Primum Non Nocere: Is CKD Staging Appropriate in Living Kidney Transplants?
Abstract and Introduction
The aphorism of Primum non Nocere ('first do no harm') was introduced to guide physicians in making difficult and potentially hazardous decisions. The application of estimating equations for glomerular filtration rate (GFR) and classification schema for chronic kidney disease (CKD) has inadvertently led to 'labeling' of many living donors as having CKD postdonation. This review examines this issue and its possible consequences. Although complete long-term studies are lacking, it appears that the 'labeling' of such donors as having CKD postdonation is common but not associated with a major effect on morbidity or mortality.
For almost 150 years physicians have been guided by an aphorism of ancient origin to help guide their difficult and potentially hazardous decisions -- Primum non Nocere or 'first do no harm'.
It seems hardly possible that only about 50 years ago the removal of one kidney from an otherwise healthy donor (an identical twin) for the purpose of kidney transplantation challenged this precept. The wisdom of this decision has been amply verified by the hundreds of thousands of similar events that have occurred subsequently, to manifest benefit for the recipients and collectively with little documented harm to the donors. With the possible exception of a slight increase in the long-term risks of persistent abnormal proteinuria and/or hypertension and the very uncommon occurrence of postdonation death and/or end-stage renal failure living renal transplant donors have enjoyed a long and fruitful postdonation life.
However, recently a new twist on the risks of kidney donation has appeared. In 2002, the National Kidney Foundation (NKF) and the Kidney Disease Outcomes Quality Initiative (KDOQI) published its schema for the classification (and ultimately diagnosis and staging) of 'chronic kidney disease' (CKD). This schema defined CKD on the basis of a 'reduced' glomerular filtration rate (GFR), as estimated from serum creatinine-based formulas (eGFR) such as the Modification of Diet in Renal Disease (MDRD) equation with or without concomitant evidence of kidney damage (proteinuria, hematuria, abnormal imaging or abnormal kidney biopsy) persisting for 3 months or longer. This classification system has been widely adopted and is now used throughout the world to both diagnose CKD, to estimate its prevalence among community living adults in epidemiological studies, and even to initiate therapy. In a later statement from the Kidney Disease: Improving Global Outcomes (KDIGO) initiative, 'clinical judgment' was advised regarding the relevance of markers of kidney damage, other than albuminuria. However, the definition of kidney damage identified by abnormalities in imaging tests was retained by KDIGO. Presumably, this would mean that donors might be regarded as having 'kidney damage' due to the presence of a solitary kidney postdonation, but this is not specifically mentioned in the document, other than the recommendation for the use of 'clinical judgement'. Nonetheless, an 'unintended' consequence of these classification systems has been to frequently diagnose 'CKD' (usually stage 3) following kidney donation in otherwise healthy living donors. This has added to the already somewhat uncertain burden of a lifetime with a solitary kidney for a living donor. This postdonation 'CKD' is based on assumptions drawn from an entirely separate cohort of subjects, where the degrees of disturbances in other very important functions of the kidney, progression of disease and comorbidities may be far different. We posit that this postdonation 'CKD' is an artifact of the classification system that uses an absolute threshold for eGFR to define stage 3 CKD irrespective of the age and gender of the individual. According to the KDOQI 2002 schema and the 2005 KDIGO refinements anyone with a persisting eGFR of < 60 mL/min/1.73 m has stage 3 CKD even in the absence of corroborating evidence of kidney damage, such as abnormal albuminuria. As pointed out elsewhere, these defects in the KDOQI and KDIGO classification systems may inadvertently inflate the prevalence of CKD in epidemiological studies and lead to overdiagnosis of CKD, particularly in elderly, female subjects, due to the normal decline in GFR with age. Since stage 3 CKD is believed to represent a risk for complications, such as cardiovascular disease, such a postdonation diagnosis of 'CKD' (based on eGFR alone) could have very important consequences for the donor and might influence future events, including, for example, insurability, expensive diagnostic tests or the needless taking of medications.
