"The physician must...have two special objects in view with regard to disease, namely, to do good or to do no harm" (Hippocratic Corpus in Epidemics:Bk. I, Sect. 11)
With the benefit of hindsight, and
the 20/20 vision that it allows, the 2006 KDOQI anemia guidelines (and the
2007 update) did not achieve the goals that the National Kidney Foundation
anticipated. The
publication of the guidelines generated a swirl of controversy and accusations
that are nicely summarized by Dr. Julie Ingelfinger and others in articles and several letters in CJASN. After
congressional hearings, a New York Times editorial, an FDA Black Box, and a slew of negative publicity the KDOQI guidelines essentially became irrelevant.
Were lessons learnt from
the anemia guideline experience? A paper published by Palmer et al in JAMA on the KDIGO Metabolic Bone Disease (MBD) guidelines earlier this year, and a commentary on the Palmer paper by Uhlig in AJKD this month would suggest otherwise.
The MBD guideline work-group and KDIGO had good intent. KDIGO’s goal was to
educate nephrologists and other health care providers on how to manage
metabolic bone disease (MBD). In 2005, KDIGO sponsored a controversies
conference “Definition, Evaluation and Classification of Renal Osteodystrophy”. The reported consensus of the attendees at this conference was that a new
guideline on CKD-MBD was warranted. A Work Group was assembled to develop the
guideline. The Evidence Review Team at the Tufts Center supported the Work Group for Kidney Disease Guideline Development and Implementation. The Work
Group met on five separate occasions over a period of 2 years, reviewing
evidence and drafting guideline recommendations. The KDIGO Board reviewed a
preliminary draft, and ultimately the final document. The guideline was also
subjected to public review and comment. The finished product is scholarly and thorough.
The MBD guidelines were controversial soon after they were published in 2009. The slant towards pharmacologic treatment is a major issue. Treatment means more money for companies that sell phosphate
binders and vitamin D (and PTH assay kits). The initial version of the MBD guidelines developed by the work-group did not recommend treatment -- it took a more conservative approach. Dr. Katrin Uhlig who oversaw the evidence review team explains this in a recent AJKD commentary:
“In the draft submitted for public review in August 2008, the KDIGO CKD-MBD guideline work group refrained from making recommendations for phosphorus, calcium, and PTH targets given the absence of high-quality evidence. This was a remarkable shift from the prior 2003 KDOQI (Kidney Disease Outcomes Quality Initiative) guideline, which recommended specific targets for calcium, phosphorus, and PTH and graded the recommendations for phosphorus and PTH targets in dialysis patients as being based on evidence rather than opinion."
In other words, the initial plan was to not make suggested treatment recommendations based on targets, but following consultation with the board and others, the thrust changed towards treatment. Indeed, in the KDIGO publication titled: “Best Practices in CKD-MBD: A Focus on Phosphorus”, sponsored by Genzyme, even the “suggested” treatment of hyperphosphatemia that is in the guideline document is dropped to just “Treat Hyperphosphatemia”. In 2009, with the publication of TREAT, a placebo controlled study, evidence emerged that treatment of mild anemia was not beneficial and could be harmful. A similar scenario is possible with MBD. Why suggest treatment for MBD when treatment may not be necessary, or may even be harmful?
“In the draft submitted for public review in August 2008, the KDIGO CKD-MBD guideline work group refrained from making recommendations for phosphorus, calcium, and PTH targets given the absence of high-quality evidence. This was a remarkable shift from the prior 2003 KDOQI (Kidney Disease Outcomes Quality Initiative) guideline, which recommended specific targets for calcium, phosphorus, and PTH and graded the recommendations for phosphorus and PTH targets in dialysis patients as being based on evidence rather than opinion."
In other words, the initial plan was to not make suggested treatment recommendations based on targets, but following consultation with the board and others, the thrust changed towards treatment. Indeed, in the KDIGO publication titled: “Best Practices in CKD-MBD: A Focus on Phosphorus”, sponsored by Genzyme, even the “suggested” treatment of hyperphosphatemia that is in the guideline document is dropped to just “Treat Hyperphosphatemia”. In 2009, with the publication of TREAT, a placebo controlled study, evidence emerged that treatment of mild anemia was not beneficial and could be harmful. A similar scenario is possible with MBD. Why suggest treatment for MBD when treatment may not be necessary, or may even be harmful?
Besides the issue of insufficient evidence for treatment, modifications in the process of developing the guidelines did not help. Changes seemed to be driven by the board of directors rather than the work-group or the evidence review team. Dr. Katrin Uhlig (article in AJKD) explains:
"...the KDIGO board also encouraged the work
group to provide recommendations even if these were based largely on expert
judgment, as long as the group could achieve consensus and the underlying
uncertainty was described clearly.... Consequently,
the KDIGO CKD-MBD work group extensively revised the guideline document,
reconvened to achieve consensus on wording and grading, and finalized the
document with recommendations for targets.”
