#NephJC Chat
Tuesday, July, 23rd, 2024, 9 pm Eastern
Wednesday, July 24th, 2024, 9 pm Indian Standard Time, 3:30 pm GMT
JAMA. 2024 May 23:e248523. doi: 10.1001/jama.2024.8523. Online ahead of print.
Hypertension and Kidney Function After Living Kidney Donation
PMID: 38780499
Introduction
In the sophisticated maze of nephrology therapies, a living donor kidney transplant remains the best kidney replacement therapy for kidney failure. Living donor kidney transplants demonstrate increased survival when compared to deceased donors. Additionally, the number of deceased donors has never kept up with patients on the transplant waitlist. With a high mortality rate on dialysis (~50% at 5 years), there is a desperate need to increase kidney donation and expanding living donation can fulfill this need. The workup for living donor candidacy is very thorough, however, counseling donors on long-term outcomes is lacking due to the limitations of large, high quality studies.
According to the Global Observatory on Donation and Transplantation, 35 000 living kidney donations take place annually, or about 39% of total kidney transplants. However, living kidney donors and patients with chronic kidney disease (CKD) present a fascinating dichotomy. A recurring question is whether kidney donors should be categorized as having CKD, and assumed to have similar associated health risks as patients with CKD from other causes. The adaptive hyperfiltration that occurs in a donor’s remaining kidney is influenced by many factors (age, sex, race, and BMI), yet it remains unclear whether donation-related eGFR decline is accelerated compared to non-donor age-related changes. (Srinivas et al, Adv Chronic Kidney Dis, 2012, Glassock et al, Trans Am Clin Climatol Assoc, 2009, Zhou et al, Kidney Int, 2008)
Figure from Ferro CJ, Townend JN, Clin Kidney J, 2022
As current evidence is inconclusive, the post-transplant landscape is open for higher-quality, long-term prospective studies to help inform kidney donors of potential health risks. Thus far, large studies have fallen short of providing physicians with individualized risk profiles for potential donors. Ideally, physicians should be able to inform each patient of their perioperative and long-term risks of donating a kidney. The current study by Garg et al aimed to determine if kidney donors have more hypertension and worse kidney outcome compared to matched non-donor control patients.
The Study
Methods
This was a multinational (Canada and Australia), prospective cohort study conducted in 17 transplant centers, including 924 living kidney donors and 396 non-donors, that aimed to determine the risk of hypertension in healthy, normotensive adults who donate a kidney compared with healthy, normotensive non-donors with similar parameters of baseline health. Data regarding the rate of eGFR decline, albuminuria and self reported health related quality of life was also collected. Recruitment was from 2004-2014, including a pilot phase from 2004-2008, and follow-up period was till November 2021. The median follow-up duration was 7.3 years.
Study Population
The study recruited adults, both donors, and nondonors, without hypertension, proteinuria, hematuria, or significant obesity (BMI < 35kg/m²). Patients were excluded if they had a history of CKD, cancer, diabetes, cardiovascular disease, pulmonary disease, or any other process that would have excluded them as kidney donor candidates.
eTable 3. Study eligibility criteria for non-donors and the screening criteria used to define standard criteria living kidney donors. Garg et al, JAMA, 2024.
Data Collection
The participants were assessed through the collection of a standardized health questionnaire, measurements of BP (in clinic and home measurements), height/weight, and laboratory testing of proteinuria and serum creatinine. Non-donors were allotted a hypothetical nephrectomy date and followed up in a similar schedule as donors.
Twelve months after donation, donors and nondonors were asked to complete 12 to 18 home BP measurements, as well as complete laboratory testing with blood and urine samples. This process was repeated for the remainder of the study. Participants also completed mailed health questionnaires at 3 and 12 months after baseline and then annually.
