Association Between Treatment of Secondary Hyperparathyroidism and Posttransplant Outcomes

Abstract

Background:

Secondary hyperparathyroidism (SHPT) affects nearly all patients on maintenance dialysis therapy. SHPT treatment options have considerably evolved over the past 2 decades, but vary in degree of improvement in SHPT. Therefore, we hypothesize that the risks of adverse outcomes after kidney transplantation (KT) may differ by SHPT treatment.

Methods:

Using the SRTR and Medicare claims data, we identified 5,094 adults (age≥18) treated with cinacalcet or parathyroidectomy for SHPT prior to receiving KT between 2007–2016. We quantified the association between SHPT treatment and delayed graft function and acute rejection using adjusted logistic models and tertiary hyperparathyroidism (THPT), graft failure, and death using adjusted Cox proportional hazards; we tested whether these associations differed by patient characteristics.

Results:

Of 5094 KT recipients who were treated for SHPT while on dialysis, 228 (4.5%) underwent parathyroidectomy and 4866 (95.5%) received cinacalcet. There was no association between treatment of SHPT and posttransplant delayed graft function, graft failure or death. However, compared to patients treated with cinacalcet, those treated with parathyroidectomy had a lower risk of developing THPT (aHR=0.56, 95%CI: 0.35–0.89) post-KT. Furthermore, this risk differed by dialysis vintage (p_interaction=0.039). Among patients on maintenance dialysis therapy for ≥3 years prior to KT (n=3,477, 68.3%), the risk of developing THPT was lower when treated with parathyroidectomy (aHR=0.43, 95%CI: 0.24–0.79).

Conclusions:

Parathyroidectomy should be considered as treatment for SHPT, especially in KT candidates on maintenance dialysis for ≥3 years. Additionally, patients treated with cinacalcet for SHPT should undergo close surveillance for development of tertiary hyperparathyroidism post-KT.

INTRODUCTION

Secondary hyperparathyroidism (SHPT), or excess production of parathyroid hormone (PTH) by the parathyroid glands, affects nearly all patients with kidney failure on maintenance dialysis therapy.¹ While modest increases in parathyroid hormone levels may reflect an appropriate compensatory response, levels above 600pg/mL have been independently associated with increased mortality, cardiovascular morbidity, bone loss, fractures, and decreased health-related quality of life.^1–5

Therefore, guidelines by Kidney Diseases Improving Global Outcomes (KDIGO) recommend treatment of SHPT to maintain a parathyroid hormone (PTH) level at 2–9 times the normal range.⁶ Current treatment strategies to reduce PTH levels in patients with SHPT include pharmacotherapy with vitamin D analogs, phosphate binders, and calcimimetics, of which cinacalcet was the first to be approved in 2004, or surgical parathyroidectomy. While treatment with cinacalcet decreases mean PTH levels by about 40%, only 43% of patients reach target PTH levels as recommended by KDIGO (<600pg/ml).^7,8 Additionally, side effects from cinacalcet can lead to problems with adherence limiting its effectiveness and utilization.⁹ Therefore, there is a lack of consensus in recommending calcimimetics as first line treatment. In contrast, parathyroidectomy is an effective method for long-term control, significantly lowering PTH levels into the normal laboratory reference range (<65 pg/mL) in 90–100% of patients.^10,11 However, parathyroidectomy is typically reserved for patients with refractory SHPT or those unable to tolerate calcimimetics even though it results in greater drop in PTH levels.^12,13

Furthermore, studies have demonstrated an association between severity of SHPT pre-KT and higher rates of post-KT graft failure and persistent hyperparathyroidism, also known as tertiary hyperparathyroidism (THPT).^14–16 Therefore, it is often recommended to delay KT until SHPT is adequately treated as KT resolves SHPT in only 57% of recipients.¹⁷ However, a paucity of evidence exists regarding an optimal treatment regimen for SHPT prior to KT to minimize adverse sequelae post-KT and decrease the incidence of THPT. The impact of the various treatment options for SHPT on other post-KT outcomes, such as mortality and graft function, is not well understood. Because the degree of PTH reduction for medical versus surgical treatment varies, we hypothesize that the risk of post-KT outcomes may differ by SHPT treatment.