The anemia guidelines were criticized for their lack of
transparency; the KDIGO MBD guidelines seem to have the same problem. The KDIGO web-site only lists
the names of it’s supporters: Amgen, Abbott, AMAG, Bristol-Meyers Squibb,
Bayer, Centocor, Chugai, Fresenius, Kyowa-Kirin, Pharmacosmos, PUMC, Roche, and
Shire. The contributions of these companies are not stated. A key recommendation of a recent Institute of Medicine (IOM) report: "STANDARD 1: Establishing transparency. 1.1 The processes by which a CPG [clinical practice guideline] is developed and funded should be detailed explicitly and publicly accessible."
No randomized trials have demonstrated that treating
hyperphosphatemia or the calcium-phosphorus product or the PTH level with
existing treatment options reduces cardiovascular events or mortality. However,
these trials have been limited by the sample size and being underpowered for
mortality and cardiovascular outcomes. Observational studies have guided
therapy and have informed the KDIGO-MBD guidelines. In the meta-analysis
published in JAMA, Palmer
et al analyzed 47 cohort studies and 327, 644 patients. They conclude “the
evidentiary basis for a strong, consistent, and independent association between
serum levels of calcium and parathyroid hormone and the risk of death and
cardiovascular events in chronic kidney disease is poor”. They point out: “the
evidence for an association between serum levels of phosphorus, calcium, or
parathyroid hormone as risk factors for outcomes in individuals with chronic
kidney disease (at any stage) is currently insufficient to inform clinical
decision making, policy, or practice guidelines.”
Dr. Uhlig from the KDIGO Evidence Review group doesn't dispute the Palmer paper. Treatment of mineral abnormalities has not been shown to be beneficial.
Dr. Uhlig from the KDIGO Evidence Review group doesn't dispute the Palmer paper. Treatment of mineral abnormalities has not been shown to be beneficial.
So where are we now? The MBD guidelines suggested treatment that at best may do no good, and at worst may be harmful and/or costly. Like the anemia guidelines, the MBD guidelines are in danger of becoming irrelevant. That aside, with every questionable nephrology guideline, confidence erodes about the value of guidelines.
In light of this, Dr Singh, what are you recommending to your patients? What are you going to say, tomorrow, to your patients with a too high phosphorus x calcium product?
ReplyDeleteP.S. I'd love to chuck my phosphorus binders. I hate those things, and they're so expensive! And if they may be harming me and not helping me, then I'll just toss them, thank you very much!
ReplyDeleteDr Singh,
ReplyDeleteThis write up is quite revealing. However what would you recommend to Nephrologists and your Nephrology fellows (at Brigham)?
Excellent commentary, Dr. Singh. I recently gave a talk at the ASN on trials that needed to be done to advance CKD-MBD. I identified at least 5 phosphate-related treatment recommendations by KDIGO that lack trial evidence, including that hyperphosphatemia should be treated at early stages of CKD, and recommending expensive non-Ca binders over Ca-binders in many patients despite the DCOR trial results.
ReplyDeleteKDIGO also made recommendations for trials that need to be done: those trials would determine if the present KDIGO guideline recommendations are actually of value. Talk about putting the cart before the horse! If such trials are not done, the Drug Companies funding KDIGO guidelines get "expert" "consensus" their drugs should be used. If those trials are eventually done and show little or no value, the drug companies have still won because they had years of KDIGO guidelines promoting their drugs prior to trials showing lack of benefit. The KDOQI anemia guidelines recommending - then demanding! - all CKD patients have anemia treated to Hgb >11g/dL is a good example of KDOQI and KDIGO favoring their drug sponsors' goals.
The KDIGO system pointedly refuses to adopt the Institute of Medicine recommendations on handling conflicts of interest on guideline groups. The IOM recommends chairs have no conflicts, and those with conflicts be greatly limited or excluded from membership. Work group members with conflicts should also not participate in decisions where their conflicts are in play. None of that appears to be adopted by KDIGO.
Having said that, the draft KDIGO anemia guideline group reflects some diversity of viewpoints finally. It appears the Chairs are attempting to rein in unsupported guideline claims from the past. It remains to be seen if the viewpoint-biased opinions (lacking trial data) make it into the final KDIGO document anyway.
Lastly, you are correct the influence of these "expert" opinion-based guidelines is waning. This is in good measure due to the willingness of scientists such a you to challenge the guidelines that lack trial support, or worse, ignored contrary trial evidence.
What should doctors do? Consider the patients, and start prioritizing treatments. Challenge anyone who says you must get the phosphate less than 5.5 mg/dl, Hgb to above 11g/dL, and keep ferritin above 200 ng/ml but never give iron if ferritin is >500ng/ml. There are no trials showing these guidelines are better than alternatives. If you choose to follow KDOQI / KDIGO advice, recognize the benefits are unproven, the potential for harm is real, and curiously the costs are usually substantial. Don't believe me? Go read the CHOIR, Normal Hematocrit Trial, and TREAT results again, then read the 2006 KDOQI anemia guideline.