Figure 1. Recruitment, follow-up, and retention of participants in the Living Kidney Donor Safety Study, from Garg et al, Can J Kidney Health Dis, 2022
Outcomes
The outcomes of this study included measurements of hypertension, kidney function, albuminuria, risk of death/kidney failure/MACE, self-reported quality of life, depression/anxiety, health behaviors, and tinnitus/sinusitis. Wait, what do tinnitus and sinusitis have to do with kidney donation? Well actually, nothing! This clever data point is a dummy outcome, and helped determine bias, read on to find out how.
Hypertension was defined as a participant-reported physician diagnosis of hypertension and use of antihypertensive medication, or an SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg based on the mean BP measurements at follow-up visits. The mean changes in SBP, DBP and mean arterial pressure (MAP) were examined over time, accounting for any use of antihypertensives. Participants with stage 1 hypertension (SBP of 130-139 mm Hg or DBP 80-89 mm Hg) at the baseline assessment were excluded. Given the known variability of intermittent blood pressure measurements in a clinical setting, this was an attempt to recruit patients without early stages of hypertension.
Kidney Function
The annual change in eGFR (based on the 2021 CKD-EPI equation) was assessed at 1, 2, and 4 years after donation (baseline). Low eGFR was classified as < 60, < 45, or < 30 mL/min/1.73 m² at every follow-up.
Albuminuria
Defined as a new ACR of ≥ 3 mg/mmol (≥30 mg/g) from a spot urine sample at any follow-up visit.
The composite endpoint of the incidence of hypertension, eGFR less than 60 ml/min/1.73m², and/or albuminuria examined the occurrence of 1, 2, or all 3 components.
Death, kidney failure (persistent eGFR <15 mL/min/1.73 m² or dialysis/kidney transplant), and major cardiovascular events (MACE: myocardial infarction, stroke, or major cardiovascular procedure) were all assessed from the annual survey and medical records. These outcomes were assessed individually and as a composite.
Self-reported quality of life over the previous 4 weeks was assessed using a 36-item short-form survey, where higher physical and mental health summary scores indicated a better quality of life. In addition, The Beck Depression Inventory and Beck Anxiety Inventory were used to assess symptoms of depression and anxiety over the previous week and month. Self-reported health behaviors were examined in participants enrolled after 2008, including a low-salt diet, smoking status, and physical activity.
Self-reported tinnitus (prespecified) and sinusitis (post hoc) were used as potential markers of self-reported bias as neither should be affected by nephrectomy; now it makes sense, doesn’t it? The difference between donors and non-donors would mark the propensity to recall and/or report health concerns.
Sample Size and Analytics
This study was designed to rule out a two-fold higher risk of hypertension in donors compared with non-donors. Effect size was chosen based on a survey of acceptable donor risk by donors, recipients, and transplant professionals (Young A et al, Kidney Int 2008). The expected median follow-up was 7.5 years, with the actual follow-up length of 7.3 years. For an overall event occurrence of 15% and a 5% loss of follow-up, > 900 standard criteria donors and 390 non-donors would provide more than 80% power to detect a hazard ratio of 1.6 (with a 2-sided α=0.05).
Inverse probability of treatment weighting on the propensity score was used to balance donors and non-donors on baseline characteristics, producing a larger weighted pseudo-sample of non-donors with a similar distribution of measured covariates as donors. Non-donors were weighted using average treatment effect in the treated weights (propensity score/(1-propensity score)), with donors receiving a weight of 1.
Hazard ratios were estimated using Cox proportional hazards regression. Unweighted adjusted analyses included covariates such as age, sex, MAP, BMI, family history of hypertension, and family history of kidney failure. Proportionality assumption was confirmed using the time-interaction test and graphing the weighted log (-log [survival functions]) vs log time. Composite outcome HR was estimated for hypertension, low eGFR, and albuminuria with death as a competing event.
Linear mixed-effect models were used to examine between-group differences in longitudinal changes in continuous variables, with random individual-specific intercepts and slopes. The absolute and percentage differences in eGFR were examined using linear regression. A post hoc sensitivity analysis of albuminuria was conducted, restricted to those with at least 2 occurrences in follow-up. Point estimates were shown with 95% CIs, without adjustment for multiple comparisons.
