Blog entry by Meguid El Nahas
An excellent editorial by Steven Rosansky and Richard Glassock review issues related to the estimation of CKD progression in renoprotection drug trials.
The authors examine the currently used parametres of CKD progression:
1. ESRD and Renal Replacement Therapy (RRT)
They note that RRT initiation as a hard point has its limitations in view of trends over the last decade in early initiation of RRT. They also note the non-GFR related alterations in serum creatinine levels including alterations in endogenous creatinine generation due to diet and sarcopenia.
the authors recommend a harder endpoint of fixed eGFR of 15ml/min/1.73m2.
Some would argue that mGFR of 10ml/min/1.73m2 is a more valid endpoint.
2. The decline in eGFR
using the decline in eGFR is fraught by the fact that many patients dont progress or progress in a non-linear fashion. Increasingly nephrologists have come to realise that CKD progression is not invariable with a significant percentage not progressing for long period of time and some progressing in a non-linear fashio whilst some even improving with time. Such non-progression or a non-linear progression RFT would make endpoints based on RFT unpredictable and difficult to interpret in a clinical trial.
They also note that some interventions such as Low protein diet (MDRD study) or ACE inhibition may lead to an initial acceleration of GFR decline before longer term stabilisation/renoprotection.
The FDA recently suggested a 30% decline and a 40% decline in GFR from baseline values as new surrogate end points in intervention studies. This on the assumption that the decline in GFR is linear.
This would be preferable to the current doubling of serum creatinine (50% decline in GFR).
3. Changes in RFT slopes:
The authors note that changes in RFT pattern in clinical trials over the duration of intervention studies usually, short term around 3 years of follow-up, may not consistently predict the progression to ESRD; Longer trials 5-10years may be required. Whether these are feasible, fundable or acceptable to Pharmaceutical sponsors in a hurry to market a new drug for CKD is questionable.
4. Patients selection:
RCTs and data analysis as well as showing intervention efficacy may benefit from a careful patients' selection favouring teh inclusion of those with faster rate of GFR decline and overt proteinuria; a faster rate of progression would require a smaller sample size and increased study power.
Also recommended is to have an observational run in phase to determine the nature and rate of the RFT and CKD progression; those with fast (-3-5ml/min/year) and linear negative (RFT) progression rate would also increase the power of the study.
The authors also touch on the most important issue in my mind, namely that fact that eGFR does NOT equate with meassured/true GFR; this was most dramatically shown in the BEAM and BEACON studies, where the impression of an improved eGFR reflected weight loss and anorexia thus changes in creatinine metabolism rather than improving renal function. Ultimately BEACON had to be stopped for toxicity of the compound (Bardoxolone) and excessive morbidity and mortality.
Also the reliability of eGFR/serum creatinine does NOT take into account the impact of renal tubular secretion of creatinine on the changes induced by an intervention on these parameters. Even ACE inhibition may affect tubular secretion of creatinine, a totally neglected variable in RAAS inhibition studies of CKD progression where GFR has seldom been measured????
Finally, when eGFR and mGFR were compared in progressive ADPKD, eGFR based on the MDRD and CKD-EPI equations was found wanting. eGFR changes poorly reflected changes in mGFR. Both formulas underestimated GFR changes by 50%.
mGFR would be preferable to eGFR, although the authors have argued that variability of a mGFR based on isotopically labelled iothalamate may be high, not withstanding the inconvenience to the patient of measuring GFR and the related cost and time requirements.
One can conclude from this excellent editorial, that to study the impact of interventions on CKD progression the ideal RCT would require:
1. Careful patients selection; preferably with fast progressors who are proteinuric. After all these are the patients who warrant intervention to stop their progression to ESRD.
2. A Run in phase to determine the nature of RFT in patients included in the trial with preference for those with fast and negative but also RFT/slopes.
3. Long follow-up as short studies relying on unpredictable (for long term outcomes/ESRD) endpoints can be misleading.
4. Validating CKD progression parameters by measured GFR as the gold standard.
Nephrologists need to think carefully and read this editorial before embarking on hastily designed and flawed clinical trials that yield inconclusive or wrongly conclusive results.