'Early' Dialysis Start Based on eGFR is no Longer Appropriate

Authors and Disclosures

Steven Rosansky & Richard J. Glassock

Dorn Research Institute, William Jennings Bryan Dorn VA Hospital, 6439 Garners Ferry Road, Columbia, SC 29209, USA (S. Rosansky). 8 Bethany, Laguna Niguel, CA 92677, USA (R. J. Glassock).

Correspondence to
S. Rosansky sjrcra@yahoo.com
From Nature Reviews Nephrology
'Early' Dialysis Start Based on eGFR is no Longer Appropriate

Steven Rosansky; Richard J. Glassock

Posted: 12/06/2010; Nat Rev Nephrol. 2010;6(12):693-694. © 2010 Nature Publishing Group
Abstract and Introduction
Abstract

The randomized, controlled IDEAL study reports no survival advantage of 'early' dialysis initiation and data from this study support an estimated glomerular filtration rate of around 7 ml/min/1.73 m2 as a guideline for dialysis initiation. The results of the IDEAL study supplement data from eight observational studies involving over 1.2 million patients which showed a comorbidity-adjusted incremental survival disadvantage of 'early' dialysis initiation.
Introduction

The long-awaited Initiating Dialysis Early and Late (IDEAL) study, a randomized, controlled trial (RCT) conducted in Australia and New Zealand comparing the 'early' and 'late' initiation of dialysis (as assessed by renal function at the time of starting therapy) did not find 'early' initiation of dialysis to be associated with a survival advantage.[1] These results supplement evidence from eight of nine published observational studies that utilized data from over 1.2 million patients and showed a comorbidity-adjusted incremental survival disadvantage of 'early' dialysis initiation.[1–3] The main advantage of the IDEAL study is that the survivor bias, or 'healthy cohort effect', which uniformly confounds comorbidity-adjusted observational studies, is eliminated by RCT methodology. The results of the IDEAL study should finally curtail the disturbing trend towards 'early' dialysis start (Figure 1). In the USA, the fraction of patients initiating dialysis at an estimated glomerular filtration rate (eGFR) >10 ml/min/1.73 m2 (estimated using the Modification of Diet in Renal Disease [MDRD] equation) increased from 20% of new dialysis starters in 1996 to 52% of new dialysis starters in 2008; an even more dramatic rise occurred during this period in the percentage of patients starting dialysis with an eGFR ≥15 ml/min/1.73 m2 (Figure 1).[4] The IDEAL study used creatinine clearance estimated using the Cockroft–Gault equation as the measure of renal function; mean eGFR was also estimated using the MDRD equation with mean values being 7.2 ml/min/1.73 m2 in the 'late' start group and 9.0 ml/min/1.73 m2 in the 'early' start group.

Figure 1. Percentage of patients initiating dialysis 'early' in the USA during the years 1996–2008. Data from the United States Renal Data System.4 Abbreviation: eGFR, estimated glomerular filtration rate.

Over a median follow-up of 3.59 years, death occurred in 155 of 424 patients (36.6%) in the 'late' start group and 152 of 404 patients (37.6%) in the 'early' start group (P = 0.75). The IDEAL researchers found that none of the deaths occurring during the study period were related to uremia. As there was no significant survival difference between the two groups, this study would support waiting for an MDRD-estimated eGFR of around 7 ml/min/1.73 m2 to initiate dialysis (rather than initiating dialysis at an eGFR >10 ml/min/1.73 m2). The approach to initiating dialysis should evolve from what appears to be a trend toward pre-emptive dialytic therapy based on calculated eGFR levels to a 'wait and see' approach utilizing signs or symptoms of uremia rather than an arbitrary level of eGFR. An important byproduct of such an approach is that it would enable physicians to wait for the maturation of a permanent dialysis access placement, thereby avoiding the troublesome trend towards early mortality that seems to be associated with the use of indwelling catheters for vascular access and the disastrous consequences of attendant infection.[4]

The design of the IDEAL study allowed patients to be started on dialysis before they reached the assigned eGFR level (which was 10–15 ml/min/1.73 m2 in the early start group and 5–7 ml/min/1.73 m2 in the late start group). This study design resulted in 322 of the 424 randomized patients allocated to a 'late' start initiating dialysis at MDRD eGFR levels >7 ml/min/1.73 m2. Uremia—in 234 patients—was the major reason given for these protocol deviations. It would be useful to know the character of the uremic symptoms and signs that led nephrologists to deviate from the protocol design in these patients.

Classic uremic complications, including pericarditis, uremic encephalopathy, and oligoanuric chronic renal failure, occur only at very low levels of renal function. Australian guidelines for dialysis initiation were published in 2005 and recommended use of malnutrition as a uremic complication that justified dialysis initiation.[1] The guidelines also recommended dialysis initiation at an average urea and creatinine clearance of 6 ml/min/1.73 m2, regardless of symptomatology. The 2006 US National Kidney Foundation guidelines recommended dialysis initiation at eGFR <15 ml/min/1.73 m2.[1] The guidelines also endorsed initiating dialysis at eGFR >15 ml/min/1.73 m2 in patients with symptoms believed to be related to both their comorbidities and their level of residual renal function.[1] Between 2000 and 2006, when IDEAL study patients were recruited, US, Australian and other similar international guidelines may have influenced the IDEAL study physicians in their decision to initiate dialysis early in some patients. The resultant small 1.8 ml/min/1.73 m2 difference in eGFR between the two study groups decreased the ability to demonstrate whether 'early' dialysis initiation has a harmful impact, as had been suggested by many large observational studies.[2,3]

