Understanding and Communicating Medical Risks for Living Kidney Donors: A Matter of Perspective

Perioperative Risks

Early donor nephrectomy complications reported by centers to the Organ Procurement and Transplantation Network (OPTN) from 2000 to 2012 indicated the need for blood transfusion in 0.4%, vascular complications in 0.3%, and reoperation in 0.5%.⁷ These frequencies are substantially lower than estimates of 3%–6% for major complications and 18%–22% for minor complications identified in a prospective Norwegian donor registry and United States hospital records,^5,8 supporting a need to expand ascertainment of perioperative complications with other information sources.

Using data from the Nationwide Inpatient Sample, Schold et al. identified a sample of United States living kidney donors from 1998 to 2010 on the basis of the International Classification of Diseases, Ninth Revision diagnosis and procedure codes.⁹ Procedure-related complications were reported after 7.9% of donations, and were more common among men, Medicare beneficiaries, and those with hypertension. Limitations of this study include the lack of confirmation of donor status through patient level linkages to the OPTN registry, and use of weighting schemes to draw inferences for a “represented” sample of all United States donors on the basis of a stratified sample of 20% of acute care hospitalizations.¹⁰ A subsequent study integrated national United States donor registry data as a source of verified live donor status with administrative records from a consortium of 97 academic hospitals (2008–2012; n=14,964), and found that 16.8% of donors experienced any perioperative complication, most commonly gastrointestinal (4.4%), bleeding (3.0%), respiratory (2.5%), and surgical/anesthesia-related injuries (2.4%).¹¹ Major complications, defined as Clavien severity level 4 or 5, were reported in 2.5% of donors. After adjustment for demographic, clinical (including comorbidities), procedure, and center characteristics, factors associated with increased risk of major complications included black race, obesity, predonation hematologic disorders, psychiatric conditions, and robotic nephrectomy, whereas greater annual hospital volume predicted lower risk. The limitations of administrative database studies including possible coding biases highlight the need for prospective collection of granular clinical data on live donor perioperative outcomes.

The risk of death after donor nephrectomy is very low. Linkage of OPTN registry data for 80,347 living donors between 1994 and 2009 with the Social Security Death Master File showed a 90-day mortality estimate of 3.1 per 10,000 that did not change significantly over the 15-year study period.¹² Mortality was higher in men, black donors, and donors with hypertension. Unfortunately, the practice of nontransfixion techniques for arterial closure in the donor nephrectomy persists¹³ despite numerous reports about the mortality risk of clip dysfunction¹⁴ and a US Food and Drug Administration recall of these devices.¹⁵ We strongly believe that the use of nontransfixion techniques for renal artery closure puts donors at unnecessary risk of hemorrhage and death. Continued efforts to elucidate and prevent modifiable causes of perioperative complications in all donors are warranted to maximize safety of the donation procedure.

Longer-Term Mortality and Cardiovascular Disease

As the OPTN registry in the United States collects living donor follow-up information for only 2 years¹⁶ and long-term donor registries are uncommon internationally, inferences on longer-term donor morbidity and mortality were historically drawn from retrospective, single-center studies with comparison to general population-based survival estimates.^17–19 The conclusion that “kidney donors live longer” was popularized on the basis of general population comparisons,¹⁷ but with obvious confounding by the fact that living donors undergo medical evaluation and rigorous selection for good health at the time of donation.

Recent studies comparing longer-term mortality in donors to that of matched “healthy” persons who did not donate support similar survival, at least over approximately 10–18 years,^12,20 including one study of donors >55 years.²¹ In the aforementioned study of linked death records for >80,000 United States donors, the risk of death over a median 6.3 years of follow-up (maximum 12 years) was not different for kidney donors and demographically matched, healthy National Health and Nutrition Examination Survey (NHANES) III participants screened for baseline good health, including among subgroups stratified by race.¹²

Garg et al. compared 2028 living kidney donors in Ontario, Canada (1992–2009) with 20,280 healthy, demographically matched nondonors, and found that the risk of death or major cardiovascular events over a median 7 years follow-up (maximum 18 years) was lower in donors than in healthy nondonors (2.8 versus 4.1 events per 1000 person-years; adjusted hazard ratio [aHR], 0.66; 95% confidence interval [95% CI], 0.48 to 0.90).²⁰ The risk of death-censored major cardiovascular events was similar among donors and nondonors (1.7 versus 2.0 events per 1000 person-years).