The primary objectives of our study were to quantify the association between treatment for SHPT prior to transplant and post-KT adverse outcomes in Medicare beneficiaries and to identify subgroups of patients who would benefit from these treatments.

MATERIALS AND METHODS

Data Sources

Data from the Scientific Registry of Transplant Recipients (SRTR) was utilized for this study. The SRTR data system includes data on all donors, waitlisted candidates, and transplant recipients in the United States, submitted by the members of the Organ Procurement and Transplantation Network (OPTN). The Health Resources and Services Administration (HRSA), US Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors. SHPT, THPT, and medical interventions were ascertained using Medicare claims data from the United States Renal Data System (USRDS).

Study Population

Adult (age≥18), first-time KT recipients between January 1, 2007 and December 31, 2016 were identified (n=142 738). Recipients who did not receive dialysis prior to KT (n=28 638) or did not enroll in Medicare Part A, B and D between dialysis initiation and KT (n=96 293) were excluded to identify recipients with Medicare as a primary payor (n=17 807). A 4-month gap was allowed for Medicare enrollment since the start date of dialysis.

Diagnosis of SHPT was then ascertained through Medicare inpatient/outpatient claims using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) of 588.81 and International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) of N25.81 codes. We used ICD-9-CM and ICD-10-CM codes as ICD-10-CM was implemented on October 1, 2015. Recipients without claims of cinacalcet, or parathyroidectomy were then excluded (n=11,378).

Among the recipients with claims of cinacalcet, or parathyroidectomy during dialysis (n=6429), those that had claims of SHPT, cinacalcet, or parathyroidectomy prior to dialysis initiation (n=1335) were further excluded. The study population included 5094 KT recipients who were followed until death, graft failure, end of Medicare coverage, end of Medicare data availability (12/31/2016) or 3-years post-KT, whichever came first (Figure 1).

Figure 1.

Flowchart Demonstrating Identification of the Study Population (n=12 372) of Adult Kidney Transplant Recipients (2007–2016) with Secondary Hyperparathyroidism (SHPT)

a. 4-month gap was allowed for Medicare enrollment. SHPT; secondary hyperparathyroidism

SHPT Treatments

Parathyroidectomy was ascertained by using a list of Current Procedural Terminology (CPT), International Classification of Diseases, Ninth Revision, Procedure Coding System (ICD-9-PCS), and International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS) codes, which were converted to approximate correspondences of ICD-9-PCS codes (Supplement). We used Medicare Part D data to obtain if patients were dispensed a prescription for cinacalcet. Patients who underwent parathyroidectomy for SHPT and did not receive cinacalcet after the operation were referred to as the “parathyroidectomy” group. Those who received cinacalcet only or continued to receive cinacalcet after parathyroidectomy were categorized as the “cinacalcet” group, as this implied that parathyroidectomy was not performed appropriately. Complications from parathyroidectomy including hungry bone syndrome, vocal cord paresis, reoperation, and cardiovascular events were also ascertained. Cardiovascular events included myocardial infarction, congestive heart failure, and stroke.