Since there is no evidence that early initiation of dialysis provides any survival benefit, and because there are multiple adverse comorbidity and quality-of-life issues related to dialysis treatment, further exploration of the issue of optimal timing of dialysis initiation is needed. Data on the reasons used by nephrologists to guide initiation of dialysis therapy would be particularly valuable. Symptoms related to renal failure are often subjective and do not necessarily correlate with indices of renal function. In one study, most patients commenced dialysis with at least three nonspecific symptoms, most commonly fatigue, nausea and anorexia.[5] Symptoms of 'uremia' might be more likely associated with comorbidities than with the level of renal function.[5] Serum albumin level is a comorbidity factor that predicts dialysis mortality and may be misused as evidence of 'malnutrition' and justification for dialysis initiation.[6] Although malnutrition is included and promoted as an indicator of uremia that is treatable by dialysis, the randomized, controlled HEMO study found this not to be the case.[7] In patients with a residual GFR <2 ml/min/1.73 m2, hemodialysis failed to influence the decline in measures of nutrition over the 3-year duration of the study.[7]

One issue that could not be adequately examined by the IDEAL study was the relationship between the rate of decline in residual renal function prior to the start of dialysis (the 'renal function trajectory'), and the decision to initiate dialysis. The median time from randomization to dialysis initiation was 1.8 months in the 'early' start group and 7.4 months in the 'late' start group. The crossovers meant that no evaluation could be made of the potential benefit of delaying dialysis in patients who maintain stable renal function and thus whose symptoms are less likely to be related to uremia and more likely to be related to comorbidities.

A second issue not addressed by the IDEAL study was the management of the only segment of the US dialysis incident population that is still growing—patients over 75 years of age.[4] Compared with younger patients, elderly patients may have a slower decline in residual renal function[8] ('renal function trajectory'), a shorter life expectancy independent of dialytic therapy, and more comorbidity-related symptoms that could possibly be interpreted as uremic in nature and result in dialysis initiation. An adequately powered RCT would be useful to determine if there is any clinically significant survival, morbidity, or quality-of-life benefit or hazard associated with 'early' dialysis initiation in the elderly population.

A third issue that the IDEAL study could not address was any differences between peritoneal dialysis and hemodialysis with respect to renal function at dialysis initiation and survival. Several studies have demonstrated that preservation of endogenous residual renal function is a strong positive predictor of survival and that residual renal function might be better preserved with peritoneal dialysis than with hemodialysis.[9]

Finally, the IDEAL study utilized rigorous multidisciplinary predialysis follow-up which can itself potentially decrease the risk of subsequent mortality and hospitalizations,[10] and emergency dialysis was largely avoided for the study participants. Since such close follow-up likely influences patient survival in a positive manner, in order to assign benefit to dialysis treatments, a third arm of an RCT such as IDEAL should include patients assigned to weekly or thrice-weekly programmed care but without dialysis treatment. However, the crossovers and immense difficulty of performing a study such as the IDEAL study make it unlikely that an RCT with such a third arm will be successfully executed.

It is hoped that the IDEAL study results, together with the results from the many observational studies on 'early' dialysis initiation, will help to reverse the epidemic of 'early' start of dialysis. Individuals involved in formulating US and international guidelines may want to consider an eGFR of 5–7 ml/min/1.73 m2 along with definitive uremia-related symptoms as the optimal point at which to initiate dialysis. Such a shift in recommendations might lead to considerable economic savings without adverse impact on patient survival, morbidity or quality of life.
[ CLOSE WINDOW ]
References

1. Cooper, B. A. et al. A randomized, controlled trial of early versus late initiation of dialysis. N. Engl. J. Med. 363, 606–619 (2010).
2. Hwang, S., Yang, W., Lin, M., Mau, L. & Hung-Chen, C. Impact of the clinical conditions at dialysis initiation on mortality in incident hemodialysis patients' national cohort study in Taiwan. Nephrol. Dial. Transplant. 25, 2616–2624 (2010).
3. Wright, S. et al. Timing of dialysis initiation and survival in ESRD. Clin. J. Am. Soc. Nephrol. doi: 10.2215/CJN.06230909.
4. National Kidney Foundation. United States Renal Data System. 2007 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Am. J. Kidney Dis. 51 (Suppl. 1), S1–S304 (2008).
5. Curtis, B. M. et al. Canadian survey of clinical status at dialysis initiation 1998–1999: a multicenter prospective survey. Clin. Nephrol. 58, 282–288 (2002).
6. Friedman, A. N. & Fadem, S. Z. Reassessment of albumin as a nutritional marker in kidney disease. J. Am. Soc. Nephrol. 21, 223–230 (2010).
7. Rocco, M. V. et al. The effect of dialysis dose and membrane flux on nutritional parameters on hemodialysis patients: Results of the HEMO Study. Kidney Int. 65, 2321–2334 (2004).
8. O'Hare, A. M. et al. Age affects outcomes in chronic kidney disease. J. Am. Soc. Nephrol. 18, 2758–2765 (2007).
9. Wang, A. Y. & Lai, K. N. The importance of residual renal function in dialysis patients. Kidney Int. 69, 1726–1732 (2006).
10. Wingard, R. L, Chan, K. E., Lazarus, M. & Hakim, R. M. The "right" of passage: surviving the first year of dialysis. Clin. J. Am. Soc. Nephrol. 4, S114–S120 (2009).

Competing interests
The authors declare no competing interests.

Nat Rev Nephrol. 2010;6(12):693-694. © 2010 Nature Publishing Group

http://www.medscape.com/viewarticle/733355?src=rss