Incidence, risk factors and survival outcomes of post-transplant tertiary hyperparathyroidism in kidney recipients

Abstract

BACKGROUND

Post-transplant tertiary hyperparathyroidism (PT-tHPT) is a well-recognized complication following kidney transplantation, characterized by persistent excessive secretion of parathyroid hormone (PTH) despite improved renal function. It is potentially associated with an increased risk of cardiovascular events, renal osteodystrophy, pathologic fractures, graft loss, and mortality.

AIM

To evaluate the incidence, risk factors, and outcomes of PT-tHPT amongst kidney transplant recipients.

METHODS

A total of 887 transplant recipients who underwent transplantation between 2000 and 2020 were evaluated. Univariable and multivariable logistic regression was performed to determine the predictors of tertiary hyperparathyroidism. Graft and recipient outcomes were assessed using multivariable Cox regression. A separate multivariable Cox regression was performed to determine the effect of treatment strategies on outcomes.

RESULTS

PT-tHPT, defined as elevated PTH (> 65 ng/L) and persistent hypercalcemia (> 2.60 mmol/L), was diagnosed in 14% of recipients. Risk factors for PT-tHPT included older age [odds ratio (OR) = 1.36, P < 0.001], Asian ethnicity (OR = 0.33, P = 0.006), total ischemia time (OR = 1.03, P = 0.048 per hour), pre-transplant serum calcium (OR = 1.38, P < 0.001) per decile increase, pre-transplant PTH level (OR = 1.31, P < 0.001) per decile increase, longer dialysis duration (OR = 1.12, P = 0.002) per year, history of acute rejection (OR = 2.37, P = 0.012), and slope of estimated glomerular filtration rate change (OR = 0.91, P = 0.001). There were a 3.4-fold higher risk of death-censored graft loss and a 1.9-fold greater risk of recipient death with PT-tHPT. The three treatment strategies of conservative management, calcimimetic and parathyroidectomy did not significantly change the graft or patient outcome.

CONCLUSION

Pretransplant elevated calcium and PTH levels, older age and dialysis duration are associated with PT-tHPT. While PT-tHPT significantly affects graft and recipient survival, the treatment strategies did not affect survival.

Key Words: Post-transplant tertiary hyperparathyroidism; Kidney transplantation; Parathyroid hormone; Parathyroidectomy; Calcimimetics; Graft survival; Risk factors; Mineral bone disorder; Fibroblast growth factor 23; Treatment outcomes

Core Tip: This study explores post-transplant tertiary hyperparathyroidism (PT-tHPT) in kidney transplant recipients, identifying elevated pre-transplant calcium and parathyroid hormone levels, prolonged dialysis duration, and acute rejection as key risk factors. PT-tHPT significantly increases the risks of graft loss and patient mortality. However, survival outcomes were comparable across treatment strategies: (1) Conservative management; (2) Calcimimetics; and (3) Parathyroidectomy. These findings highlight the need for optimized pre-transplant care and vigilant post-transplant monitoring to reduce PT-tHPT-related complications. Further research is essential to determine the most effective treatment strategy to improve outcomes and minimize morbidity and mortality associated with PT-tHPT.

INTRODUCTION

Post-transplant tertiary hyperparathyroidism (PT-tHPT) is a common complication following kidney transplantation, affecting metabolic parameters and accompanied by morbidity[1]. Tertiary hyperparathyroidism (tHPT) occurs when there is an excessive secretion of parathyroid hormone (PTH) by the parathyroid glands. This is usually associated with hypercalcemia. The tHPT typically develops after prolonged secondary hyperparathyroidism (sHPT). It is commonly seen in patients with end-stage kidney disease (ESKD)[2]. The condition occurs when the over-stimulated parathyroid glands, caused by chronic kidney disease (CKD)-related sHPT, can no longer switch off hormone production once serum calcium concentration has normalized. As a result, they continue to produce large amounts of PTH.

Before transplantation, up to 75% of patients with CKD develop sHPT, with prevalence ranging from 32% in CKD stage 3 to 93% ESKD[3,4]. The sHPT arises due to impaired phosphate homeostasis in CKD, leading to overproduction of fibroblast growth factor 23 (FGF23) and reduced levels of 1,25-dihydroxy vitamin D [1,25(OH)₂D]. Consequently, serum calcium and PTH levels become dysregulated[5,6].

The decline in 1,25(OH)₂D, combined with elevated phosphate levels, reduces intestinal calcium absorption and stimulates excessive PTH secretion. Prolonged stimulation of the parathyroid glands promotes cellular growth, resulting in diffuse or nodular hyperplasia. Nodular hyperplasia, in particular, reduces sensitivity to calcium and vitamin D feedback mechanisms, leading to persistent PTH overproduction even after the initial triggers have resolved[2,5,6].

The sHPT is associated with poor outcomes in CKD. Patients with sHPT have twice the risk of major adverse cardiac events and a fivefold higher risk of CKD progression[7]. Additionally, CKD patients with sHPT are four times more likely to initiate dialysis or die compared to those without sHPT[8].

Following kidney transplantation, most patients experience normalization of PTH levels. However, approximately 25% fail to do so and develop PT-tHPT[9,10]. This condition is characterized by persistently elevated PTH and calcium levels, alongside reduced phosphate levels, despite successful transplantation. Variations in diagnostic thresholds have contributed to inconsistent reports on the incidence and impact of PT-tHPT[2,11].

