Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

MEDICAL OPTIONS in Chronic Kidney Disease

Chronic Kidney Disease-Mineral Bone Disorder: The Kidney Disease Improving Global Outcomes Initiative

Dr. David Mendelssohn, University of Toronto, Ontario

Exploring the Rationale Behind the Need for Non-calcium-Containing Phosphate Binders

Dr. Anthony B. Hodsman, University of Western Ontario, London, Ontario

Building a Case for a Potent Non-calcium-Containing Phosphate Binder

Dr. Denis Ouimet, Université de Montréal, Quebec

Practical Lessons for the Practicing Nephrologist

Prof. Rob Horne, University of London, UK

CHRONIC KIDNEY DISEASE-MINERAL BONE DISORDER: THE KIDNEY DISEASE IMPROVING GLOBAL OUTCOMES INITIATIVE

Editorial Overview:

David Mendelssohn MD, FRCPC, FACC

Head, Division of Nephrology, Nephrolife: Complete Kidney Care Centre, Humber River Regional Hospital, Associate Professor of Medicine, University of Toronto, Toronto, Ontario

Clinical practice guidelines have been developed in Canada, the US, Europe, Britain and Australia on how to manage various aspects of chronic kidney disease (CKD) and associated abnormalities. Inevitably, the guidelines have been published at different stages in the evolution of CKD management and they are not always concordant in their recommendations.

In an effort to harmonize the different guidelines, the Kidney Disease: Improving Global Outcomes (KDIGO) was initiated in 2004 with a stated purpose of forming evidencebased clinical practice guidelines for the care of CKD patients. One of the first initiatives was to address a growing body of literature linking disorders in mineral metabolism and bone disease to significant morbidity and decreased quality of life associated with vascular calcification in CKD patients.

This included more extensive abnormalities than are encompassed in the term renal osteodystrophy, which KDIGO members felt should be reserved exclusively to define alterations in bone morphology associated with CKD. As Dr. Sharon Moe¹, Professor of Medicine, Indiana University School of Medicine, Indianapolis, reported in and KDIGO members concluded, manifestations of mineral and bone abnormalities are diverse and include biochemical abnormalities, increased bone fragility leading to fractures and extraskeletal calcification. These abnormalities are interrelated and after deliberation, the group felt this systemic disorder should be referred to as chronic kidney disease-mineral bone disorder (CKD-MBD), although the term renal osteodystrophy still applies to abnormalities of bone.

CKD-MBD may be manifested by one or a combination of laboratory abnormalities of calcium (Ca), phosphorus (P), parathyroid hormone (PTH) or vitamin D metabolism; bone turnover, mineralization, volume, linear growth or strength; and calcification of the vasculature or other soft tissues.

At the time of the consensus conference, KDIGO members noted that their next priority would be to develop clinical practice guidelines for the diagnosis, evaluation, prevention and treatment of CKD-MBD. As one of the reviewers, I received an advance copy of this draft. Unfortunately, the current document contains almost no recommendations, sets no targets for correction of mineral abnormalities, as our own guidelines have done, and it makes no recommendations on the use of the currently available phosphate binders. It may in part have been because the KDIGO initiative was global and since many of the new medications that have emerged over the past five to six years for the treatment of CKD-MBD are expensive, global cost considerations had to be taken into account.

What is CKD-MBD?

As the kidneys fail, they are unable to excrete the normal daily load of P. As it accumulates, there is a reciprocal reduction in serum Ca and in response, the parathyroid gland is activated. In addition, abnormalities in vitamin D metabolism develop such that the renal-dependent final step, conversion of vitamin D to its active form is attenuated so patients do not make enough activated vitamin D. A deficiency in active vitamin D is recognized as a major contributor to the development of secondary hyperparathyroidism.

