Risk of Malignant Neoplasm in Patients with Primary Hyperparathyroidism: A Systematic Review and Meta-analysis

This study aimed to evaluate the prevalence and risk of malignant neoplasm in primary hyperparathyroidism (PHPT) patients. Potentially eligible studies were retrieved from PubMed and Embase databases from inception to November 2023 using search strategy consisting of terms for “Primary hyperparathyroidism” and “Malignant neoplasm”. Eligible study must report prevalence of malignant neoplasm among patients with PHPT or compare the risk of malignant neoplasm between patients with PHPT and comparators. Point estimates with standard errors were extracted from each study and combined using the generic inverse variance method.A total of 11,926 articles were identified. After two rounds of systematic review, 50 studies were included. The meta-analysis revealed that pooled prevalence rates of overall cancer was 0.19 (95%CI: 0.13–0.25; I2 94%). The two most prevalent types of malignancy among patients with PHPT ware papillary thyroid cancer (pooled prevalence: 0.07; 95%CI: 0.06–0.08; I2 85%) and breast cancer (pooled prevalence: 0.05; 95%CI: 0.03–0.07; I2 87%). Subgroup analysis of studies focusing on patients undergoing parathyroidectomy reported a fourfold higher prevalence of papillary thyroid cancer than the remaining studies (0.08 versus 0.02). The meta-analysis of cohort studies found a significant association between PHPT and overall cancer with the pooled risk ratio of 1.28 (95%CI: 1.23–1.33; I2 66.9%).We found that the pooled prevalence of malignant neoplasm in PHPT was 19%, with papillary thyroid cancer and breast cancer being the most prevalent types. The meta-analysis of cohort studies showed that patient with PHPT carried an approximately 28% increased risk of malignancy. Supplementary Information The online version contains supplementary material available at 10.1007/s00223-024-01219-y.


Introduction
Primary hyperparathyroidism (PHPT) is a disorder of mineral metabolism characterized by increased or unsuppressed levels of parathyroid hormone (PTH) due to excessive secretion from one or more abnormal parathyroid glands [1,2].This condition is a relatively common endocrine disorder, with prevalence estimates ranging from 1 to 7 cases per 1,000 adults and an incidence varying from 0.4 to 21.6 cases per 100,000 person-years [3,4].
There are few established risk factors for PHPT, which include female sex, exposure to ionizing radiation, lithium exposure and genetic predisposition [2].Over 95% of PHPT cases are sporadic, with less than 5% associated with germline mutations predisposing individuals to the development of parathyroid tumor [5].Complications of PHPT include hypercalcemia, bone loss and hypercalciuria leading to kidney stones, nephrocalcinosis and kidney failure [1,2].Furthermore, evidence from observational studies suggests an association between PHPT and increased risks of medical comorbidities including cardio-metabolic disease and psychiatric disorders [6].Although the exact causal relationship between PHPT and these conditions are undetermined, it has been thought to be mediated by abnormal calcium homeostasis [7].
Interestingly, studies have suggested a link between PHPT and the risk of developing malignant neoplasm, which is thought to be mediated by different mechanisms including the tumorigenic effects of PTH, abnormal calcium and vitamin D homeostasis, as well as shared genetic and environmental risk factors [8].From an epidemiological standpoint, multiple studies have revealed varying prevalence rates of various types of malignant neoplasm among patients with PHPT, including thyroid, breast, lung, and colon cancer, among .Nevertheless, most of the included studies were performed in a single institution and had small sample sizes, limiting the confidence and generalizability of the findings.Additionally, some studies suggest an increased risk of malignant neoplasm in patients with PHPT compared to those without [45,[56][57][58][59][60][61][62][63][64][65][66].However, the reported degree of association between PHPT and the risk of malignant neoplasm vary across the studies.Furthermore, it remains unknown whether the risk of malignancy is different in patients with PHPT requiring parathyroidectomy (PTX) compared to mild PHPT, or if the preoperative assessment for PTX leads to increased screening and detection of malignancy, particularly thyroid cancer.
Using a systematic review and meta-analysis technique, the objective of this study is to identify all available data on the prevalence of each type of malignant neoplasm in patients with PHPT, including a subgroup of studies in patients undergoing PTX, and combine them together.Additionally, we aim to determine the pooled effect size of the risk malignant neoplasm in patients with PHPT compared with individuals without PHPT.

Search Strategy
Three investigators (C.W., W.P., S.S.) independently conducted searches in the Embase and PubMed databases from inception until November 2023.The search terms were generated from terms related to 'Primary Hyperparathyroidism' and 'Malignant Neoplasm.'The comprehensive search strategy is presented in Supplementary Material 1.There were no restrictions on language.This study adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, as illustrated in Supplementary Material 2.