PT-tHPT has been linked to adverse outcomes, including an increased risk of renal graft dysfunction and graft failure[12]. However, optimal management remains uncertain. Treatment options range from conservative strategies such as dietary modifications and immunosuppression adjustments to pharmacological interventions like calcimimetics and surgical parathyroidectomy. The lack of robust evidence complicates treatment decisions, and standardized guidelines are lacking.

Given its potential impact on graft function and recipient survival, further investigation is essential.

To identify risk factors associated with PT-tHPT, evaluate the impact of PT-tHPT on allograft and patient survival, and to assess the effects of different treatment strategies on survival outcomes.

MATERIALS AND METHODS

Study design and cohort

This retrospective cohort study focused on kidney transplant recipients (KTRs) who underwent transplantation from January 2000 to December 2020. The cohort was identified from a comprehensive database of KTRs regularly followed up at a single tertiary nephrology centre. Exclusion criteria included patients experiencing graft loss within the first three months post-transplantation, individuals with a prior diagnosis of primary hyperparathyroidism (HPT), a history of parathyroidectomy before transplantation, or those receiving calcimimetic treatment for indications other than HPT (such as calcific uremic arteriolopathy). A flow chart of subject selection is depicted in Figure 1.

Figure 1 Flow diagram of the study cohort selection. A total of 963 kidney transplant recipients transplanted between 2000 and 2020 were screened. Of these, 76 patients were excluded due to early graft loss or missing data (n = 54), primary hyperparathyroidism (HPT) (n = 2), calciphylaxis treated with cinacalcet (n = 1), or pre-transplant tertiary HPT (tHPT) (n = 19). The final analysis included 887 patients, of whom 125 developed post-transplant tHPT and 762 did not. HPT: Hyperparathyroidism; KTR: Kidney transplant recipient; tHPT: Tertiary hyperparathyroidism.

Explanatory variables

The explanatory variables included age at transplantation, gender, ethnicity, body mass index, primary kidney disease, dialysis vintage, history of pre-transplant diabetes, allograft type received, degree of human leucocyte antigen mismatch, immunosuppressive regimen, and smoking history. Pre-transplant bone profile parameters were also recorded. These were measured as the median of all the test results over the pre-transplant period and divided into deciles of increments. They included serum calcium (mmol/L), phosphate (mmol/L), and PTH (ng/L). Additionally, the immunosuppression regime, history of biopsy-proven acute rejection, history of opportunistic (DNA) viral infections and smoking history were also recorded. The baseline estimated glomerular filtration rate (eGFR), defined as eGFR at 3 months post-transplantation and calculated using the modification of Diet in Renal Disease equation, was also recorded. This served as a marker of graft function. We also calculated the slope of eGFR decline for each recipient during follow up.

Outcome measures

There is no consensus on a PTH level that clearly defines the presence of persistent post-transplant HPT. In our study, PTH levels above the upper limit of normal (65 pg/mL or 6.9 pmol/L) were considered elevated[13].

PT-tHPT was defined as elevated PTH along with persistent hypercalcemia (serum adjusted calcium > 2.6 mmol/L) for at least 6 months post-transplantation, accompanied by low or normal phosphate levels[13]. Graft survival was evaluated as death-censored graft survival, measuring the time until kidney transplant failure leading to dialysis or re-transplantation; all-cause graft loss was defined as a composite of death-censored graft loss and recipient death with a functioning graft and recipient survival with a functioning graft.

Immunosuppression

All recipients in the study received maintenance immunosuppression, which comprised a calcineurin inhibitor (CNI) based regimen utilizing either tacrolimus or cyclosporin. An anti-proliferative agent, specifically mycophenolic acid or azathioprine was also administered for individuals without contraindications. The administration of corticosteroids was tailored based on the immunologic risk status of the recipients. Those with a standard immunologic risk profile received a brief course of corticosteroids lasting 1-2 weeks at transplantation. In contrast, recipients deemed to be at high immunologic risk, characterized by factors such as younger recipients or older donors, a calculated panel reactive antibody exceeding 20 percent, the presence of a donor-specific antibody, or delayed onset of graft function, underwent an extended duration of maintenance corticosteroid therapy.

Treatment strategies

Treatment strategies for PT-tHPT were categorized into conservative management, calcimimetic therapy (using calcimimetic agents e.g., cinacalcet), and partial or total parathyroidectomy, either as standalone treatments or in combination with calcimimetics.

Data collection

Data was extracted from electronic patient records, and collection was continued until clinical endpoints were reached, including allograft failure, recipient death, recipient loss to follow-up, or the end of the study period in December 2021.

Statistical analysis

Continuous variables with a normal distribution were presented as mean (SD), while continuous variables with non-normal distributions were described using the median and interquartile range (IQR). Categorical variables were presented as frequencies and percentages. Group comparisons were conducted using two-tailed t-tests for parametric variables and Wilcoxon's sign-rank test for non-parametric variables. Categorical variables were compared using Pearson's χ² test or Fisher's exact test.

The rate of eGFR change was expressed as a continuous variable (slope over time), for each patient, was determined by employing a linear mixed-effects regression model. This model was constructed based on the yearly eGFR values for each patient, utilizing the final eGFR value for each year following transplantation. This approach allowed for the calculation of the slope of eGFR change, where higher values reflect slower decline or improving graft function, providing a quantitative measure of the progression of renal function over time for individual patients.