Funding
The study was funded by the Canadian Institutes of Health Research (CIHR), with additional partnership funding from Astellas Canada and Novartis. The sponsors did not influence the study design, data collection, analysis, interpretation, or manuscript preparation and approval.
Changes in the study protocol
The study protocol underwent several changes to enhance reliability and address practical changes. The cumulative incidence function was estimated only at 5 and 10 years due to insufficient 15-year data. MAP was used instead of both systolic and diastolic BP in the propensity score model to avoid correlation issues, and additional baseline characteristics (ethnicity and recent smoking) were included to better balance the groups. Health-related quality of life variables were also assessed at 3 years to ensure trend consistency. For HR estimation, zero-event outcomes led to a method using random event addition and iterative simulation. Finally, the study examined sustained eGFR decline <60ml/min by checking for occurrences at least twice or in consecutive visits.
eTable 1. Summary of changes to the published protocol. Garg et al, JAMA, 2024.
Results
Study Population
The study enrolled 1,042 living kidney donors before their surgery. Out of these, 924 were standard-criteria donors. At the same time, a group of 396 healthy non-donors was enrolled. 5% of donors and 10% of non-donors were lost from follow-up. Due to resource limitations, a 1:1 ratio of donors to non-donors couldn’t be achieved.
Figure 1. Participant Selection in a Study of Living Kidney Donation. Garg et al, JAMA, 2024.
The donors were 66% female, 6% Asian, 88% White, with 84% having a BMI under 30. The average age was 47 years, and the average SBP/DBP was 120/73 mmHg with a mean eGFR of 100 mL/min/1.73m². Most participants were either relatives or friends of kidney recipients. For donors 51% were genetically related to the recipient; 44% emotionally related; 6% non-directed. For non-donors, 57% are genetically or emotionally related to a recipient (20% genetically, 37% emotionally). Importantly, half of the donors and almost 1/4 of non-donors had family medical history of kidney failure.
Table 1. Baseline Characteristics Before and After Weighting. Garg et al, JAMA, 2024.
Table 1- continuation. Garg et al, JAMA, 2024.
Hypertension
Hypertension occurred in 161 donors and 158 non-donors, with an incidence of 2.7 and 2.5 events/ 100 person-years respectively (weighted HR 1.11, 95% CI 0.75-1.66).
Figure 2. Weight Mean Systolic and Diastolic Pressures. Garg et al, JAMA, 2024.
The result was consistent across sensitivity and subgroup analyses. There were no meaningful between-group differences for longitudinal changes in SBP, DBP or MAP.
Table 2. Risk of Hypertension and Other Outcomes in Donors and Non Donors. Garg et al, JAMA, 2024.
Estimated GFR
Donors and non-donors had a similar mean increase in SBP, attributed to aging (0.4 and 0.5 mm Hg respectively). The mean change in eGFR after donor nephrectomy was -32 mL/min/1.73m² but from 12 months after donation through end of follow up, the mean annualized change in donors was only 0.1 mL/min/1.73m² versus -1.2 mL/min/1.73m² in non-donors.
eTable 16. Median eGFT at pre-donation/baseline and years 1, 3, 5, 7 in donors versus non-donors. Garg et al, JAMA, 2024.
While no participants during the follow-up examination had an eGFR below 30 mL/min/1.73m², donors more often (47% donors vs 5% non-donors) had an eGFR between 30 to 60 mL/min/1.73m² (e-table 9, e-figure 8). Donors were also more represented in the 60-89 mL/min group as well and often seen early post-donation due to the loss of nephron mass.
eFigure 8. Weighted percentage of donors and nondonors in categories of estimated glomerular filtration rate (eGFR) (mL/min per 1.73m2) over time. Garg et al, JAMA, 2024.