Elevated PTH levels are strongly implicated in the development of soft tissue and vascular calcification. Abnormalities of PTH are also a significant contributor to abnormalities in bone that can manifest as part of CKD-MBD. Extremes in bone turnover that can occur in CKD patients also gravely affect bone fragility and the risk for subsequent fracture are well-recognized complications of CKD. For example, as Dr. Jorge Cannata-Andia² reminds us, high bone turnover, which occurs in secondary hyperparathyroidism, as well as low bone turnover or adynamic bone disease can increase abnormal Ca-P deposition in soft tissues, vessels and cardiac valves, and both have been associated with vascular calcification and increased mortality risk.

The last component of CKD-MBD is vascular calcification (Figure 1). Age and duration of dialysis are uniformly accepted as risk factors for coronary artery calcification, but hyperphosphatemia, elevated Ca x P product and excessive Ca load from phosphate binders have all been identified as additional risk factors in most (though not all) studies. Of all of the mineral abnormalities that may be manifest in CKD-MBD, hyperphosphatemia is likely the most important contributor to vascular calcification and is considered by many to be a major nontraditional risk factor for cardiovascular (CV) disease in CKD patients.

Figure 1.

Why elevated P levels? Because in the presence of either high P or high serum Ca concentrations or both, vascular smooth muscle cells can differentiate into osteoblastlike cells, triggering signals that can induce mineralization and which leads to vascular calcification.³ A number of studies have implicated hyperphosphatemia as an independent risk factor for mortality—mostly from CV causes—in dialysis and even in predialysis patients, the latter carried out by Kestenbaum et al.⁴ Done retrospectively, a total of 95,619 veterans were screened, out of which 7021 participants met their criteria for CKD, 3490 of whom had had their serum phosphate levels measured during the previous 18 months.

After adjusting for all confounders, investigators found a serum phosphate level >1.13 mmol/L (>3.5 mg/dL) was associated with a significantly increased risk of death, the risk increasing linearly with each subsequent 0.16 mmol/L (0.5 mg/dL) increase in serum phosphate levels. Block et al.⁵ in turn analyzed data on 40,538 hemodialysis patients with at least one serum P and Ca determination during the last three months of 1997. After adjustment for case mix and laboratory variables, investigators found that serum P concentrations >0.16 mmol/L (>0.5 mg/dL) were associated with an increased relative risk of death, that risk more than doubling at the highest levels of serum P greater than 2.90 mmol/L (=9.0 mg/dL). In this study, both higher adjusted serum Ca concentrations as well as moderate to severe hyperparathyroidism (PTH =63.18 ng/L or =600 pg/mL) were both associated with an increase in the relative risk
ure 2).

Figure 2.

<img2613|center>

More recently, the DOPPS (Dialysis Outcomes and Practice Patterns Study) provided relevant and current insights about the impact of disorders in mineral metabolism across some 10 years during which data were collected in 12 participating countries. As reported by Tentori et al.⁶ investigators identified mortality risk based on different levels of serum Ca, albumin-corrected Ca, P and PTH levels in 25,588 patients on hemodialysis for longer than 180 days. Lower Ca and P levels were observed in the most recent phase of DOPPS (2005-2007)—perhaps a reflection of improved mineral disorder management and newer therapies. Survival models identified the lowest mortality risk for patients with serum Ca between 2.15 to 2.50 mmol/L (8.6 and 10.0 mg/dL), an albumin-corrected Ca of between 1.9 to 2.3 mmol/L (7.6 and 9.5 mg/dL), a serum P of between 1.16 to 1.61 mmol/L (3.6 and 5.0 mg/dL) and a PTH of between 10.63 and 31.59 ng/L (101 and 300 pg/mL).

In contrast, the greatest mortality risk was found among patients with a serum Ca or an albumin-corrected Ca level >2.50 mmol/L (>10.0 mg/dL), a serum P level >2.26 mmol/L (>7.0 mg/dL) and a PTH level >63.18 ng/L (>600 pg/mL) as well as in patients with combinations of high-risk levels of each of these three end points.

