared to the control group. Nevertheless, PHPT
patients with breast calcification had lower PTH and urinary calcium levels but were
older compared to those without calcification. Additionally, as noted earlier,
calcification in PHPT patients correlated positively with age and negatively with
PTH levels and 24-hour urinary calcium levels. Differences in the physiological
characteristics of various tissues, the variable effects of PTH on these organs or
tissues, the selection of different patient groups (e.g., normocalcemic PHPT, mild
PHPT, asymptomatic PHPT), and diagnostic criteria (e.g., imaging methods, serum
calcium range) may explain the diversity of results in the literature (35).
The duration of PHPT can affect the course of some clinical and laboratory
parameters. PHPT patients managed nonoperatively showed an increased annual rate of
renal stone events in the year before and after diagnosis, with the rate decreasing
after five years (48). Some
studies have shown that a proportion of patients with normocalcemic PHPT progressed
to hypercalcemia over time, while other studies did not find this progression
(49). In our study,
nephrolithiasis was more frequent, and the highest calcium levels were more
prominent in the group with longer disease duration compared to the group with
shorter duration.
Despite our promising findings, this study has some limitations. Due to the
increasing access to laboratory and imaging modalities, patients are diagnosed at
earlier stages in recent years. Consequently, most of the patients in the study had
mild hypercalcemia. This may have contributed to possibly lower rates of
calcium-phosphate product levels and, therefore, a lower rate of breast
calcification. Additionally, the cross-sectional design might have resulted in a low
number of breast cancer patients. Given the often asymptomatic or subclinical nature
of PHPT, underestimation of the disease duration is possible in the current study.
Consequently, the disease duration may be inaccurately short in some cases, possibly
because the disease started earlier than the first record available in the hospital
database. This may affect some analyses performed based on disease duration,
particularly its effect on the frequency of breast calcification. The lack of data
on ionized calcium measurements, calcification subtypes, and the physical
examination findings of the breast were other limitations of the study.
In conclusion, it can be inferred from this study that breast calcification may not
be considered a target organ involvement, and the presence of calcification is not
related to an increased incidence of breast cancer in PHPT patients. Although not
statistically significant, the fact that all patients with non-benign cytology
results were in the PHPT group may suggest that PHPT patients should be carefully
screened for breast cancer. Nonetheless, these results should be confirmed in
patients with different levels of hypercalcemia, in different age groups, and among
patients with longer follow-up periods and a larger cohort.