Albuminuria
New albuminuria occurred in 132 donors and 95 non-donors, with an incidence of 2.4 and 1.7 events/100 person-years respectively (HR 1.46, 95% CI 0.97-2.21) as seen in eTable 9. There was no significant between group difference in ACR as a continuous variable (weighted geometric mean ratio 1.02, 95% CI, 0.88-1.19), see eTable 9, and eTable 19.
eTable 19. Post-hoc sensitivity analysis: albumin-to-creatinine ratio ≥ 3mg/mmol at least twice during follow-up. Garg et al, JAMA, 2024.
Composite outcomes
Hypertension, eGFR <60, and/or albuminuria occurred in 59% of donors and 29% of nondonors, with an incidence of 13.2 and 4.5 events/100 person-years respectively (weighted HR 2.87, 95% CI 2.23-3.68)-table 2.
The composite outcome (with all three components) occurred in 24 donors and 0 non-donors. Death, kidney failure and/ or MACE occurred in 27 donors and 15 non-donors (weighted HR 1.86, 95% CI 0.04-98.84).
eTable 9. Pre-specified sensitivity analyses of medical outcomes. Garg et al, JAMA, 2024.
There were 9 deaths in donors cohort, with an incidence of 0.1 events per 100 person-year versus 0 events in non-donor group (HR 7.11, 95% CI 0.93-54.14)- table 2. However, one single death was registered in the first year post-donation, but it wasn’t declared as perioperative death (e-table 7a,b).
eTable 7a. Reported cause of death, 7b- year of death. Garg et al, JAMA, 2024.
Health related quality of life
The weighted physical health summary scores were lower in donors than non-donors 3 months after donation/baseline, but it bounced back subsequently, and were similar in subsequent years. There were no clinically meaningful differences in anxiety and depression.
eFigure 14. Average SF-36 physical component summary (PCS) score in a longitudinal analysis with splines in donors and nondonors. Garg et al, JAMA, 2024.
Health behaviors and markers of self-report bias
Donors were also more likely to adopt a low-salt diet, and self-report tinnitus. There was no difference in self-reported sinusitis.
eTable 25. Potential markers of self-report bias in donors versus nondonor. Garg et al, JAMA, 2024.
Discussion
"The quality of mercy is not strained; it droppeth as the gentle rain from heaven." - William Shakespeare
The most precious gift that can be given is, arguably, a living organ to a person in need. Kidney transplantation remains the best modality of renal replacement therapy and there is an ever-increasing demand for organ donation. The safety of living donor nephrectomy is essential to the continued success, growth, and sustainability of the clinical practice of living donor kidney transplantation. This study, spanning seven years, informs us about the balance between the selfless act of kidney donation and the inherent risks therein.
This prospective study provided comprehensive insights into the health outcomes of living kidney donors compared to non-donors. Key findings indicate that both groups had comparable rates of developing hypertension (17%) and similar modest increases in mean systolic and diastolic blood pressure over the study duration. Donors experienced an initial drop in estimated glomerular filtration rate (eGFR) post-nephrectomy, but their eGFR decline rate slowed thereafter. At the final follow-up, the median eGFR was 67 mL/min/1.73 m² in donors versus 91 mL/min/1.73 m² in non-donors. Kidney donors generously give away roughly half of their nephrons. A lower eGFR results from the nephrectomy, but glomerular hypertrophy and hyperfiltration of the remaining kidney allow for compensatory recovery of 40% of lost eGFR. (Lenihan, et al, J Clin Invest, 2015) In other models of kidney disease, these physiologic changes can be associated with increased proteinuria, hypertension, and more rapid loss of kidney function. Kidney donors, however, are not typical patients with CKD and often don’t suffer from pre-existing hypertension, rapidly declining eGFR, proteinuria, RAAS activation, and arterial stiffness.