Post-KT Adverse Outcomes

Tertiary hyperparathyroidism (THPT) was defined as having ICD-9-CM or ICD-10-CM codes of THPT (ICD-9-CM, 252.08; ICD-10-CM, E21.2) or hypercalcemia (ICD-9-CM, 275.42; ICD-10-CM, E83.52) at least 6-months after KT. Prior studies have demonstrated that therapy for hyperparathyroidism should be instituted 3 months after KT because the most dramatic fall of parathyroid hormone occurs within the first 3 months after kidney transplant, and that PTH levels are unlikely to return to normal when elevated between 6 months to 1 year post-KT.^18–20 Due to a lack of consensus definition for tertiary hyperparathyroidism and ambiguity regarding a time point for diagnosis, we also performed a sensitivity analysis defining THPT as any claim of THPT or hypercalcemia 1-year posttransplant. Delayed graft function (DGF) was defined as requiring dialysis within the first week after KT. Because SRTR does not provide a date for onset of acute rejection (AR), it was defined it as experiencing either biopsy confirmed or treated AR within 1-year post-KT.^21,22 Death-censored graft failure (DCGF) was defined as resuming dialysis or being retransplanted, censoring for death. All-cause graft failure (ACGF) was defined as either death or graft loss. Death was obtained from multiple sources, including follow-up reports from transplant centers, Centers for Medicare & Medicaid Services ESKD Death Notification Form (CMS 2746), and the Social Security Death Master File.

Statistical Analysis

Cox proportional hazard models were utilized to compare hazard of THPT, DCGF, ACGF, and death, separately by treatment. Logistic models were used to compare the odds of DGF and AR, separately by treatment. Time-to-discharge was treated as a time-to-event analysis to compare length of stay (LOS) by treatment.²³ Parametric survival models assuming exponential distribution were used to compare LOS by treatment.

All models were weighted using propensity score to adjust for confounding.²⁴ As some transplant-related factors and all donor-related factors are typically determined after the treatment of interest, they might not meet the conditions for being confounders.²⁵ Therefore, several multivariable propensity score models including different lists of confounders were fit. The first model was included for age at listing, sex, race, education, BMI, cause of kidney failure, peak panel reactive antibody (PRA), history of any malignancy, years on dialysis and calendar year of dialysis initiation. The second model included for transplant-related factors (HLA mismatch, cold ischemia time, ABO incompatibility, calendar year of transplant and donor-related factors: age, sex, race, deceased donor, expanded criteria donor) in addition to the covariates included in the previous model. Missing values were handled by including indicator variables for missingness.

We tested whether the associations between treatment for SHPT and post-KT adverse outcomes differed by age at listing (18–54 versus 55+ years), sex, race, BMI (<25, 25–29.9, and 30+ Kg/m²), PRA (0–79 versus 80–100), calendar year of transplant (2007–2010, 2011–2013, and 2014–2016), years on dialysis (<3 versus 3+ years), hemodialysis modality, and donor type (living versus deceased). P-value for interaction term less than 0.05 indicated that the association between treatment and THPT differed across the subgroups. A p-value less than 0.05 was considered statistically significant. All analyses were performed using Stata 16.0/MP for Linux (College Station, Texas) and R version 3.6.2.

RESULTS

Characteristics of KT Recipients with SHPT

Among patients who were treated for SHPT, the median age at listing for transplant was 47 years (IQR: 35–58), 39.8% were female, 34.3% were white, 39.1% were African American, and 19.6% were Hispanic/Latino. Median BMI was 29.0 Kg/m² (IQR: 25.0–33.0). Cause of kidney failure was hypertension in 29.2% patients, diabetes in 27.8% patients, and glomerulonephritis in 20.5% patients. Patients were on dialysis for a median of 4.0 years (IQR: 2.6–5.8). Of this cohort, 89.7% had a history of hypertension, 34.2% had a history of diabetes, and only 4.8% had a history of any malignancy.

SHPT Treatment

Of 5094 patients with SHPT, 205 (4.0%) underwent parathyroidectomy and 4866 (96.0%) received cinacalcet (Table 1). There were significant differences in most patient characteristics between the 2 treatment groups. Patients who underwent parathyroidectomy were younger, with median age of 40 (IQR: 31–50) years, compared to those treated with cinacalcet (median=47 years, IQR: 36–58, p<0.001). Slightly over half of patients who underwent parathyroidectomy were female (52.6%), compared to only 39.2% receiving cinacalcet (p<0.001). A higher proportion of African American patients underwent treatment with parathyroidectomy (44.7%) compared to treatment with cinacalcet (38.8%). Median BMI was similar between the cinacalcet and parathyroidectomy groups (29.0 Kg/m², IQR: 25.0–33.0 and 28.0 Kg/m², IQR: 24.0–32.0, respectively).