Eligibility Criteria
For the meta-analysis of the prevalence of malignant neoplasm in patients with PHPT, eligible studies must include a cohort of patients with PHPT and report the proportion of patients with overall or any type of malignant neoplasm.For the meta-analysis of the risk of malignant neoplasm in patients with PHPT versus comparators without PHPT, eligible studies must include a cohort of patients with PHPT and another cohort of comparators without PHPT.The study must then compare the risk of overall or any type of malignant neoplasm between the two groups.Studies that focused exclusively on patients with parathyroid carcinoma, multiple endocrine neoplasia or genetic forms of PHPT were excluded from both meta-analyses.Additionally, case reports and case series with sample sizes of less than 5 were excluded from the meta-analyses.In cases where multiple studies utilized the same database, only the one with a more comprehensive inclusion of data would be included to avoid duplication of data points.
The retrieved articles were independently evaluated for eligibility by two investigators (N.C and C.W.).Disagreements were resolved through discussion with the methodologist (T.R.).The quality of each included cohort study was assessed by two investigators (N.C., T.R.) using the Newcastle-Ottawa quality assessment scale for cohort study [67].

Data Extraction
A standardized collection form was employed for data extraction.The collection form included the following information: last name of the first author, year of publication, country of study, number of PHPT patients, the number/ proportion of each malignant neoplasm, study institution/ data source, whether or not included patients underwent parathyroidectomy, mean age of patients, the percentage of female patients and mean serum PTH and serum calcium.For studies comparing the risk of malignant neoplasm in PHPT patients versus comparators, additional variables were extracted, including number of comparators, diagnosis of PHPT, diagnosis of malignant neoplasm, follow-up criteria, variables adjusted in the multivariate analysis, and effect estimates representing the relative risk of overall and each type of malignant neoplasm.

Statistical Analysis
All data analyses were conducted using the StataMP15.Proportions with standard errors for count data were extracted from each included study.The pooled effect sizes and 95% confidence interval were computed using DerSimonian and Laird's generic inverse variance method [68].A randomeffect model was employed instead of fixed-effect model given the diverse background populations and protocols across the included studies.Statistical heterogeneity was assessed using the I 2 statistics, where I 2 values of 0-25% indicated insignificant heterogeneity, 26-50% low heterogeneity, 51-75% moderate heterogeneity and > 75% high heterogeneity [69].In the meta-analysis of prevalence of malignant neoplasm in patients with PHPT, subgroup analyses were performed to explore if studies focusing in patients undergoing PTX yielded different results from the remaining of the studies.Sensitivity analyses were performed by excluding each of the studies with overlapping data to assess the robustness of the results.

Search Results
A total of 11,926 articles were initially retrieved from the EMBASE and PubMed databases.After removing 2253 duplicated articles, 11,926 articles remained for title and abstract review.At this stage, 11,789 articles were excluded as they did not meet the eligibility criteria based on article type and study design.This exclusion left 137 potentially eligible articles for full-text review.Subsequently, 79 articles were further excluded due to the lack of the outcome of interest, resulting in 58 articles that fulfilled the eligibility criteria.
One study [55] was excluded from the meta-analysis as it reported the prevalence of thyroid cancer in a PHPT patient cohort that was a subset of another study [37].Seven studies utilized the Swedish Cancer Registry database to report the risk of cancer among PHPT patients in comparison with the general Swedish population [56,[60][61][62][63][64][65].Therefore, only one study with the most comprehensive inclusion of data from 1958 to 2008 was included in the meta-analysis [56].Additionally, two studies [59,66] utilized the database of residents of Tayside, Scotland; therefore, the one with the more comprehensive inclusion of data from 1997 to 2019 was included [59].
Finally, a total of 50 articles were included in the metaanalysis.Out of these, 46 were included in the meta-analysis of the prevalence of malignant neoplasm in PHPT , and 5 were included in the meta-analysis comparing the risk of malignant neoplasm in patients with PHPT versus comparators [45,[56][57][58][59]. Figure 1 demonstrates the study identification and literature review process.The characteristics of the eligible studies are shown in Tables 1 and 2.

Prevalence of Malignant Neoplasm in Patients with Primary Hyperparathyroidism
Among the 46 prevalence studies included in the metaanalysis, 37 studies were conducted on PHPT patients who underwent PTX.The mean age ranged from 48 to 65 years and the percentage of female participants ranged from 57-89%.The mean ± SD of serum PTH levels across studies was 234 ± 124 pg/mL (from 20 studies), while the mean ± SD of serum calcium levels was 11.4 ± 0.5 mg/dL   3).List of studies included in the meta-analysis is presented in Supplementary material 3.