The independent risk factors for PT-tHPT were determined using multivariable binary logistic regression. The multivariable analysis included all covariates that exhibited associations with PT-tHPT in the univariate model with P < 0.20 and progressed through a stepwise backward elimination process. We sought to mitigate the risk of both of overfitting and potentially excluding clinically meaningful variables by using a hybrid approach combining statistical thresholds and clinical relevance. To avoid prematurely excluding potentially important predictors, we first included all variables associated with PT-tHPT in univariate analysis at a P value threshold of < 0.20. We then used stepwise backward elimination guided by the Wald χ² test, with a removal threshold of P > 0.05. This approach helped streamline the model while retaining statistically significant predictors. Importantly, we reviewed the final model for clinical plausibility and retained key covariates that are known to be biologically or clinically relevant, even if marginally non-significant, to preserve interpretability and reduce the risk of omitting important variables.

Multicollinearity diagnostics were performed using generalized variance inflation factors (GVIF). All predictors in the multivariable model demonstrated low collinearity, with adjusted-GVIF values close to 1.0, suggesting no significant inflation of variance due to correlated predictors (Supplementary Table 1).

The performance of the logistic regression model was also evaluated using the receiver operating characteristic (ROC) analysis. The ROC curve was generated by varying the classification threshold, and the true positive rate (sensitivity) and false positive rate (1-specificity) were calculated at multiple threshold values. The area under the curve (AUC) was computed to quantify the model's discriminatory ability.

The impact of PT-tHPT on graft and recipient survival was determined. Kaplan-Meier survival curves were used to illustrate the unadjusted survival of patients and grafts, and comparisons were made using the log-rank test. Multivariable Cox proportional hazards regression models adjusted for confounding variables were used to evaluate the effects of PT-tHPT on death censored graft survival (DCGS), all cause graft survival (ACGS) and recipient survival. Confounding variables for the outcome models were chosen based on previous knowledge through a literature review. A separate Cox regression model was used to assess the effect of the different treatment strategies on graft and recipient outcomes. Statistical significance was determined at a P value less than 0.05.

A sensitivity analysis was conducted after excluding recipients who developed tHPT after three years of transplantation. Cox proportional hazard model was applied to the data to assess the effect of early PPT-tHPT on outcomes.

All statistical analyses were performed using R Statistical Software (version 4.3.0) with the utilization of various R packages, including the “arsenal”, "survival", “pROC”, and “glm” packages.

RESULTS

Characteristics of the cohort

Amongst 963 subjects in the initial study cohort, 76 subjects were excluded after applying the exclusion criteria leaving 887 subjects in the final analysis. Table 1 presents an overview of the cohort characteristics. The mean age at transplantation was 47 ± 15 years, 39% were women, and 81%, identified their ethnicity as white. The cause of kidney failure was glomerulonephritis in 27% of recipients, unknown diagnosis in 17%, adult polycystic kidney disease in 13%, reflux disease in 13% and diabetes in 12%.

Table 1 Characteristics of the study cohort according to the occurrence of post-transplant tertiary hyperparathyroidism, n (%).

Characteristics	PT-tHPT (n = 125)	Non-tHPT (n = 762)	Total (n = 887)	P value
Demographics
Age (years), mean (SD)	50.5 (13.2)	47.0 (15.5)	47.5 (15.2)	0.016
Sex				0.617
Male	74 (59.2)	469 (61.5)	543 (61.2)
Female	51 (40.8)	293 (38.5)	344 (38.8)
Ethnicity				0.028
White	110 (88.0)	609 (79.9)	719 (81.1)
Asian	11 (8.8)	125 (16.4)	136 (15.3)
Black	4 (3.2)	12 (1.6)	16 (1.8)
Other	0 (0.0)	16 (2.1)	16 (1.8)
Pretransplant body mass index, mean (SD)	27.4 (5.0)	27.0 (8.3)	27.1 (7.9)	0.634
Donor characteristics
Donor type category				0.006
Deceased donor	101 (80.8)	523 (68.7)	624 (70.4)
Living donor	24 (19.2)	238 (31.3)	262 (29.6)
Total human leucocyte antigen mismatch, median (IQR)	3.0 (2.0-3.0)	3.0 (2.0-3.0)	3.0 (2.0-3.0)	0.425
Total ischaemia time, mean (SD)	14.1 (7.1)	12.5 (7.7)	12.7 (7.6)	0.037
Recipients’ characteristics
Primary renal disease				0.033
Adult polycystic kidney disease	12 (9.6)	104 (13.6)	116 (13.1)
Glomerulonephritis	32 (25.6)	207 (27.2)	239 (26.9)
Diabetic kidney disease	9 (7.2)	99 (13.0)	108 (12.2)
Hypertensive kidney disease	8 (6.4)	52 (6.8)	60 (6.8)
Reflux/chronic pyelonephritis	28 (22.4)	90 (11.8)	118 (13.3)
Unknown	21 (16.8)	130 (17.1)	151 (17.0)
Other	15 (12.0)	80 (10.5)	95 (10.7)
Dialysis vintage (pre-emptive transplants excluded), median (IQR)	39.0 (20.0-63.0)	23.0 (11.0-38.0)	25.0 (12.0-44.0)	< 0.001
Pretransplant diabetes	20 (16.0)	136 (17.8)	156 (17.6)	0.615
Transplant number, mean (SD)	1.2 (0.5)	1.1 (0.4)	1.1 (0.4)	0.013
Pre-emptive transplant	24 (19.8)	246 (34.2)	270 (32.1)	0.002
Pre-transplant calcium (mmol/L), median (IQR)	2.5 (2.4-2.5)	2.3 (2.2-2.4)	2.3 (2.2-2.4)	< 0.001
Pre-transplant phosphate (mmol/L), median (IQR)	1.6 (1.4-1.9)	1.6 (1.3-1.9)	1.6 (1.4-1.9)	0.153
Pre-transplant PTH (ng/L), median (IQR)	415.8 (232.0-717.9)	185.0 (97.8-307.8)	202.0 (107.0-360.0)	< 0.001
Post-transplant PTH (ng/L), median (IQR)	135.8 (107.6-240.1)	83.7 (58.3-119.9)	88.4 (61.7-138.1)	< 0.001
Smoking history				0.073
Non-smoker	110 (92.4)	604 (86.5)	714 (87.4)
Current smoker	9 (7.6)	94 (13.5)	103 (12.6)
Immunosuppression regime
Main immunosuppression				0.557
Tacrolimus	110 (88.7)	674 (90.1)	784 (89.9)
Mammalian target of rapamycin inhibitor	3 (2.4)	9 (1.2)	12 (1.4)
Cyclosporin	11 (8.9)	65 (8.7)	76 (8.7)
Antimetabolite				0.739
None	14 (11.3)	80 (10.7)	94 (10.8)
Mycophenolic acid	97 (78.2)	571 (76.3)	668 (76.6)
Azathioprine	13 (10.5)	97 (13.0)	110 (12.6)
Steroid regime				0.016
< 6 monhs of glucocorticoid	54 (45.4)	405 (57.2)	459 (55.5)
> 6 months of glucocorticoid	65 (54.6)	303 (42.8)	368 (44.5)
Other
Baseline eGFR (mL/minute/1.73 m2), median (IQR)	50.0 (37.2-62.2)	51.0 (40.8-65.0)	51.0 (40.0-64.0)	0.373
History of acute rejection	20 (16.5)	78 (10.9)	98 (11.7)	0.074
The eGFR slope (mL/minute/year), median (IQR)	1.4 (3.20.0)	0.8 (2.70.5)	0.9 (2.70.5)	0.027