Table 1.

Characteristics of Patients with Secondary Hyperparathyroidism (SHPT) who Underwent Kidney Transplant (KT) Between 2007–2016 Stratified by SHPT Treatment (Parathyroidectomy or Cinacalcet).

	Parathyroidectomy (n = 205)	Cinacalcet (n = 4,866)
Patient Characteristics
Age at listing, years	40.0 (31.0, 50.0)	47.0 (36.0, 58.0)
Female	52.6%	39.2%
Race
White	34.2%	34.3%
African American	44.7%	38.8%
Hispanic/Latino	17.5%	19.7%
Other/multiracial	3.5%	7.2%
Attended college	39.0%	36.6%
Body mass index, Kg/m2	28.0 (24.0, 32.0)	29.0 (25.0, 33.0)
Cause of kidney failure
GN	25.4%	20.3%
DM	14.5%	28.5%
HTN	28.5%	29.2%
Others	31.6%	22.0%
Peak panel reactive antigen	2.5 (0.0, 29.5)	0.0 (0.0, 21.0)
History of hypertension	89.5%	89.7%
History of diabetes	18.1%	34.9%
Any malignancy	2.6%	4.9%
Years on dialysis	5.4 (3.4, 6.9)	4.0 (2.6, 5.7)
Calendar year of dialysis initiation
2005–2009	74.1%	64.9%
2010–2012	24.1%	28.6%
2013–2016	1.8%	6.5%
Transplant Characteristics
Zero HLA mismatch	4.8%	4.8%
Cold ischemia time, h	14.1 (8.2, 20.6)	14.8 (8.1, 21.3)
ABO incompatibility	0.9%	0.8%
Calendar year of transplant
2007–2010	16.7%	20.4%
2011–2013	29.4%	32.7%
2014–2016	53.9%	46.9%
Donor age, y	36.0 (24.5, 47.0)	39.0 (26.0, 51.0)
Donor female	42.1%	43.6%
Donor race
White	68.0%	65.7%
African American	17.5%	15.6%
Hispanic/Latino	12.7%	15.1%
Other/multiracial	1.8%	3.6%
Deceased donor	86.8%	82.3%
Donation after circulatory deatha	20.2%	18.4%
Expanded criteria donora	8.6%	14.6%

For continuous variables, median (interquartile range) is presented.

^aAmong deceased donor

The median number of years on dialysis was 5.4 (IQR 3.4–6.9) years for the parathyroidectomy group compared to 4.0 (IQR: 2.6–5.7) years for those on cinacalcet (p<0.001). While nearly 90% in each treatment group had hypertension, diabetes was less common in those who underwent parathyroidectomy compared to the cinacalcet group (18.1% versus 34.9%, p<0.001). A lower proportion of patients who had any history of a malignancy underwent parathyroidectomy compared to the cinacalcet groups (2.6% versus 4.9%).

Among patients that underwent parathyroidectomy, 16.7% underwent KT between 2007–2010, 29.4% between 2011–2013, and increased to 53.9% between 2014–2016. Among patients treated with cinacalcet, 20.4% underwent transplant between 2007–2010, 32.7% between 2011–2013, and increased to 46.9% between 2014–2016. Complication rates from parathyroidectomy included hungry bone syndrome in 48 patients (21.1%), and cardiovascular events in 24 patients (10.5%). No patients required a reoperation or had a vocal cord paresis. Of 228 underwent parathyroidectomy, no one received vitamin D analogs (calcitriol, doxercalciferol, or paricalcitol) between parathyroidectomy and KT and 2 (0.9%) received vitamin D analogs after KT.