The HEMO (Hemodialysis) study by Wald et al.⁷ was actually a randomized controlled trial designed to determine whether increased dialysis dose or the use of high-flux dialysis membranes improved outcomes in long-term hemodialysis patients. Nevertheless, results lend further compelling support linking disorder mineral metabolism and adverse outcomes. Specifically, Wald et al. again found that serum P levels >1.94 mmol/L (>6 mg/dL) were independently associated with about a 25% increase in mortality risk compared with patients whose P levels were between 1.32 and 1.61 mmol/L (4.1 and 5.0 mg/dL). Similar to long-term DOPP analysis, a serum Ca in excess of 2.74 mmol/L (11 mg/dL) independently increased the risk of death by 56% compared with a reference level of 2.27 to 2.50 mmol/L (9.1 to 10 mg/dL)—findings that suggest that persistently high levels of serum Ca over time are particularly dangerous, as study authors pointed out.

A Ca x P product >4 mmol2/L2 (>50 ng2/dL2) was also independently associated with a heightened mortality risk but albumincorrected serum Ca levels were associated with adverse outcomes only when levels exceeded 2.74 mmol/L (11 mg/dL) and only when this parameter was evaluated as a time-varying or cumulative time-varying covariate.

This evidence serves to underscore the importance of achieving better P control than the majority of CKD patients are now able to attain, and it also argues for more effective phosphate binders as the ones widely used today are clearly unable to offer good phosphate control in far too many patients.

Calcium Overload

Much of the blame for the positive Ca balance CKD patients frequently experience is levelled at the large doses of Ca-containing phosphate binders they require to control P levels but we should not forget that the administration of large doses of vitamin D sterols to treat secondary hyperparathyroidism also contribute to episodes of both hypercalcemia and hyperphosphatemia, both of which serve to aggravate vascular calcification, according to Goodman et al.⁸ In fact, as Goodman pointed out, the total body Ca balance can be quite positive in dialysis patients without overt increases in serum Ca concentrations— suggesting that the absence of hypercalcemia does not guarantee patients still are not retaining excess amounts of Ca.

An on-line précis of a study by Danese et al.9 from the American Society of Nephrology showed that consistent control of mineral and bone disorders in 24,803 hemodialysis patients was associated with a markedly reduced risk of death. Specifically, patients who achieved KDOQI targets for PTH, serum Ca and serum P had a 51% lower risk of death compared to patients who achieved none of these targets. For patients who met only one KDOQI target, the risk of death was 35% to 39% higher compared to those who achieved all three targets and the risk was 15% to 21% higher for patients meeting any two KDOQI targets. The same study showed that the longer a patient met one or more targets, the better.

At the same time, we need to recognize that the mineral metabolism story in CKD patients is only a small part of what should be an overall CV risk reduction strategy from early CKD onwards. In order to make a real difference in patient outcomes, we need to have physicians refer patients to nephrologists early on in their disease and not wait until patients are ready for dialysis. Patients with early CKD should also be targeted with aggressive and global CV risk factor reduction strategies aimed at optimizing blood pressure, lipids and blood sugar control and we should be using medications known to have a protective effect on the kidney and the heart. We should also not forget that anemia is a common complication of early and late CKD and treat it as well, and manage mineral disorders including hyperphosphatemia, hypercalcemia and elevated PTH with appropriate medications that hopefully do no harm.

Once patients require dialysis, alternative regimens, including daily nocturnal hemodialysis or short-duration hemodialysis carried out six days a week, both manage excessive P concentrations much more effectively than our current practice of dialyzing patients three times a week. Regrettably, there is not yet a wealth of evidence to suggest that if we controlled CKD-MBD perfectly, we would significantly improve CV risk and reduce mortality in our patients with CKD.

Yet with the report from Danese et al.9, it certainly appears as if optimal control of the key contributors to early mortality in CKD— namely hyperphosphatemia, hypercalcemia and elevated PTH—does significantly reduce mortality risk in dialysis patients. With that study giving purpose to our efforts, we can collectively hope that better control of CKDMBD will improve survival in dialysis patients and with improved survival, perhaps extend the window of opportunity for transplantation, the optimal treatment for patients with endstage renal disease.