There are studies going back decades that have looked at the consequences of unilateral nephrectomy. In fact, a commonly cited retrospective study of World War II veterans, who suffered traumatic kidney loss during the war, did not show any signals of increased incidence of hypertension or mortality (versus healthy controls), nearly 45 years later! In addition, in veterans who underwent autopsy after death, there was no obvious increased glomerular sclerosis. (Narkum-Burgess et al, Kidney Int 1993) More recently, however, two landmark studies published in 2014 demonstrated an increased risk of ESKD in donors compared to matched, healthy controls. (Mjoen et al, Kidney Int 2014 and Muzaale et al, JAMA 2014) In Norway, Mjoen and colleagues compared the risk of ESKD in a cohort of 1901 donors to a cohort of 32,621 healthy non-donors over a median of 24.9 years and found an eleven-fold higher risk in donors (HR 11.38, 95%CI 4.37–29.6). Similarly, in the US, Muzaale and colleagues compared the risk of ESKD in a cohort of 96,217 donors to a cohort of 20,024 healthy non-donors and found a ten-fold higher risk (ESKD risk 15 years after donation was 30.8 per 10,000 versus 3.9 per 10,000 in non-donors; P < .001). It is important to note that although donors have a higher risk of ESKD compared to healthy non-donors, the absolute number of ESKD cases is quite small. Notably, some subgroups do have a higher long-term risk of ESKD including males, Black race/ethnicity, higher BMI, and pre-donation hypertensive donors (aHR 3.04; 95%CI: 1.28–7.22; P=.01). (Ammary et al, Clin J Am Soc Nephrol 2019) The current study is reassuring in that donors actually had a slower rate of eGFR decline than non-donor controls, but this cohort of donors would be considered low risk for ESKD based upon the above, previously known, risk factors. Prior studies have suggested a higher incidence of hypertension in kidney donors, however, the authors of the current study call into question the quality of that data based upon a 2016 review. (Slinin et al, Transplantation 2016) Concerns about the quality of evidence stemmed from outcome bias and significant loss of patients to follow-up. To combat these issues seen in prior studies the current multicenter prospective cohort study was designed to try to limit bias. Controls, without contraindication to donation, were specifically chosen. In addition, a statistical weighting technique was used to ensure balance between the donor and non-donor study groups. The authors also tried to limit loss to follow-up through regular contact and home visits. In the end, the lack of risk for development of hypertension seen in donors in this study agreed with a 2018 systematic review that only included more recent, high quality studies. (O’Keeffe et al Annals of Internal Medicine 2018)
There were several limitations to the current study. The disparity in sample sizes between donors and non-donors could have affected statistical efficiency, leading to wide and imprecise confidence intervals for some outcomes. In addition, the study's generalizability is also limited, as it focused on standard-criteria donors and predominantly White individuals with access to universal healthcare. The racial mix and access to healthcare may not reflect the experiences of donors with more complex medical backgrounds or from different regions.
This study was marked by several protocol changes and complex statistical adjustments. The authors did place a lot of effort and emphasis on keeping patients in the study and did an admirable job of having a very low dropout rate. In the end, the findings on hypertension and eGFR decline were not significantly different from prior recent studies. One final comment on the reporting of tinnitus and sinusitis on patient self-reporting bias. The study identified a higher reported incidence of tinnitus among kidney donors, which did not correlate with eGFR levels in either group, indicating no biological link between nephrectomy and tinnitus. There were no differences in the reporting of sinusitis. This suggests that donors may have a greater tendency to report somatic symptoms. Using objective measures when assessing risks attributable to donation is certainly the standard, as subjective self-reports can be misleading and biased. However, there are many measures (including quality of life) that can only be gathered by self-reporting on questionnaires, and use of symptoms unrelated to the procedure being studied is a clever way to estimate self-reporting bias.
Conclusion
In this study of living kidney donors, the authors did not find a higher risk of high blood pressure or loss of kidney function compared to non-donors. Although donors' kidney function dropped initially after surgery, the subsequent decline in eGFR was slower than that seen in non-donors at follow-up. These results support the safety of donating a kidney and can help reassure people considering living donation about potential risks. As such, knowledge of these long-term outcomes has the potential to increase living donation.
Summary prepared by
Manish Lalwani
Assistant Professor, Department of Nephrology
Christian Medical College,
Vellore, India
Sejal Lakhani
Internal Medicine Resident,
Lehigh Valley Health Network,
Allentown, PA
NSMC interns Class of 2024