Association Between SHPT Treatment and Post-KT Adverse Outcomes

When THPT was defined as persistent hyperparathyroidism 6 months after KT, the cumulative incidence of THPT at 3-years post-KT was 21.6% among patients treated with cinacalcet and 14.1% among those who underwent parathyroidectomy. After adjustment, patients who underwent parathyroidectomy had a lower risk of developing THPT when defined at 6 months post-KT (aHR=0.56, 95%CI: 0.35–0.90) and also when THPT was defined at 1-year post-KT (aHR=0.50, 95% CI:0.27–0.92) compared to those treated with cinacalcet (Table 2). The 3-year cumulative incidences of death, fracture and cardiovascular events were lower among recipients who underwent parathyroidectomy compared to those treated with cinacalcet (5.8% versus 7.1% for death; 1.0% versus 1.8% for fracture;16.3% versus 17.4% for cardiovascular events).

Table 2.

Risk of Tertiary hyperparathyroidism (THPT), defined as occurring at 6 months and at 1 year, by Pre–Kidney Transplant (KT) treatment for Secondary Hyperparathyroidism (SHPT).

Tertiary Hyperparathyroidism (Definition)	Unadjusted HR (95% CI)	Adjusted for Pre-KT Factorsa HR (95% CI)	Adjusted for Additional Perioperative KT factorsb HR (95% CI)
6 months	0.58 (0.36–0.93)*	0.59 (0.37–0.95)*	0.56 (0.35–0.90)*
1 year	0.51 (0.29–0.94)*	0.54 (0.30–0.99)*	0.50 (0.27–0.92)*

^*p < 0.05

^aAdjusted for age, sex, race, education, body mass index, cause of kidney failure, peak panel reactive antigen, history of malignancy, diabetes, hypertension, calendar year of dialysis initiation

^bAdjusted for (¹) + years on dialysis, human leukocyte antigen mismatch, cold ischemic time, ABO incompatibility, calendar year of transplant, type of induction, and donor factors (age, sex, race, deceased donor, expanded criteria donor)

There was no association between parathyroidectomy and DGF (aOR=1.04 95% CI: 0.76–1.41). There was no difference in the odds of AR after parathyroidectomy (aOR=0.82, 95% CI: 0.51–1.31) compared to cinacalcet treatment.

There was no association between parathyroidectomy and DCGF (aOR=1.50, 95% CI: 1.00–2.27), ACGF (aOR=1.37, 95% CI: 0.96–1.95), or death (aOR=1.15, 95% CI: 0.64–2.07) compared to those treated with cinacalcet prior to KT (Table 3).

Table 3.

Association Between Secondary Hyperparathyroidism Treatment (Parathyroidectomy vs Cinacalcet) and Post–Kidney Transplant (KT) Adverse Outcomes. Odds Ratio of Acute Rejection and Delayed Graft Function, Hazard Ratio of Death-Censored Graft Failure, All-Cause Graft Failure and Death.

	Unadjusted HR (95% CI)	Adjusted for pre-KT factorsa HR (95% CI)	Adjusted for additional peri-operative KT factorsb HR (95% CI)
Discharge	1.21 (1.06–1.39)*	1.13 (0.91–1.41)	1.13 (0.91–1.41)
Delayed graft function	1.00 (0.74–1.35)	1.13 (0.83–1.53)	1.04 (0.76–1.41)
Acute rejection	0.93 (0.58–1.48)	0.78 (0.49–1.26)	0.82 (0.51–1.31)
Death-censored graft failure	1.51 (1.01–2.27)*	1.40 (0.93–2.11)	1.50 (1.00–2.27)
All-cause graft failure	1.25 (0.89–1.77)	1.31 (0.92–1.86)	1.37 (0.96–1.95)
Death	0.87 (0.49–1.54)	1.16 (0.65–2.08)	1.15 (0.64–2.07)

^*p < 0.05

^aAdjusted for age, sex, race, education, body mass index, cause of kidney failure, peak panel reactive antigen, history of malignancy, diabetes, hypertension, calendar year of dialysis initiation

^bAdjusted for (1) + years on dialysis, human leukocyte antigen mismatch, cold ischemic time, ABO incompatibility, calendar year of transplant, type of induction, and donor factors (age, sex, race, deceased donor, expanded criteria donor