One study with longer follow-up compared cardiovascular and all-cause mortality in 1901 kidney donors with a control group of 32,621 healthy, matched controls selected from a population-based survey in Norway (Nord-Trøndelag Health Study I).²² Mortality curves were similar for donors and nondonors over the first 15 years but subsequently diverged. At 25 years after donation, the cumulative all-cause mortality was approximately 18% among donors and 13% among healthy nondonors (aHR, 1.3; 95% CI, 1.1 to 1.5). Limitations of this study include differences in accrual periods and in baseline characteristics (including age) between the donors and nondonors.^23,24 Continued study is needed to assess the impact of donation on long-term survival in large, representative cohorts.

Limited data are available on the effect of donation on the pathophysiology of cardiovascular disease. The Chronic Renal Impairment in Birmingham (CRIB)-Donor study included 68 donors at two United Kingdom centers (2011–2014), of whom 90% were white, and prospectively examined changes in left ventricular (LV) mass and other cardiovascular disease surrogate markers at 12 months postdonation versus predonation, compared with changes in these parameters in healthy nondonors.²⁵ Among the findings, donors had larger increases in LV mass (+7±10 versus −3±8 g; P<0.001) and LV mass/volume ratio (+0.06±0.12 versus −0.01±0.09 g/ml; P<0.01), and decreased aortic distensibility; donors were also more likely than controls to develop detectable highly sensitive troponin T. The increase in LV mass among donors was independently associated with the magnitude of decrease in measured GFR. These pilot observations warrant replication efforts in larger cohorts with longer follow-up to better define the effects of donation on cardiovascular disease surrogates and clinical events.

Kidney Disease

Although 50% of functioning renal mass is removed during donor nephrectomy, compensatory hypertrophy in the remaining kidney returns GFR to approximately 70% of baseline within several weeks postdonation.^26,27 In a prospective study of 182 predominantly white (94.6%) kidney donors and paired control individuals considered eligible for kidney donation in the Assessing Long Term Outcomes in Living Kidney Donors (ALTOLD) study, GFR measured by iohexol clearance declined 0.36 ml/min per year in control individuals but increased 1.47 ml/min per year in donors during the period between 6 and 36 months of follow-up.^26,28 The trajectory of change in GFR beyond 36 months postdonation requires characterization in larger, demographically diverse samples.

The risk of ESRD after kidney donation does not exceed ESRD rates in the general population.^18,29,30 An analysis of living kidney donors in the United States between 1994 and 2003 quantified a postdonation ESRD rate of 0.134 per 1000 person-years over an average follow-up of 9.8 years, which was not higher than the ESRD rate in the general population.³¹ However, the general population is a limited comparison group given the robust medical evaluation that donors undertake. Recent studies comparing to individuals selected for baseline good health suggest that donation is associated with an increase in the risk of ESRD, although the risk increase is small and the absolute postdonation risk remains low. In comparing 1901 kidney donors with 32,621 healthy, demographically matched controls, Mjøen et al. reported that nine donors (0.47%) developed ESRD versus 22 healthy nondonors (0.07%) (aHR, 11.38; 95% CI, 4.37 to 29.63).²² On the basis of linking national ESRD reporting forms to United States donor registry for 96,217 donors and to records for healthy participants in NHANES III, the cumulative incidence of ESRD at 15 years was estimated as 30.8 per 10,000 in donors compared with 3.9 per 10,000 in matched, healthy nondonors (risk attributable to donation of 26.9 per 10,000).³² Considered by race, black living donors had the highest absolute incidence of ESRD as well as the highest risk attributable to donation.