Continuous variables: Presented as mean (SD) or median (interquartile range) and compared using the independent t-test test or Wilcoxon’s sign-rank test. Categorical variables: Presented as n (%) and compared using Pearson’s χ² test. eGFR: Estimated glomerular filtration rate; IQR: Interquartile range; PTH: Parathyroid hormone; PT-tHPT: Post-transplant tertiary hyperparathyroidism; tHPT: Tertiary hyperparathyroidism.

One-third of recipients (32%) underwent pre-emptive transplantation. Among those who did not receive pre-emptive transplantation, median duration receiving dialysis was 25 months (IQR: 12.0-44.0 months). Two-thirds of the cohort (68%) received deceased donor allografts. Most recipients received a tacrolimus-based immunosuppression regimen (90%), with 45% of subjects receiving long-term corticosteroid therapy. The median pre-transplant PTH was 202 ng/L (IQR: 107-360 ng/L) [21.4 pmol/L (IQR: 11.3-38.2 pmol/L)].

Incidence and predictors of PT-tHPT

Over a median follow-up period of 7.3 years (IQR: 4-12 years), 125 subjects (14%) developed PT-tHPT. Table 2 shows the result of the univariable and multivariable logistic regression analysis.

Table 2 Univariable and multivariable logistic regression analysis of the predictors of post-transplant tertiary hyperparathyroidism.

Variables	Univariable analysis				Multivariable model
	n	OR	95%CI	P value	OR	95%CI	P value
Age (per decade)	887	1.17	1.03-1.33	0.016	1.36	1.15-1.62	< 0.001
Gender: Female	887	1.10	0.75-1.62	0.62
Ethnicity	871			0.033	-	-	-
White		1	-		-	-	-
Asian1		0.49	0.24-0.89		0.33	0.14- 0.70	0.006
Black		1.85	0.51-5.41		0.79	0.16-3.02	0.75
Other		0.00	0.00-0.00		0.00	0.00-0.00	0.98
Pretransplant body mass index	887	1.00	0.98-1.03	0.71	-	-	-
Primary renal disease	887			0.044	-	-	-
Adult polycystic kidney disease		1	-		-	-	-
Glomerulonephritis		1.34	0.68-2.81		-	-	-
Diabetic kidney disease		0.79	0.31-1.94		-	-	-
Hypertensive kidney disease		1.33	0.50-3.43		-	-	-
Reflux/chronic pyelonephritis		2.70	1.32-5.79		-	-	-
Unknown		1.40	0.67-3.06		-	-	-
Other		1.63	0.72-3.73		-	-	-
Pre-emptive transplant	887	0.50	0.31-0.78	0.002
Pretransplant diabetes	887	0.88	0.51-1.44	0.61	-	-	-
Donor type: Living donor	887	0.52	0.32-0.82	0.004	-	-	-
Total human leukocyte antigen mismatch	887	1.05	0.92-1.20	0.44	-	-	-
Immunosuppression	887			0.75	-	-	-
Tacrolimus		1	-		-	-	-
Mammalian target of rapamycin inhibitor		1.68	0.37-5.47		-	-	-
Cyclosporin		0.98	0.48-1.84		-	-	-
Antimetabolite	887			0.68	-	-	-
None		-	-		-	-	-
Mycophenolic acid		0.96	0.54-1.83		-	-	-
Azathioprine		0.74	0.33-1.68		-	-	-
Corticosteroid regime	887			0.023	-	-	-
< 6 months of glucocorticoid		-	-		-	-	-
> 6 months glucocorticoid		1.55	1.06-2.28		-	-	-
Total ischaemia time	887	1.03	1.00-1.05	0.038	1.03	1.00-1.07	0.048
Dialysis vintage	887	1.19	1.12-1.26	< 0.001	1.12	1.04-1.20	0.002
Baseline eGFR (per 10 mL/minute increase)	887	0.98	0.90-1.05	0.66	-	-	-
History of acute rejection	838	1.62	0.93-2.72	0.086	2.37	1.19-4.60	0.012
The eGFR slope	887	0.93	0.89-0.97	0.001	0.91	0.87-0.96	0.001
Median pretransplant calcium (per decile increase)	887	1.35	1.25-1.46	< 0.001	1.38	1.26-1.51	< 0.001
Median pretransplant phosphate (per decile increase)	887	1.05	0.99-1.12	0.13	-	-	-
Median pretransplant parathyroid hormone (per decile increase)	887	1.26	1.17-1.36	< 0.001	1.31	1.20-1.43	< 0.001
Posttransplant diabetes	831	1.50	0.91-2.39	0.11	-	-	-
Smoking history	816			0.077			0.070
None-smoker		-	-		-	-
Current smoker		0.53	0.24-1.02		0.48	0.20-1.02