Association Between SHPT Treatment and THPT Across Subgroups

Association between SHPT treatment (parathyroidectomy versus cinacalcet) and THPT differed by number of years on dialysis (p_interaction=0.039) and by calendar year of transplant. Treatment with parathyroidectomy was associated with a lower risk of developing THPT (aHR=0.43, 95% CI: 0.24–0.79, p_interaction=0.039) for those who were on maintenance dialysis therapy for ≥3 years prior to KT (Figure 2). In contrast among those on maintenance dialysis for < 3 years there was no association between SHPT treatment and THPT (aHR=1.22, 95%CI: 0.57–2.59). Treatment with parathyroidectomy was also associated with a lower risk of developing THPT when KT occurred between 2014 and 2016 (aHR=0.26, 95% CI: 0.08–0.82, p_interaction=0.01). In contrast, it did not differ when KT occurred between 2007 and 2010 (aHR=1.41, 95% CI: 0.70–2.84) or between 2011 and 2013 (aHR=0.53, 95% CI: 0.25–2.12). Age at listing, sex, race, BMI, PRA, dialysis modality, and donor type did not modify the risk of developing THPT among patients treated with parathyroidectomy.

Figure 2.

Association Between Secondary Hyperparathyroidism (SHPT) Treatment (Parathyroidectomy vs Cinacalcet) and Development of Tertiary Hyperparathyroidism (THPT) in Patients with SHPT who Underwent Kidney Transplant between 2007–2017. THPT was defined as any claim of THPT or hypercalcemia 6-months posttransplant.

Hazard ratios were adjusted for age, sex, race, education, body mass index, cause of kidney failure, peak panel reactive antigen, history of malignancy, diabetes, hypertension, calendar year of dialysis initiation were shown.

Interaction P-value less than 0.05 indicates the association was significantly different across levels.

Sum of subgroups was not equal to the total (n=5094) due to missingness; Body mass index (n=28, 0.6%) and panel reactive antigen (n=1644, 32.3%).

DISCUSSION

In this national study of 5049 Medicare beneficiaries diagnosed with SHPT on maintenance dialysis therapy, patients who underwent parathyroidectomy as treatment for SHPT had a 44% lower risk of developing THPT post-KT than those who received cinacalcet. This risk reduction was greatest in patients on maintenance dialysis for 3 or more years and for those who underwent transplant between 2014–2016. These findings emphasize the need for both consideration of parathyroidectomy in patients with SHPT on dialysis for ≥3 years, as well as closer surveillance of KT recipients who were treated with cinacalcet prior to transplantation for the development of THPT.

Similar to our study, prior studies have also found an association between pre-KT cinacalcet use and the development of THPT, with higher PTH and calcium levels persisting up to 4 years post-KT, as well as a higher likelihood of requiring parathyroidectomy.^26–28 However, we built upon these findings and uniquely identified a group of patients who would derive the greatest benefit from surgical treatment. Parathyroidectomy offered greatest risk reduction for developing THPT in patients who were on maintenance dialysis therapy for 3 or more years compared to less than 3 years. Therefore, the use of cinacalcet for some patients may be postponing the necessary definitive treatment of hyperparathyroidism.²⁸

We also found that patients transplanted between 2007–2010 had higher risk of developing THPT and that this risk was significantly decreased in those transplanted later, between 2014–2016. Interestingly, this finding corresponds to the number of patients treated with parathyroidectomy during those respective time periods. Recipients transplanted between 2014–2016 were more likely to have been treated with parathyroidectomy for SHPT prior to transplant (53.9%) compared to those transplanted between 2007–2010 (16.7%). A possible explanation for a higher proportion of patients treated with parathyroidectomy between 2014–2016 may be the rise in median PTH values in kidney failure patients during this time frame despite the increased prescription of cinacalcet.^29,30 Also, the aggressiveness in treating patients at this time was much less than the later years of the study and practices have changed.