Recent studies have clearly shown that a variable fraction of living donors postdonation will have eGFR values < 60 mL/min/1.73 m and that the 'risk' of such postdonation 'CKD' varies with the age of the donor at the time of uni-nephrectomy, the baseline predonation eGFR and perhaps also the gender and ancestry of the donor. Many programs reassess GFR postdonation routinely in donors (usually at 3 months to 1 year), but relatively few systematically follow the GFR for many years or decades following donation. Therefore, the identification of the 'risk' of postdonation 'CKD' is limited to the first year postdonation in most studies. The magnitude of post-uni-nephrectomy compensatory functional 'hypertrophy' in the remaining kidney will be a determinant of the stable value of GFR reached postdonation. For most circumstances this compensatory hypertrophy has reached its zenith within a few weeks and remains stable for at least a year postdonation. Compensatory hypertrophy of function may be more complete in males and in those younger than 30 years of age. Postdonation 'CKD' has been estimated to occur in between 6% to over 90% of donors depending on their age at the time of uninephrectomy and the methods used for determining (or estimating) the GFR. Many transplant programs require that living donors have a GFR (measured or estimated) of greater than 80 mL/min/1.73 m prior to kidney donation, in addition to having no detectable evidence of a renal disease. If a lower threshold of GFR were set for such qualification then the prevalence of postdonation 'CKD' would increase accordingly. Wan et al. found that 74% of donors with a measured GFR of > 80 mL/min/1.73 m at the time of donation developed an eGFR of < 60 mL/min/1.73 m (stage 3 CKD according to KDOQI and KDIGO) when reassessed 1 year following donation. Garg et al. carried out a comprehensive systematic review of the literature and found an average decline in GFR of 26% postdonation and a pooled prevalence of stage 3 CKD of about 12% in 5048 donors from 48 transplant programs in 22 countries. Parasuraman and Venkat found a prevalence of 59% for postdonation stage 3 CKD in 22 African-American donors and only 12.5% in 32 Caucasian donors, all of whom had predonation GFR values less than 100 mL/min/1.73 m predonation. The differences were attributed to a lower extent of 'compensatory functional hypertrophy' in African-Americans compared to Caucasian donors, but the average age and genders of the donors of different ancestry were not reported. In a preliminary report, Barri, Parker and Glassock examined GFR (measured and estimated) in 196 living donors (all of whom had a measured GFR greater than 80 mL/min/1.73 m predonation) before and at 3 months postdonation. According to the measured GFR (I Iothalamate clearance or iGFR) or estimated GFR (eGFR MDRD), the postdonation prevalence of stage 3 CKD was 10% in donors age < 30 years (n = 31) by iGFR and 6% for eGFR (MDRD). This prevalence of postdonation stage 3 CKD rose steadily with age of the donor to values of 91% for donors over age 60 years (n = 11) by iGFR and 100% for eGFR (MDRD). The overall prevalence of postdonation stage 3 CKD by iGFR was 53/196 or 27%. No difference in the prevalence of stage 3 CKD postdonation between males and females was observed.
Thus, it appears that stage 3 CKD as defined by KDOQI (2002) and KDIGO (2005) is common in live donors postdonation of single kidney, irrespective of how one interprets the presence of a solitary kidney as evidence of kidney damage. The major questions arising from these observations are: Does postdonation 'CKD' lead to (1) an increased future risk of progressive kidney disease or (2) to an enhanced future risk for cardiovascular events. In sum, does it represent a potential 'harm' to altruistic kidney donation, thus violating the modern and ancient aphorism of Primum non Nocere? Only a very long-term longitudinal (many decades in duration) and complete follow-up of a well-defined cohort of living donors stratified by varying levels of postdonation GFR can answer this question definitively. Ibrahim and colleagues from the University of Minnesota have very recently reported on a single-center study of the long-term consequences of kidney donation in 3698 individuals 3-45 years after uni-nephrectomy. The average patient survival postdonation was quite similar to that of the general population. Treated ESRD was observed in only 11 donors (180 per million persons per year), which was about two thirds the rate of treated ESRD seen in the general population. A subgroup of 255 donors (selected to be representative of the entire group of surviving donors) was examined 12.2 ± 9.2 years after uni-nephrectomy, 15% had an eGFR of < 60mL/min/1.73m, 32% had hypertension and 13% had abnormal proteinuria. Quality-of-life scores were excellent. Caveats regarding this study need to be emphasized. The donors were 98% Caucasian. The average age of the selected subset was only 41 years at the time of donation, and 57% of this subset was followed for less than 10 years after donation. Thus, the long-term risks of kidney donation in older and non-Caucasian donors cannot be reliably assessed by this study. Nevertheless, it does provide a strong measure of reassurance that for