¹Ethnicity was treated as a categorical variable with "White" as the reference group, and other groups (Asian, Black, other) were included as dummy variables in the logistic regression model.

Table 2 shows the unadjusted odds ratio with 95%CI of the factors associated with post-transplant tertiary hyperparathyroidism (PT-tHPT) by the univariable and multivariable logistic regression model. The independent predictors of higher odds of PT-tHPT include age at transplantation, dialysis vintage, history of acute rejection, negative estimated glomerular filtration rate slope, higher pretransplant serum adjusted calcium and higher pretransplant serum parathyroid hormone. Asian ethnicity was associated with lower odds of PT-tHPT. eGFR: Estimated glomerular filtration rate; OR: Odds ratio.

The univariate logistic regression analysis identified several factors associated with posttransplant PT-tHPT. These include age at transplantation, ethnicity (Asian ethnicity), primary renal disease, longer duration of dialysis (per year), preemptive transplantation, longer total ischemia time (per hour), longer duration of glucocorticoid therapy (greater than 6 months), higher pretransplant calcium and PTH levels and higher rate of eGFR decline (per year).

In the multivariable regression model, the factors associated with increased risk of PT-tHPT were older age at transplantation [odds ratio (OR) = 1.36, P < 0.001], prolonged total ischemia time (per hour increase) (OR = 1.03, P = 0.48), dialysis vintage (per year) (OR = 1.12, P = 0.002), history of acute rejection (OR = 2.37, P = 0.012), pretransplant median serum calcium (OR = 1.38, P < 0.001) per decile increase, and pretransplant PTH levels (OR = 1.31, P < 0.001) per decile increase. Conversely, a positive eGFR slope (the rate of eGFR change over time) (OR = 0.91, P = 0.001) and Asian ethnicity (OR = 0.33, P = 0.006) were associated with lower odds of PT-tHPT per unit eGFR change was associated with lower odds of PT-tHPT. The AUC for our model is 0.83 (Figure 2), indicating strong discrimination between individuals who developed PT-tHPT and those who did not.

Figure 2 Receiver operating characteristic curve for the logistic regression model predicting post-transplant tertiary hyperparathyroidism. The X-axis represents the false positive rate (specificity), and the Y-axis represents the true positive rate (sensitivity). The diagonal line (dashed) represents random guessing, while the solid line represents the model's performance. The model demonstrates good discriminatory ability with an area under the curve of 0.834, indicating high accuracy in distinguishing between patients who developed post-transplant tertiary hyperparathyroidism and those who did not. AUC: Area under the curve; ROC: Receiver operating characteristic.

Outcomes associated with PT-tHPT

A total of 138 (15.5%) deaths and 87 (9.8%) graft losses were recorded over a median follow-up period of 7.3 years (88 months). The median death-censored graft survival was more than 20 years in both the tHPT and the non-tHPT group. The median recipient survival was more than 20 years in the no- tHPT group and 16 years in the tHPT group (Supplementary Figure 1).

In the unadjusted analysis, the 1-year, 5-year, 10-year, 15-year, and 20-year death censored allograft survival rates for recipients without PT-tHPT were 99.7%, 98%, 91%, 82%, and 68%, respectively. In contrast, recipients with PT-tHPT exhibited lower allograft survival rates of 99%, 89%, 73%, 58%, and 51% at the corresponding time points (log rank P < 0.001). Similarly, the unadjusted recipient survival rates at 1-year, 5-year, 10-year, 15-year, and 20-year were 100%, 95%, 82%, 73%, and 57%, respectively in recipients without PT-tHPT. In contrast, the survival rates in those who experienced PT-tHPT were 99%, 91%, 71%, 51%, and 35%, respectively (P = 0.006).