Similar to other studies, those who underwent parathyroidectomy for SHPT were more likely to be younger, African American, have a lower BMI, and have spent more years on dialysis, likely reflective of a select group of patients with refractory SHPT.^31–33 Our study additionally found that patients with any history of malignancy or diabetes were less likely to undergo parathyroidectomy, highlighting the possibility that parathyroidectomy was performed in patients who may have been better operative candidates. However, there was no difference seen in mortality between groups arguing against channeling bias.

Of the 5094 patients in our cohort, 727 (14.3%) developed THPT post-KT, of which 97.5% were treated with cinacalcet pre-KT and 127 (2.5%) underwent parathyroidectomy. This rate of THPT is lower than the rates reported in other recent studies of approximately 17–50% and likely reflects our study population being limited to KT recipients with SHPT treatment as opposed to including all KT recipients.^17,34,35 It may also be reflective of our definition of THPT, which was any claim of THPT or hypercalcemia 6 months post-KT. There is currently a lack of consensus regarding an exact definition for THPT. A study by Evenepoel and colleagues demonstrated that the most dramatic fall of parathyroid hormone occurs within the first 3 months after KT and therefore therapy should be instituted after 3 months.²⁰ Other studies have demonstrated that THPT should be treated at 6 months to 1 year after KT as parathyroid hormone is unlikely to normalize 1 year after KT.^18–20 Due to the ambiguity of definitions, we performed a sensitivity analysis with a definition of THPT at 1 year, and our conclusions remain unchanged.

Hypercalcemia of THPT promotes tubulointerstitial, vascular, and soft tissue calcifications. THPT is additionally associated with increased mortality, graft failure, cardiovascular complications, nephrocalcinosis, fractures, and other known symptoms of hyperparathyroidism.^{14–16,36,37} In our population the incidence rates, cases per 100 person-years of mortality, cardiovascular events, and fracture in patients with THPT as compared to those without was as follows: 2.9 versus 2.3 for death, 6.3 versus 6.7 for CVE, and 4.4 versus 8.2 for fractures. Prior studies have demonstrated that persistent hyperparathyroidism post-KT is associated with adverse outcomes. Among patients with optimal transplant function, elevated PTH level as early as 10 weeks posttransplant was associated with a composite outcome of cardiovascular events, graft loss, and all-cause mortality.³⁸ Elevated PTH 1-year posttransplant was associated with an increased risk of long-term death-censored graft failure.³⁷ Additionally, PTH ≥ 150 pg/mL 3 months posttransplant was associated with worse allograft function up to 3 years posttransplant in addition to increased risk for overall mortality and for death with a functioning graft.³⁹ In our study population, the 3-year cumulative incidences of death, fracture, and cardiovascular events were lower among recipients who underwent parathyroidectomy as compared to cinacalcet (5.8% versus 7.1% for death; 1.0% versus 1.8% for fracture; 16.3% versus 17.4% for cardiovascular events). However, we believe that the sequelae of THPT warrant in-depth investigation as part of a more comprehensive study including rigid time points and PTH levels. While those treated with cinacalcet pre-KT should be monitored closely for development of THPT and its sequelae, it is important to note that parathyroidectomy after transplant might be associated with an early decline in allograft.^40,41 Additionally, parathyroidectomy was found to be both less expensive and more cost-effective that treatment with cinacalcet at 7.25 ± 0.25 months on dialysis.⁴²

In contrast to prior studies, which demonstrated an association between pre-KT hypercalcemia or elevated PTH level and the development of DGF posttransplant, our study demonstrated no difference in developing DGF by treatment.^43,44 Several large, observational studies that demonstrated a 15–57% decrease in mortality for patients who underwent parathyroidectomy, our study did not identify a significant difference in AR, DCGF, ACGF, or mortality between the SHPT treatment groups.¹ While data regarding the use of cinacalcet and mortality has been conflicting, a large, prospective study did not show a survival benefit associated with cinacalcet.⁴ Unlike our study, Callender, et al demonstrated that parathyroidectomy pre-KT decreased post-KT graft failure when compared to all patients who did not undergo parathyroidectomy. However, the follow-up period for that study was only 1 year.¹⁶ Further studies have reported that parathyroidectomy performed prior to KT is associated with less allograft damage compared to after KT.^45,46 Unlike our study, these previous studies did not compare the outcomes of patients who underwent parathyroidectomy to those who received cinacalcet, but rather combined groups who were not treated with those who were treated with cinacalcet.^34,47,48 This difference in reference groups may account for the differing findings.