Summary

It is now widely felt that disorders of mineral and bone metabolism, now referred to as CKDMBD, are at least partly responsible for the increased prevalence of CV disease in CKD patients. Among the most powerful influences in this systemic disorder are elevated serum P levels. Now that studies have implicated the prolonged presence of hyperphosphatemia, hypercalcemia and elevated PTH as significant drivers of mortality risk in dialysis patients, there is a biologically plausible argument that improved control of CKD-MBD should reduce that risk and extend the lives of CKD patients. At the same time, there is an urgent need to intervene early and aggressively as soon as a patient is diagnosed with CKD and not wait until they require dialysis. Efforts are now underway in Canada to extend access to care through multidisciplinary predialysis clinics and hopefully optimize care at an early stage when it stands to do the most good.

References:

1. Moe S. Medscape Nephrology 2006;3(2), posted 11/30/2006.

2. Cannata-Andia J. Medscape Nephrology, posted 05/14/2007.

3. Jono et al. Circ Res 2000;87:E10-E17.

4. Kestenbaum et al. J Am Soc Nephrol 2005;16:520-8.

5. Brock et al. J Am Soc Nephrol 2004;15:2208-18.

6. Tentori et al. Am J Kidney Dis 2008;52(3):519-30.

7. Wald et al. Am J Kidney Dis 2008;52(3)531-40.

8. Goodman et al. Am J Kidney Dis 2004;43:572-9.

9. Danese et al. Clin J Am Soc Nephrol 2008;3(5):1423-9.

EXPLORING THE RATIONALE BEHIND THE NEED FOR NON-CALCIUM-CONTAINING PHOSPHATE BINDERS

Editorial Overview:

Anthony B. Hodsman, MD, FRCPC

Professor, Department of Medicine, Division of Nephrology, University Hospital, University of Western Ontario, London, Ontario

From my point of view, there is little medical ground for debate between calcium and non-calcium containing phosphate binders. If non-calcium containing binders were more readily accessible, most nephrologists would use them rather than a calcium-containing binder unless there was a good indication not to. The simple reason is that calciumbased phosphate binders such as calcium carbonate, widely used in Canada today, are a very inefficient means of achieving optimized control of serum phosphate. These binders were originally recruited as phosphate chelators to avoid the use of aluminum-containing binders, which are toxic to brain and bone in a small proportion of dialysis patients.

As many uremic patients are hypocalcemic, calcium (Ca)-based binders also appeared to make good empirical sense. Although the fractional absorption of Ca from insoluble Ca compounds is very small, small corrections of the serum Ca would have the added benefit of suppressing secondary hyperparathyroidism. However, most patients require large amounts of Ca salts to adequately bind dietary phosphate; because Ca-based binders are relatively inefficient, significant amounts of excess Ca are absorbed and subsequently deposited in soft tissue. Most dialysis patients require 4.5 to 6.5 g of elemental Ca to reduce serum phosphate to the target levels of 1.13 to 1.78 mmol/L (3.5 to 5.5 mg/dL) recommended by National Kidney Found
s Quality Initiative (K/DOQI) guidelines (Figure 1).

Figure 1.

<img2614|center>

Furthermore, K/DOQI recognizes that excessive consumption of Ca in any form is detrimental to the bone and mineral health of dialysis patients. Consequently, they recommend physicians limi t elemental Ca intake to approximately 1.5 g/day. Unfortunately, this recommendation does not make good clinical sense. Several provincial formularies now cover the higher cost of non-calcium based binders, but only in the face of hypercalcemia as a manifestation of demonstrated toxicity. The only alternatives on the open formularies are aluminum-based (approximately $0.10 to $0.12 per 500-mg capsule of aluminum hydroxide in Ontario).