What would you recommend to your PATIENTS? I have my bi-monthly appt with my neph next week, so what should I be asking him? My phos was below 5.5 but he put me on binders, anyway...should I be challenging this? What would you say to a patient who comes to you with this blog post?
ReplyDeleteAre my binders actually harming me?
Do you want to take binders? They probably are of benefit in reducing development/worsening of hyperparathyroidism. They might even lower your risk of cardiac problems. Or not. They may just be expensive. Or worse, they may interfere with other medications being absorbed. Your doctor feels binders are best for you. Listen to your doctor, but if you have problems with binders, ask him if it is really worth it.
ReplyDeleteThe problem is not your doctor recommending binders, but Guidelines telling you and your doctor to use binders despite the lack of trials showing benefits. Guidelines like that drive up drug costs, allow companies to avoid proving efficacy and safety, and allows companies to promote drugs off-label - like binders in CKD. In my opinion, good guideline groups guard against being used like that, while bad guideline groups promote their willingness to do that. Such groups make millions profiting from promoting opinion-based guidelines to companies who want to avoid doing trials. Such guidelines are forceful opinions written by "experts, and reflect consensus" who just so happen agree with the viewpoints of (and be consultants to) the pharma companies that funded the guideline group, and may have funded many of the guideline group's executive board members. In reality, science is never consensus, and more trial results undermine the prevailing expert opinion, than support it. Yet KDOQI wrote "consensus" anemia guidelines that harmed patients, and actually conflicted with available evidence. This is also why, in my opinion, KDIGO clings to expert panels which have so many conflicts of interest. Pharma companies funding the guidelines need to be sure their views on treatment are on the workgroups, and also exclude people who knowledgeably oppose their views. If the science was so clear, smart scientists without conflicts would reach the same conclusions and write similar guidelines - so why are KDOQI and KDIGO afraid of conflict-free panels?
Meanwhile, these same standards completely ignore the PREPONDERANCE of evidence that much earlier application sodium bicarbonate and/or bicitra safely slows CKD, *without* worsening hypertension, in the overwhelming majority of patients. Unlike the phosphorus recommendations, this *is* supported by RCT's combined with large cohort studies. But, why should the KDIGO guidelines panels or their pharma funders have any interest in treatments that cost $20/year? There's no money in it. Besides, preventing ESRD is bad for business. The standards guiding nephrology have very little credibility, yet pass themselves off as "evidence-based" medicine.
ReplyDeleteWell, yay. I take a shedload of sodium bicarb each day. I guess I can feel good about THAT.
ReplyDeleteDr. Singh,
ReplyDeleteI agree that the conflict-of-interest issues you raise mean KDIGO, like all, standards should be taken with a grain of salt. So too though, should meta-analyses, which often just measure the inconsistency between between studies, with a mathematical tendency to wash-out even the real effects. Poorly designed studies with larger samples, can "cancel out" better designed studies with smaller samples, etc. If the majority of studies in your pool are poor, then this becomes your interpretation of everything, despite any few higher quality gems.
Thus, the entire methodology often creates more uncertainty than it resolves, merely serving to convince ourselves that we know nothing, so we may as well just throw up our hands.
Looking at more recent specific cohort studies, such as Kovesdy's VA study (doi: 10.1053/j.ajkd.2010.06.011), or the Japanese study by Konta (DOI: 10.1002/dat.20495), both of which found substantial risk reductions --on the order of 40% and greater, in both pre-dialysis and dialysis cohorts, respectively.
While it is true that these cohort studies fail to meet the "Holy Grail" of RCT, they do have virtues of long baselines and robust samples, combined with high statistical power for the observed effects. Is it *possible* the studies are confounded: Yes - but is it PROBABLE? Anyway, albeit less than perfect, it is at least EVIDENCE for benefit from binders.
More importantly, these data actually EXIST, unlike hypothetical imaginary RCT's of the future. In contrast, your assertion that binders "at worst may in fact be harmful" is at this point just *speculation*, for which there is NO real experimental evidence, (i.e, in the form of convincing studies actually showing *worse* outcomes on binders rather than without.) You might fairly make still make the case that binders are *useless*, but you have not made the case they are actually harmful (other than perhaps wasting money, if indeed useless.)
Given the choice between admittedly imperfect *evidence* vs. even weaker *speculation*, I will continue for now to prefer the evidence, since at least SOME of the recent studies *strongly* indicate a benefit, and few (if any) studies even weakly demonstrate actual harm.
This is not really a fair parallel with the EPO debacle, for which at least some clear evidence of potential harm DID exist--from the very beginning--and was subsequently swept under the rug and forgotten during the craze.