Discussion
This is the first systematic review and meta-analysis that compiles all available evidence on malignant neoplasm in patients with PHPT.Our pooled analysis showed that the prevalence of any type of malignant neoplasm among PHPT patients was approximately 20%.Papillary thyroid cancer and breast cancer were the two most common and frequently reported types of malignant neoplasm with the pooled prevalence rates of 7% and 5%.These prevalence rates are considered disproportionately higher than those of the global population, given the global age-standardized incidence rates of thyroid cancer of 2.05 cases per 100,000 population in 2019 and breast cancer of 100.9 cases per 100,000 population in 2020 based on data from the Global Burden of Disease (GBD) database [70,71].
Other less prevalent types of malignant neoplasm included lung cancer, gynecological cancer, kidney cancer, colon cancer, prostate cancer, urinary tract cancer, skin basal cell carcinoma, hematologic malignancy and gastric cancer, with pooled prevalence rates of 0-2%.In addition, the metaanalysis of 5 cohort studies utilizing population databases from European countries revealed that patients with PHPT carried approximately 28% increased risk of malignancy when compared with comparators.
These observations may suggest the link between PHPT and tumor development outside of known tumors associated with multiple endocrine neoplasia.There are a number of possible explanations to these findings.The first perceivable explanation is selection bias as the rate of screening and detection of cancer may be increased as a result of clinical care in patients with PHPT.This is particularly true in the (continued) Fallah et al. [56] Ghosh et al. [57] Ogard et al. [58] Palmieri et al. [45] Soto-Pedre et al. [59] Newcastle-Ottawa score   case of thyroid cancer.As thyroid ultrasound or neck CT scan are parts of routine evaluation in patients with PHPT undergoing PTX [72], those with concurrent asymptomatic thyroid cancer may have been screened and diagnosed through this process.This hypothesis can be supported by the result of our subgroup meta-analysis showing a fourfold higher pooled prevalence rate of papillary thyroid cancer in studies focusing on patients undergoing PTX compared with studies that included PHPT patients regardless of PTX status (8% versus 2%).It is also possible that the increased pooled prevalence rate of breast cancer was secondary to increased cancer screening in the studied population.
In addition to selection bias, there are a number of biological explanations connecting PHPT to the development of malignant neoplasm.First, experimental studies have demonstrated that increased PTH-1 receptor signaling can lead to increased cell proliferation and survival in various cancer cell types, including breast cancer, prostate cancer and renal cell carcinoma [73].Second, it is conceivable that PHPT and non-parathyroid tumors may share certain risk factors.For instance, neck radiation, a well-known risk factor for thyroid cancer, has been shown in multiple studies to contribute to the development of parathyroid tumor [74].It is also important to note that approximately 5-10% of clinically diagnosed multiple endocrine neoplasia type 1 (MEN1) patients have negative genetic test results [75] and that MEN1 has been proposed to predispose individuals to the development of breast cancer [76].This could potentially explain the observed increased risk of breast cancer in patients with PHPT, even after excluding studies focusing on patients with MEN1.However, it remains to be elucidated whether there is a shared genetic predisposition between PHPT and malignant neoplasm outside of known genetic forms of PHPT.Finally, vitamin D deficiency, characterized by low levels of serum  [77,78].Numerous observational studies have revealed association between vitamin D deficiency and an increased risk of cancer [80].This association is believed to be explained by the antiproliferative and prodifferentiative effects of vitamin D receptor activation in various cancer cells [79].However, evidence from large-scale clinical trials does not support a causal association between vitamin D supplementation regardless of baseline vitamin D status in the general population and a reduction in cancer risk [81,82].
This study has certain limitations that warrant acknowledgment.First, while several studies on the prevalence of malignant neoplasms were identified, our study's pooled prevalence results only reflect the prevalence reported in the included studies, with participants having a mean age in the 50s-60s and being predominantly women.As a result, the lifetime risk of malignant neoplasms cannot be extrapolated from our findings.Second, the number of included cohort studies was limited, and all studies were conducted in European countries.This limitation restricts the generalizability of our results to the global population.Additionally, data on the relative risk of specific types of malignant neoplasms were scarce, with information available only from 2 studies [45,56].Furthermore, there was no data from cohort studies assessing the impact of PTX on malignancy risk.Moreover, the meta-analysis showed moderate to high statistical heterogeneity.Differences in participant characteristics and study design likely served as the primary contributors to the observed heterogeneity.Finally, due to the limited number of included cohort studies, it was not possible to evaluate publication bias using funnel plots.

Conclusion
The current systematic review and meta-analysis compiled all existing evidence regarding malignant neoplasms in patients with PHPT.The pooled analysis showed that the prevalence of any form of malignant neoplasm in PHPT patients was around 20%.Notably, papillary thyroid cancer and breast cancer emerged as the most prevalent types, with pooled prevalence rates of 7% and 5%, respectively.These rates are considered disproportionately higher than the rates reported in the general population, which may suggest a new clinical entity that requires further investigations.Notably, the high prevalence of thyroid cancer is likely attributed to increased screening and diagnosis rates a part of the preoperative evaluation for PTX, since the prevalence was approximately 4 times higher among studies focusing on PHPT patients undergoing PTX.Furthermore, a metaanalysis of 5 cohort studies demonstrated an approximately 28% increased risk of malignancy in patients with PHPT compared with comparators.

Fig. 1
Fig. 1 Study identification and literature review process

Fig. 2
Fig.2Forest plot of the meta-analysis of risk of overall cancer in patients with primary hyperparathyroidism versus comparators

Table 1
Characteristics of studies reporting prevalence of malignant neoplasm in patients with primary hyperparathyroidism The results of meta-analysis of prevalence of malignant neoplasm in patients with PHPT are presented in Table3.Based on the results of 9 studies involving a total of 2,377 PHPT patients, the pooled prevalence of overall cancer was 19% (95%CI: 13-25%), with I 2 of 94% indicating high statistical heterogeneity.