Implications of Using Calcium vs. Non-calcium-Based Phosphate Binders

Here we review the implications of using Ca vs. non-Ca-based phosphate binders to control hyperphosphatemia in what constitutes the majority of patients with chronic kidney disease (CKD) stage 5. It is no secret that the majority of CKD patients die from cardiovascular (CV) disease, yet traditional CV disease risk factors, including elevated cholesterol levels, elevated blood pressure and smoking, do not fully explain this excessive risk.¹

A series of studies have implicated high serum P levels in the development of atherosclerosis, at least in part because phosphorous is a direct stimulant of parathyroid hormone (PTH) and parathyroid hyperplasia. Prolonged hyperphosphatemia leads not only to secondary hyperparathyroidism and elevated PTH levels, it also fosters the development of widespread soft tissue and vascular calcification due again in part to an increase in Ca-phosphate (Ca x P) product. (It is important to note that whatever other mineral abnormalities co-exist, soft tissue calcification is generally not seen unless hyperparathyroidism is present). Elevated P levels can also have a direct calcifying effect on vascular smooth muscle cells, cementing the pathway to vascular calcification. At the same time, parathyroid hyperplasia and elevated PTH levels contribute to high-turnover bone disease and other adverse consequences of excess PTH.

Serum P levels in excess of 2.10 mmol/L (6.5 mg/dL)—and in many series in excess of 1.94 mmol/L (6.0 mg/dL)—have been directly associated with increased overall and CV disease mortality in hemodialysis patients. There is even evidence that mild hyperphosphatemia, defined as serum P of 1.62 to 2.10 mmol/L (5.01 to 6.5 mg/dL), may be an independent risk factor for mortality in the same patient group. In one study by Rodríguez-Benot et al.², a serum P level in excess of 1.61 mmol/L (5.0 mg/dL)—the upper level of the established normal range of serum phosphate concentrations—was independently associated with a twofold increased risk of death. Indeed, research has shown that for each mg/dL increase in mean serum P levels, the relative risk of death increases by 6%.³

Calcium Overload: A High Risk for Coronary Events

Compounding the issue is exposure to Ca in the dialysate. As pointed out in an editorial by Dr. Bryan Kestenbaum, University of Washington School of Medicine, Seattle, end-stage renal disease (ESRD) represents a state of altered Ca homeostasis, where skeletal mineralization is impaired and urinary Ca excretion absent. With no means to remove excess Ca from the body, dialysis patients are at risk for significant Ca overload due to both excessive Ca intake and exposure to high Ca concentrations in the dialysate.

This might be of mere academic interest were it not for studies which suggest that patients with very high Ca scores (=1000) on screening by electron beam tomography (EBT) are at significant risk for coronary events even when they do not have CKD. As reported by Wayhs et al.⁵, hard clinical end points, defined as myocardial infarction (MI) or coronary death, were prospectively compared between 98 asymptomatic individuals who underwent EBT and historical controls who had severe abnormalities on myocardial perfusion imaging.

At an average follow-up of 17 months, 36% of this asymptomatic cohort experienced a hard clinical end point, and the likelihood that they would do so was strongly weighted in favour of those who had higher initial Ca scores. The annualized event rate in this cohort of asymptomatic subjects with a Ca score of at least 1000 was also significantly greater than for historical controls with severe perfusion abnormalities at 25% vs. 7.4%, respectively (P<0.0001).

Raggi et al.⁶ in turn compared dialysis patients with and without clinical evidence of atherosclerotic vascular disease and determined correlates of the extent of vascular and valvular calcification using validated statistical techniques. In this cross-sectional analysis of 205 maintenance hemodialysis patients, baseline EBT studies revealed a median coronary artery Ca score of 595—values consistent with a high risk of coronary artery disease (CAD) in the general population, as the authors observed. These scores were directly and significantly related to the prevalence of both MI and angina (P<0.0001 for both end points), while the aortic Ca score was directly related to the prevalence of claudication (P=0.001) and aortic aneurysm (P=0.02).

In concordance wi th the “ rever s e epidemiology” phenomenon in a CKD population, neither total cholesterol, LDL-C, HDL-C nor triglycerides were significantly related to the extent of coronary artery calcification in this hemodialysis cohort. In contrast, length of dialysis was significantly associated with the prevalence of valvular calcification, suggesting that coronary artery calcification is common, severe and significantly associated with ischemic CV disease in adult ESRD patients, likely the result of ESRD dysregulation of mineral metabolism, as the authors suggested.