Figure 3 shows the impact of PT-tHPT on DCGS, all-cause graft survival and recipient survival adjusted for confounding variables. After multivariable Cox proportional hazard regression (adjusted for age, donor type, history of acute rejection and baseline eGFR), DCGS was 3.4 times higher in recipients who experience PT-tHPT [adjusted hazard ratio (aHR) = 3.41; 95%CI: 2.09-5.59, P < 0.001; Figure 3A]. Similarly, after adjusting for age, donor type, history of acute rejection, baseline eGFR and post-transplant smoking history, ACGS was 2.5 times higher in recipients with a history of tHPT (aHR = 2.51, 95%CI: 1.78-3.53, P < 0.001; Figure 3B). Recipient survival adjusted for age, history of diabetes, duration of requiring dialysis, history of acute rejection, baseline eGFR and post-transplant smoking history, was 1.94 times higher in those who had tHPT (aHR = 1.94, 95%CI: 1.25-3.01, P = 0.003).

Figure 3 Adjusted survival analysis comparing outcomes in kidney transplant recipients with post-transplant tertiary hyperparathyroidism vs those without tertiary hyperparathyroidism. A: Death-censored graft survival was significantly lower in patients with tertiary hyperparathyroidism (tHPT) [adjusted hazard ratio (aHR) = 3.41, 95%CI: 2.09-5.59, P < 0.001], adjusted for age, donor type, acute rejection, and baseline estimated glomerular filtration rate (eGFR); B: All-cause graft survival was significantly reduced in the tHPT group (aHR = 2.51, 95%CI: 1.78-3.53, P < 0.001), adjusted for age, acute rejection, baseline eGFR, and post-transplant smoking; C: Recipient survival was also significantly lower in the tHPT group (aHR = 1.94, 95%CI: 1.25-3.01, P = 0.003), adjusted for age, pre-transplant diabetes, dialysis duration, history of acute rejection, baseline eGFR, and post-transplant smoking. aHR: Adjusted hazard ratio; eGFR: Estimated glomerular filtration rate; tHPT: Tertiary hyperparathyroidism.

PT-tHPT: Treatment and outcomes

Table 3 and Figure 4 summarized the characteristics and outcomes of the recipients who experienced PT-tHPT stratified by their different treatment modalities.

Figure 4 Survival probability (adjusted for pre-transplant parathyroid hormone). A: Recipient survival by treatment modality; B: Graft survival by treatment modality; C: All-cause survival by treatment modality. Adjusted hazard ratio and P values are shown in the legend for each treatment group, using conservative management as the reference category. There was no statistically significant difference in recipient survival in those who received active treatment compared to those managed conservatively. Cox regression survival curves were adjusted for baseline pre-transplant parathyroid hormone levels using multivariable Cox proportional hazards models. aHR: Adjusted hazard ratio; PTH: Parathyroid hormone.

Table 3 Characteristics and outcomes of recipients with post-transplant tertiary hyperparathyroidism stratified by treatment modalities, n (%).

Variables	Conservative management (n = 40)	Calcimimetics (n = 42)	parathyroidectomy (n = 41)	Total (n = 123)	P value
Age (years), mean (SD)	52.1 (14.2)	50.4 (11.6)	49.6 (12.8)	50.7 (12.8)	0.679
Duration of pre-transplant renal replacement therapy (months), median (IQR)	24.0 (6.0-54.0)	48.5 (24.0-66.0)	24.0 (1.5-50.0)	31.0 (7.0-57.0)	0.049
Baseline eGFR (mL/minute/1.73 m2), median (IQR)	50.4 (37.0-59.2)	53.5 (43.0-71.8)	48.0 (38.0-59.0)	50.8 (39.0-62.5)	0.174
Median pre-transplant calcium (mmol/L), median (IQR)	2.5 (2.4-2.5)	2.4 (2.3-2.5)	2.5 (2.4-2.6)	2.5 (2.4-2.6)	0.184
Median pre-transplant phosphate (mmol/L), median (IQR)	1.6 (1.4-1.9)	1.7 (1.4-2.0)	1.6 (1.4-1.8)	1.6 (1.4-1.9)	0.621
Median pre-transplant parathyroid hormone (pmol/L), median (IQR)	295.0 (138-478)	643.5 (340-1163)	374 (190-599)	374 (216-719)	< 0.001
The eGFR slope (mL/minute/year), median (IQR)	-1.02 (-1.94 to 0.38)	-1.78 (-3.80 to 0.01)	-1.09 (-3.21 to -0.18)	-1.27 (-3.67 to -0.09)	0.378
History of acute rejection	4 (10.5)	5 (12.5)	10 (24.4)	19 (16.0)	0.186

eGFR: Estimated glomerular filtration rate; IQR: Interquartile range.

Three management strategies were employed to address PT-tHPT among the recipients. Of the total recipients who experienced PT-tHPT, a third (n = 40) were managed conservatively, a third (n = 42) were treated with calcimimetics (cinacalcet), and a third (n = 41) underwent parathyroidectomy (with 22 having a partial parathyroidectomy and 19 undergoing total parathyroidectomy). Data on treatment modality was not available for two recipients. There were no significant differences in recipients’ age, baseline eGFR and pretransplant bone profile among the treatment groups. Of the three groups, the calcimimetic group had spent the longest period receiving dialysis and had the highest median pretransplant PTH levels. Amongst the three modalities of treatment, acute rejection, DCGS and ACGS did not differ significantly although the parathyroidectomy group had numerically higher rejection rates, lower DCGS and lower ACGS.