young, healthy, Caucasian kidney donors the long-term risks of uni-nephrectomy to life and health are minimal. Regrettably, such a study has not yet been conducted. Cross-sectional and retrospective analyses of varying postdonation follow-up duration, have provided some cautious reassurances that kidney donation in an otherwise healthy individual does not represent a major long-term hazard. Garg et al. found a very low prevalence of ESRD (< 0.2%) among former kidney donors followed for 1-25 years (average 7 years) -- a value is not very different from the prevalence of ESRD in the general population. The theoretical counterbalancing effects of a greater risk for intrafamilial occurrence of renal disease and the lowering of risk stemming from a comprehensive evaluation of normality at the time of uninephrectomy factor into the determination of postdonation risk of ESRD. In systematic review of the literature from 1966 to 2006, Young and coworkers noted that living donors studied up to 11 years postnephrectomy seldom develop the metabolic, endocrine, hematology or microinflammatory disturbances commonly associated with established CKD caused by parenchymal renal disease leading to ESRD. More worrisome is the prevalence of postdonation overt proteinuria and elevated blood pressure (increased relative to matched epidemiologic data in the nondonor population). Increased urinary albumin excretion rate above the normal range was found in 34% of 47 living donors followed for 12-31 years postdonation in a study by Saran et al. Despite these unsettling findings, the actual mortality rate among donors (all cause and cardiovascular related) does not appear to be appreciably altered in comparison to a matched group of nondonors, studied retrospectively. Garg and coworker retrospectively examined the 6.2 year risk of death or major CVD events in 1278 living donors compared to 6359 matched healthy adults and found no increase in the hazard ratios for death or CVD. There was a slight increase in the risk of hypertension (Hazard Ratio: 1.4; 95% confidence interval = 1.2-1.7) but this may have been a biased estimate since the donors were seen more frequently examined by their primary care physicians. So far, to our knowledge, no one has examined the long-term risk of cardiovascular events in cohorts of donors stratified by postdonation GFR and urinary albumin excretion rate. Both sustained elevations of systolic blood pressure and a persistent abnormal increase of urinary albumin excretion rate are well-established risk factors for cardiovascular events in the general population. It is very reassuring to note that the PREVEND study has recently shown no additional risk for CVD associated with an eGFR of 30-59 mL/min/1.73 m (stage 3 CKD) compared to an eGFR of > 60 mL/min/1.73 min the absence of concomitant proteinuria in a large study of community living adults in Groningen.
For the present, it might be safe to presume that the long-term risk for premature death from CVD in stage 3 CKD in the absence of proteinuria or an elevated blood pressure postdonation is negligible and should not interfere with longevity or the ability to obtain life or health insurance. Is it then appropriate to label such individual as having a 'chronic disease' -- stage 3 CKD in this instance? We think not. Such a label can have far-reaching 'harmful' effects -- such as unnecessary anxiety and worry, generating questions regarding insurability or employment, and needless focus on a laboratory value for a functional parameter without proven relevance to disease risk.
Nonetheless, the observation of GFR values into the stage 3 CKD range in a substantial proportion of living donors postdonation calls for better and longer-term prospective studies. Relatively few long-term, prospective studies have been performed in living donors and a large fraction of donors are 'lost-to-follow-up' in the limited number of long-term reports available. Providing all living donors with permanent (life-long) health and life insurance policies heavily subsidized by the government, with once a year visits to a qualified medical practioner required for annual renewal of the policies would help to resolve the difficulty in assuring complete long-term follow-up of donors. In addition, living donors should be immediately'exempted' from the provision of the KDOQI (2002) and KDIGO (2005) CKD classification systems if they have an eGFR of < 60 mL/min/1.73 m in the absence of other evidence of kidney damage (excluding a solitary kidney). Furthermore, we believe that and steps should be undertaken to revise these criteria to deal with the potential 'harms' created by inappropriate 'labeling' of donors postdonation as having an advanced stage (stage 3 of 5 stages) of a 'chronic kidney disease'. In our opinion, it is not likely that the use of new formulas for calculation the eGFR used for classification of CKD (such as the CKD-EPI or Mayo Clinic Formulas) will resolve the conundrum. The mechanism for a lowering of the GFR in these altruistic and heroic individuals is quite different from those who lower their GFR through progression of an intrinsic renal disease. These donors should not be penalized for their act of generosity.
CLICK HERE for subscription information about this journal.