Controlling Phosphorus, Reducing Calcification and Limiting Events

Patients do not have to reach ESRD to have evidence of vascular calcification. According to Block et al.⁷, two-thirds of patients starting on hemodialysis have evidence of vascular calcification and it is this group who is most likely to progress while on dialysis. Interestingly, the investigators also observed that across the length of their study, patients with no evidence of calcification at onset of hemodialysis did not progress.

Russo et al.⁸ also followed 53 patients with CKD stage 3 to 5 but who had not yet started on dialysis. At 24 months’ followup, investigators found that serum P was the only variable associated with progression of coronary artery calcification, which was prominent in this group of predialysis patients. Patients also experienced more fatal and nonfatal CV events than a control group with normal renal function.

Perhaps because of these and related findings, the use of phosphate binders to treat hyperphosphatemia in CKD stage 4 patients was recently recommended in the US. For CKD stage 3 and 4 patients, K/DOQI recommends serum P be maintained between 0.87 and 1.49 mmol/L (2.7 and 4.6 mg/dL).

Whether or not attenuation or amelioration in coronary artery calcification scores with phosphate chelating therapy is associated with a reduction in mortality requires further study but evidence so far is mixed. Block et al.⁹, for example, found that treatment with a non-Ca-containing phosphate binder was associated with a significant survival benefit in 127 patients who received the anionic polymer sevelamer rather than a Ca-containing phosphate binder. After a median follow-up of 44 months, a total of 34 deaths had occurred: 23 in the Ca-containing binder arm and 11 in the non-Ca containing binder arm (mortality rate: 5.3/100 patients-years vs. 10.6/100 patientyears for the two arms, respectively, P=0.05). Again, baseline coronary artery Ca scores were a significant predictor of mortality, with increased mortality proportional to baseline Ca scores (P=0.002).

In contrast, a survival benefit was not observed in the DCOR (Dialysis Clinical Outcomes Revisited) study in which over 2100 dialysis patients received either a non- Ca containing binder or a Ca counterpart.¹⁰ However, sevelamer did reduce mortality by 23% in DCOR patients who were =65 years of age compared to those who received a Ca binder, and there was a trend towards fewer hospitalizations and fewer days in hospital in the same non-Ca binder group.

Sevelamer also attenuated the progression of coronary and aortic calcification
s in the Treat-to-Goal study and extension by Chertow et al.¹¹ (Figure 2).

Figure 2.

<img2615|center>

Although both phosphate binders achieved similar control of serum P in the Treatto- Goal study, hypercalcemia was more common in the Ca binder arm at 16% vs. the non-Ca binder arm at 5% (P=0.04) and more patients in the Ca binder group had endof- study intact PTH levels below the target of 15.8 to 31.6 ng/mL (150 to 300 pg/mL) than the comparator arm at 57% and 30%, respectively (P=0.001). More relevant to the discussion, median absolute Ca scores in the coronary arteries and in the aorta increased significantly in the Ca-treated group but not in sevelamer-treated subjects (coronary arteries: 36.6 vs. 0, P=0.03 and aorta: 75.1 vs. 0, P=0.01, respectively). The median per cent change in coronary artery Ca score (25% vs. 6%, P=0.02) and aortic Ca score (28% vs. 5%,P=0.02) was also significantly greater with Ca than with the non-Ca containing binder. Interestingly, Asmus et al.¹² also found that two years after randomization, both trabecular and cortical bone density increased in the sevelamer group vs. a loss in bone density for patients taking a Ca-based binder (P<0.05).