On the basis of these findings, the likelihood of a small increase in ESRD risk should be discussed with donor candidates, and such counseling is endorsed by a recent AST LDCOP consensus statement.³ Efforts for risk stratification and mitigation in black donors warrants particular attention. Higher risks of ESRD and pre-endstage kidney complications in black compared with white donors³³ may relate to the incidence of hypertension and diabetes,^34,35 access to care and other environmental factors, and/or the distribution of kidney disease risk alleles such as apo-L1.^6,36

Integrated Risk Assessment

To date, the evaluation of living donor candidates has generally involved consideration of individual risk factors in isolation. For example, most centers exclude donors with a body mass index exceeding a predetermined threshold, usually between 30 kg/m² and 35 kg/m², without consideration of additional donor characteristics or risk factors. To develop an approach for assessing risk on the basis of simultaneous consideration of a profile of demographic and clinical characteristics, a recent Kidney Disease Improving Global Outcomes (KDIGO) work group formulating guidelines for the “Evaluation and Care of Living Kidney Donors” collaborated with the CKD-Prognosis Consortium. Risk associations from a meta-analysis of seven general population cohorts capturing data for nearly 5 million healthy persons were calibrated to the population-level incidence of ESRD and mortality in the United States.³⁷ The resulting models were used to project the estimated long-term incidence of ESRD (baseline risk, in the absence of donation) according to ten demographic and health characteristics, and programmed into an online risk tool.

Baseline risk projections were higher in the presence of a lower eGFR, higher albuminuria, hypertension, current or former smoking, diabetes, and obesity. In the model-based lifetime projections, the risk of ESRD in the absence of donation was highest among persons in the youngest age group, particularly among young black persons. In contrast, application of the tool may demonstrate that more older adults, even with some medical abnormalities, may be appropriate donor candidates with respect to their future ESRD risk. The online tool is available at: www.transplantmodels.com.

Further research is needed to quantify the effect of donation itself on lifetime risk according to donor characteristics and add other unmeasured risk factors (e.g., family history) to the online tool as data become available, to validate the results, and to extend the models to populations outside the United States and other racial groups. This study and the risk assessment tool advance a framework for donor evaluation centered on simultaneous consideration of many clinical factors relevant to ESRD risk, but application in the clinical setting requires clinician insight and interpretation. The KDIGO work group presents this framework as a starting point, and advocates strongly for continued efforts to improve the precision, tailoring, and generalizability of pre- and postdonation risk estimates.³⁸

Hypertension

Although it is well documented that BP rises with aging,³⁹ GFR reduction from kidney donation may accelerate the risk or progression of hypertension over time to a greater extent than expected from normal aging, possibly due to physiologic alterations (hyperfiltration in the remaining kidney, changes in vascular tone and renin-angiotensin-aldosterone regulation), and possibly with the additional artifact of heightened detection at donor follow-up.^40,41

Existing retrospective studies examining the effect of kidney donation on hypertension risk have been limited by short follow-up durations, high rates of loss to follow-up, and comparisons to unselected general rather than healthy populations, which may fail to capture donation-related effects. Use of antihypertensive medications was lower in a cohort of privately-insured donors compared with age- and sex-matched unscreened beneficiaries in the same insurance plan.⁴² In contrast, a systematic review including data for 5145 predominantly white donors estimated 6 mmHg higher weighted mean systolic BP and 4 mmHg higher weighted mean diastolic BP in donors compared with controls after an average of 7 years.⁴⁰ An administrative claims linkage study of 1278 (primarily white) living donors in Ontario, Canada followed for a mean of 6 years found a higher incidence of hypertension diagnoses (16.3% versus 11.9%; hazard ratio, 1.4; 95% CI, 1.2 to 1.7) among donors compared with matched controls who were screened for the absence of baseline comorbidities through administrative claims.⁴¹

In United States cohorts including black donors, higher rates of postdonation hypertension diagnoses and antihypertensive medication use have been reported in black compared with white donors.^34,35,42 Although these patterns parallel hypertension prevalence differences in the general population,³⁴ one small study found higher rates of postdonation hypertension among 103 black donors compared with race-matched “healthy” nondonors (41% versus 18% at an average of 6.8 years postdonation).⁴³ Notably, many donors in this study were unaware of their hypertension.