Although there was a trend toward lower recipient survival in the conservative group (Figure 4A), this difference was not statistically significant and should be interpreted with caution. The observed pattern may reflect underlying differences in patient characteristics or care pathways, but further studies with larger sample sizes are needed to determine whether this association is clinically meaningful.

Sensitivity analysis

In the sensitivity analysis, we applied a Cox regression analysis to the data to assess graft, and recipient outcomes after excluding recipients who developed PT-tHPT after the 3^rd year of transplantation. The result was not different from that obtained with all the occurrences of PT-tHPT (Supplementary Figure 2).

DISCUSSION

The observed incidence of PT-tHPT in our cohort was 14%, lower than previously reported rates ranging from 17% to 86%[11,12,14]. Among affected patients, 15% experienced graft loss, and 24% died over a median follow-up of 7.3 years.

Significant risk factors for PT-tHPT included older age, elevated pretransplant serum calcium and PTH levels, longer dialysis duration, prolonged ischemia time, a history of acute rejection, and a faster eGFR decline.

Notably, Asian ethnicity was negatively associated with PT-tHPT, with an OR suggesting a potentially protective effect. While the biological plausibility of this association is not immediately clear, several hypotheses may explain this finding. Cultural and dietary factors such as higher intake of plant-based foods, lower dietary phosphorus load, or differing patterns of vitamin D intake and level exposure to ultraviolet light due to cultural factors may influence post-transplant mineral metabolism and warrant further exploration. Additionally, genetic differences, including variations in genes regulating calcium-phosphate homeostasis or PTH signalling pathways, may contribute to the observed effect. Although our dataset did not include detailed dietary or genetic information, these potential explanations will be important to investigate in future studies. We therefore recommend caution in interpreting this association and suggest that future research explore ethnic-specific risk factors in more detail.

The strong correlation between pretransplant serum calcium and PTH levels in our study aligns with prior findings. Hong et al[15] previously identified pretransplant intact PTH, serum calcium levels, and dialysis duration as key factors associated with post-transplant HPT. The link between prolonged dialysis and PT-tHPT supports the idea that chronic parathyroid gland stimulation during the pretransplant period leads to autonomous HPT post-transplant. The sHPT in advanced CKD results from multiple factors, including hyperphosphatemia, hypocalcaemia, low 1,25(OH)₂D levels, and skeletal resistance to PTH. These factors result in continuous stimulation of PTH synthesis and secretion. The parathyroid hyperplasia that develops is initially diffuse and polyclonal and usually responds to vitamin D therapy and calcimimetics. However, with prolonged sHPT, over time, there is down regulation of vitamin D receptors and calcium-sensing receptors in the parathyroid tissue, and hyperplasia becomes monoclonal and nodular in nature[16,17]. In such patients, PTH synthesis and secretion become autonomous with minimal response to therapeutic agents. This is often associated with hypercalcemia. Parathyroid imaging often shows that diffuse polyclonal hyperplastic glands are significantly smaller than nodular monoclonal glands[18].

With successful transplantation and higher eGFR, most of these stimuli of parathyroid hyperplasia abate. This often leads to a gradual decline in PTH concentrations. However, reports have shown that up to 25%-80% of patients still have inappropriately high PTH beyond 1-year post transplant[16]. Pre-transplant cinacalcet use, development of nodular hyperplasia, and dialysis vintage are associated with high PTH levels after transplantation. Thus, in transplant recipients, the clinical manifestations of HPT resemble primary HPT and are characterised by hypercalcemia and hypophosphatemia due to the effects of PTH on the kidney. Whereas nontransplant patients with CKD with HPT typically have hyperphosphatemia and hypocalcaemia.

In multivariable analysis, a higher eGFR slope was associated with a reduced risk of PT-tHPT (OR = 0.91, 95%CI: 0.87-0.96, P = 0.001), indicating that preserved or improving renal function is protective. This association may be explained by the development or persistence of sHPT in the context of worsening allograft function. Better-preserved or improving graft function may reduce the biochemical stimuli, such as hyperphosphatemia and impaired vitamin D metabolism, which drive PTH secretion. In contrast, declining graft function perpetuates mineral dysregulation, contributing to sustained parathyroid stimulation, progressive glandular hyperplasia, and autonomous hormone secretion.

Our most significant finding was the association between PT-tHPT and an increased risk of graft loss and recipient mortality. Recipients with PT-tHPT had a 3.4-fold higher risk of death-censored graft loss, a 2.5-fold higher risk of all-cause graft loss, and a 1.9-fold higher risk of recipient death compared to those without PT-tHPT. These results mirror previous studies[15] and emphasize the importance of early intervention and proactive management[19-23].

The exact mechanisms linking PT-tHPT to graft loss and mortality remain unclear. However, evidence suggests that persistent hypercalcemia in PT-tHPT contributes to renovascular calcification, nephrocalcinosis, and progressive graft dysfunction[23,24]. Elevated PTH levels also directly promote systemic vascular and cardiac complications, including arterial calcification, hypertension, left ventricular hypertrophy, and cardiovascular dysfunction. These effects are mediated through increased intracellular calcium, mitochondrial dysfunction, oxidative stress, and inflammation, ultimately leading to cardiac myocyte necrosis[21,25]. Moreover, PTH upregulates pro-atherosclerotic and pro-inflammatory mediators such as receptors for advanced glycation end products, interleukin-6, and vascular endothelial growth factor, which may contribute to vascular remodelling and atherosclerosis[24,26,27].