The main strengths of our study are large population size, patient-specific follow-up utilizing data from SRTR and Medicare claims, comparison of 2 treatment options for SHPT (parathyroidectomy and cinacalcet), and the availability of diagnostic and procedure codes for parathyroid-related disease as well as other comorbidities, allowing for the incorporation of both patient and transplant-related factors in our analysis. We also accounted for potential confounding by indication using inverse probability treatment weighting and excluded untreated SHPT patients to improve upon prior studies. This is the largest study to evaluate how treatment method of SHPT impacted post-KT adverse outcomes, specifically with the development of THPT, and identify which patients would benefit the most.

There are several limitations to this study in addition to those inherent to retrospective database studies including coding errors, missing values, confounding by indication, and selection bias. Because Medicare is not the primary healthcare payer for all patients on hemodialysis in the United States, results of this study may not be generalizable to the entire population of hemodialysis patients. However, Medicare is the largest single insurer for KT recipients as all patients requiring hemodialysis are eligible, and therefore Medicare coverage is often used as an inclusion criterion for studies of this population.^49,50 Another potential limitation is the possibility that some patients may have been on cinacalcet, which was covered by a second payer. However, all patients included in this study had Medicare claims for other medications, and therefore it was presumed that they would receive all of their prescriptions through Medicare. Although our models were adjusted for series of known confounders, the possibility that “unmeasured confounders” exist could not be excluded (Supplement). Finally, PTH values corresponding to each patient were not available in the database. Therefore, severity of SHPT, effectiveness and adherence to cinacalcet, and determining if parathyroidectomy was performed appropriately could not be factored into our analysis. Detailed information regarding number of parathyroid glands removed and thus adequacy of the operation could not be assessed. Because the pathophysiology of tertiary hyperparathyroidism involves autonomous proliferation of a parathyroid gland in the setting of polyclonal proliferation of all parathyroid glands, the consensus operation for these patients is a subtotal or total parathyroidectomy with autotransplantation. Therefore, patients who have an initial limited parathyroidectomy (removal of only 1–2 parathyroid glands) typically have persistent or recurrent THPT.^19,51 While we were able to assess complications of parathyroidectomy, we were unable to assess side effects and adherence to cinacalcet.

As the prevalence and severity of SHPT in the United States continue to rise, it is critical to understand risks and benefits of each treatment option.^29,30 This work will be instrumental to guide future decision-making for treatment of SHPT. Patients who undergo parathyroidectomy are less likely to develop THPT when compared to patients who received the oral calcimimetic, cinacalcet. This benefit is greatest in those who are on maintenance dialysis for 3 years or longer prior to KT. Given that THPT is associated with increased morbidity and mortality, these findings should be incorporated in shared decision-making and encourage consideration of parathyroidectomy for those on maintenance dialysis for 3 or more years. Additionally, transplant nephrologists and surgeons should closely follow parathyroid markers in KT recipients treated with cinacalcet for SHPT.

ABBREVIATIONS:

AR

Acute Rejection

ACGF

All-Cause Graft Failure

DCGF

Death-censored graft failure

DGF

Delayed graft function

HRSA

Health Resources and Services Administration

KDIGO

Kidney Diseases Improving Global Outcomes

KT

Kidney Transplant

LOS

Length of Stay

OPTN

Organ Procurement and Transplantation Network

PRA

Panel Reactive Antibody

SHPT

Secondary Hyperparathyroidism

SRTR

Scientific Registry of Transplant Recipients

THPT

Tertiary Hyperparathyroidism

USRDS

United States Renal Data System