On the basis of these data, donor candidates should be counseled that donation may accelerate the rise in BP and need for antihypertensive treatment over that expected with normal aging, especially among black persons and those whose BP is high-normal before donation.

Gestational Hypertension

Two retrospective cohort studies, one from the United States and the other from Norway, reported an increased frequency of gestational hypertension and preeclampsia among groups of women with pregnancies after donation compared with groups of women with pregnancies before donation.^44,45 A caveat to interpreting these studies is that as women age, their risk of pregnancy complications increases, and some women became pregnant for the first time after donation.⁴⁶ Comparing the outcomes of predonation to postdonation pregnancies even with statistical adjustment may not clearly define the incremental risk attributable to donation. Studies on the basis of donors’ recollections may also be affected by inaccurate recall of past events.

To address these methodologic concerns, a retrospective cohort study was performed comparing 85 donors (131 pregnancies) in Ontario, Canada, with 510 healthy nondonors (788 pregnancies) who were screened for baseline health conditions to simulate donor selection.⁴⁷ Each donor was then matched to six healthy nondonors on characteristics that might be associated with the risk of gestational hypertension or preeclampsia. Over median follow-up of 11 years, gestational hypertension or preeclampsia was more common in donors compared with nondonors (11% versus 5%; odds ratio, 2.4; 95% CI, 1.2 to 5.0). However, there were no differences in rates of preterm birth, low birth weight, caesarean section, or postpartum hemorrhage.

For women of child-bearing potential considering donation, we agree with recommendations of a recent AST LDCOP consensus statement that counseling should include the possibility of increased risk of gestational hypertension and preeclampsia after donation compared with experience in otherwise similar healthy women.³ A balanced presentation, however, should also note that in the available studies, most women had uncomplicated pregnancies after kidney donation.

Gout

Kidney donors may be at higher risk for developing hyperuricemia and gout compared with equally healthy nondonors. Two small, prospective, controlled cohort studies reported higher uric levels in donors compared with healthy controls,^25,26 beginning at 6 months and persisting through the 36-month follow-up available for one of the cohorts.²⁸ In a study of 1988 living kidney donors from Ontario, Canada who were followed for a median of 8.4 years (maximum 20.8 years), living kidney donors were more likely to be diagnosed with gout compared with healthy, matched nondonors (3.5 versus 2.1 events per 1000 person-years; hazard ratio, 1.6; 95% CI, 1.2 to 2.1). At 8 years, the cumulative incidence of gout was 1.4% higher in donors compared with nondonors (3.4% versus 2.0%).⁴⁸

The risk of gout after kidney donation may vary on the basis of race. A study of 4650 kidney donors from the United States, including 13.1% black donors, found that, by 7 years, black donors were almost twice as likely to develop gout as white donors (4.4% versus 2.4%; aHR, 1.8; 95% CI, 1.0 to 3.2).⁴⁹ Postdonation gout risk also increased with older age at donation and was nearly three times higher in men. Compared with matched donors without gout, donors with gout had more frequent diagnoses of renal conditions including ARF, CKD, and other disorders of the kidney.