Beyond these established pathways, recent evidence highlights a complex interplay between PTH and FGF23. PTH stimulates FGF23 production in bone, while FGF23, in turn, suppresses PTH secretion via FGFR-Klotho signalling in the parathyroid glands[28]. In patients with CKD or post-transplant mineral dysregulation, this feedback loop may become impaired, particularly due to reduced Klotho expression in hyperplastic parathyroid tissue, contributing to sustained PTH elevation and resistance to FGF23. This dysregulation may further exacerbate phosphate retention, vascular calcification, and cardiovascular injury[29]. Furthermore, expression of FGF23 has been demonstrated in various cardiovascular cell types, including cardiac myocytes, cardiac fibroblasts, coronary artery endothelial cells, vascular smooth muscle cells, and inflammatory macrophages. Moreover, the heart itself has been identified as a source of FGF23 release. Experimental and clinical data suggest that FGF23 exerts paracrine effects that promote pro-hypertrophic, pro-fibrotic, and pro-inflammatory signalling pathways. Elevated circulating levels of FGF23 have been clinically associated with adverse cardiovascular conditions such as acute decompensated heart failure, ischemic and dilated cardiomyopathy, and myocarditis[30]. These findings demonstrate the impact of the FGF23-PTH axis on the graft loss and mortality.

Management of PT-tHPT

The optimal management strategy for PT-tHPT remains unclear. In our cohort, three treatment approaches were used, but patient outcomes did not significantly differ between treatment modalities. However, recipient survival appeared numerically lower in the conservatively managed group. These findings are consistent with the EVOLVE trial, which examined cinacalcet use in dialysis patients and found no survival benefit compared to placebo[31].

Studies comparing surgical and medical management of PT-tHPT have produced mixed results. Some have reported a survival advantage with parathyroidectomy over medical therapy in KTRs[2,32], while others found no significant difference[26]. A systematic review by Dulfer et al[33], noted a higher biochemical cure rate with parathyroidectomy but no clear difference in graft survival compared to cinacalcet therapy. In contrast, a retrospective single-centre study by Finnerty et al[34], reported improved graft survival in patients treated with parathyroidectomy vs those receiving cinacalcet. However, these studies are not directly comparable. Dulfer et al’s review[33] included 47 studies, most of which were observational and heterogeneous in design, patient populations, treatment protocols, and follow-up duration. Due to this heterogeneity, the authors were unable to conduct a meta-analysis, which limits the strength of the conclusions that can be drawn. In contrast, Finnerty et al's findings[34] were derived from a retrospective cohort of 133 renal transplant recipients at a single centre followed up between 2000 and 2017. Its smaller sample size and retrospective design also limited the generalizability of its findings. Given the methodological heterogeneity and limitations in current evidence, it remains unclear which treatment modality offers superior outcomes in PT-tHPT. Well-designed randomized controlled trials and prospective longitudinal studies are warranted to establish the optimal therapeutic strategy.

Limitations

This study has several limitations that should be acknowledged. First, its observational design precludes definitive causal inference, and residual confounding may persist despite adjustment for measured baseline characteristics.

Second, serum vitamin D levels were not routinely measured in our transplant population, and data on calcium and vitamin D supplementation were inconsistently recorded. These unmeasured variables may have influenced PTH levels and treatment responses, limiting our ability to account for their potential confounding effects.

Third, accurately characterising long-term immunosuppression exposure was challenging due to frequent regimen modifications. CNI doses vary with therapeutic drug monitoring, antimetabolites are often interrupted due to infections, cytopenias, or malignancy, and glucocorticoids may be reintroduced in selected cases. These dynamic changes limit the precision with which immunosuppressive burden could be assessed in relation to PT-tHPT outcomes.

Fourth, treatment allocation was not randomised. Patients in the calcimimetic group had higher baseline PTH levels, and other differences in clinical characteristics may have influenced the choice of therapy and subsequent outcomes. Importantly, propensity score methods were not employed to adjust for these differences, and unmeasured confounders such as frailty, comorbidity burden, and clinician-driven selection bias may have affected treatment comparisons.

Fifth, this was a single-centre study based in Northwest England and included transplants performed between 2000 and 2020. These temporal and geographical constraints may limit the generalisability of the findings to other populations or clinical settings.

Sixth, patients with early graft loss (within three months post-transplant) were excluded from the analysis. This may have introduced survivorship bias and resulted in underrepresentation of individuals with early post-transplant complications, who may follow different clinical trajectories than those with long-term stable graft function. As PT-tHPT is typically a long-term complication, our findings may not fully reflect the spectrum of transplant-related mineral-bone disorders.

Finally, the absence of statistically significant survival differences between treatment modalities may reflect the influence of unmeasured confounding. Given that important clinical characteristics affecting both therapy selection and outcomes were not systematically recorded or adjusted for, the findings should be interpreted with caution. This underscores the need for prospective, longitudinal, and ideally randomised studies incorporating standardised data collection to better evaluate the comparative effectiveness of surgical and medical management strategies in PT-tHPT.

CONCLUSION

In conclusion, our findings underscore the burden of PT-tHPT in KTRs, emphasizing its adverse impact on allograft and patient survival. The study emphasizes the significance of pretransplant management of sHPT, advocating for early intervention and a proactive, patient-tailored approach to PT-tHPT management.