Metabolic Bone Disease

The effect of a modest decrease in GFR on the development of metabolic bone disease among kidney donors is not clear. One study showed that kidney donors had higher serum concentrations of fibroblast growth factor-23 and greater fractional excretion of inorganic phosphate compared with nondonor controls.⁵⁰ The ALTOLD prospective cohort study found a 23% higher parathyroid hormone concentration among 201 donors assessed at 6 months postdonation compared with levels among 198 healthy controls (52.7 versus 42.8 pg/ml, respectively)²⁶ that persisted at 36-month follow-up.²⁸ Similarly, the CRIB study of 68 donors at two United Kingdom centers reported larger increases in serum fibroblast growth factor-23 and parathyroid hormone levels compared with prospective changes in these parameters among healthy nondonors.²⁵ Despite these early differences in metabolic parameters, 2015 kidney donors in Ontario, Canada, experienced no increase in the risk of nontrauma-related upper- or lower-extremity fractures and no difference in prescriptions for bisphosphonates compared with healthy controls.⁵¹

Risk Communication

Education, risk communication, and assessment of a donor candidate’s knowledge and understanding of the possible outcomes of donation typically occurs through discussions with healthcare professionals. Under an ideal framework, donor candidates will be provided with individualized quantitative estimates of short-term and long-term risks from kidney donation whenever possible, including recognition of associated uncertainty, and in a manner that is easily understood.^52–54 If a donor candidate’s postdonation risk exceeds the transplant center’s acceptable threshold, risk is not acceptable for donation. If the donor candidate’s postdonation risk is below the transplant center’s acceptable threshold, then an informed, autonomous donor who understands risks may procced with donation. Optimal methods to assess understanding in living donor candidates are not well defined,^55,56 but general techniques for comprehension assessment may include use of “teach back,” in which patients are asked to “teach back” what they have learned during their visit.⁵⁷ An instrument for assessing comprehension during informed consent for living liver donation has been piloted,⁵⁸ and provides a model for developing similar instruments for comprehension assessment in living kidney donor candidates. Donor candidates should have adequate time to consider the information they are provided during the evaluation process, although the duration of adequate time is not well defined, and may vary according to donor characteristics. Some, but not all transplant centers, require donor candidates to exercise a minimal period for this adequate consideration, referred to as a “cooling-off” period.⁵⁹ Ongoing efforts to develop best practices for the disclosure of donor risks and verification of comprehension are important research priorities.⁶⁰ Future research efforts should seek to define what methods (e.g., verbal, written, and/or graphical) are most useful to impart donation-related risk information. These efforts should include how outcomes and any uncertainty in estimates should be expressed and variation in communication needs according to patient characteristics (e.g., age, ethnicity, health literacy) to better ensure that risk communication supports well informed decision-making and prepares donor candidates for all phases of donation. Recommendations for risk communication to patients developed in the general medicine and cancer literature offer a useful structure for formal study of tools and strategies among donor candidates (Table 3).⁶¹

Future Directions

Live donor transplantation is an important treatment option for kidney failure that benefits many patients and society. Responsibilities of care providers and researchers include an obligation to study and present risks and benefits for donors, and to support donor candidates in autonomous, well informed decision-making. The study and communication of donor risks must be grounded in recognition of the perspective of comparison, distinguishing comparisons within the donor population (e.g., outcome differences according to race), comparisons against general population experience, and most critically, assessment of donation-attributable risks through comparisons against “healthy” controls. Continued efforts are needed to strengthen the evidence base for counseling and selection, including longer follow-up in representative cohorts and well designed use of integrated secondary data sources,^51,52 and to advance tools for tailored risk prediction according to donor characteristics. Ideally, risk prediction tools should include comparison to similar healthy nondonors. At the present time, transplant centers and the organizations that regulate transplant practice should evaluate and disclose risks to the best of currently available knowledge, respect the donor’s autonomy, including autonomy to take risks (within a center’s/regulator’s upper bounds of acceptable risk), and embrace a long-term relationship with the donor, because some risks are uncertain or evolving.

Disclosures

K.L.L. and D.L.S. receive support from the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases, “Long-Term Health Outcomes After Live Kidney Donation in African Americans,” grant R01-DK096008.