Not all studies confirm the finding that limiting the use of non-Ca-containing phosphate binders arrest the progression of vascular calcification. Asmus et al.¹² just reported that coronary artery calcification scores increased by a mean of 35% at the end of one year in dialysis patients with known coronary calcification at baseline assigned to Ca acetate and by a mean of 39% in those assigned to sevelamer, even though both groups received intensive lipid-lowering therapy with a statin. However, according to Kestenbaum in an accompanying editorial, the rationale for this particular study is questionable because there is limited evidence that LDL-C actually contributes to coronary calcification in patients with or without CKD and at least in hemodialysis patients, LDL-C concentrations are not related to either coronary calcification or its progression. As he also pointed out, there is no effective treatment for vascular calcification and until there is, nephrologists need to try to continue to control serum P levels as well as they can, recognizing at the same time that there is a potential to induce dystrophic calcification that may arise as a result of excessive Ca loading.

Perhaps the best policy is to approach patients with known calcification or risk factors for it including higher serum phosphate and Ca levels with greater caution. In the advent that EBT does become more readily available outside the clinical trials setting, findings of calcification on EBT imaging may serve to guide clinical practice away from Ca-based phosphate binders and towards more effective serum P control.

Summary

K/DOQI guidelines support intensive control of serum P in patients with CKD but studies suggest that fewer than 30% of dialysis patients are able to achieve P levels in the suggested range. Patient education may help increase compliance to phosphate-binder therapy. However, current practice dictates that to achieve control in the majority of patients a high pill burden, which often serves to undermine adherence. While some studies suggest that phosphate binders have the potential to offset the CV disease risk due to soft tissue and vascular calcification, most notably non-Ca-containing binders, regulatory measures are needed to guarantee that patients with hyperphosphatemia have access to more effective and well tolerated phosphate binders.

References

1. Cheunq et al. Kidney Int 2000;58:353-62.

2. Rodríguez-Benot et al. Am J Kidney Dis 2005;46: 68-77.

3. Kestenbaum et al. J Am Soc Nephrol 2005;16: 520-8.

4. Kestenbaum B. Am J Kidney Dis 2008;51:877-9.

5. Wayhs et al. J Am Coll Cardiol 2002;39:225-30.

6. Raggi et al. J Am Coll Cardiol 2002;39:695-701.

7. Block et al. Kidney Int 2005;68:1815-24.

8. Russo et al. Am J Nephrol 2007;27:152-8.

9. Block et al. Kidney Int 2007;71:438-41.

10. St. Peter et al. Am J Kidney Dis 2008;51:445-54.

11. Chertow et al. Kidney Int 2002;62:245-52.

12. Asmus et al. Nephrol Dial Transplant 2005;20: 1653-61.

13. Qunibi et al. Am J Kidney Dis 2008;51:952-65.

BUILDING A CASE FOR A POTENT NON-CALCIUM-CONTAINING PHOSPHATE BINDER

Editorial Overview:

Denis Ouimet, MD, FRCPC

Associate Clinical Professor of Medicine, Université de Montréal, Staff Nephrologist, Hôpital Maisonneuve-Rosemont, Montreal, Quebec

All phosphate binders have been found to control hyperphosphatemia equally well in open-label randomized studies, though the number of tablets required to achieve target phosphorus (P) goals can vary significantly between the different types of phosphate chelators. The first phosphate binders were aluminum-based and are indeed very potent chelators of alimentary P. Unfortunately, aluminum-based binders significantly increase the normal amount of aluminum absorbed from the gut. This aluminum load requires a functioning kidney for elimination and in dialysis patients, aluminum levels build up over time, increasing the risk of osteomalacia as well as dementia. While aluminum-based binders are the least attractive of currently available options, because of formulary constraints, we are still using these potent phosphate binders not infrequently although with limited dosage. Concerns about calcium (Ca)-based phosphate binders and the risks of a positive Ca balance have already been addressed in other segments of this report. Thus, we will focus on the safety and related constraints for each of the two currently available non- Ca-based phosphate binders, sevelamer hydrochloride (HCl) and lanthanum carbonate.

Gastrointestinal, Liver, Kidney and Bone Effects

As the first of the non-Ca-non-aluminumcontaining phosphate binder to become available, sevelamer’s toxicity profile is clearly more acceptable than that of an aluminum-based binder or of Ca-containing phosphate binders. However, it does not bind phosphate very well and even less so at lower levels of intestinal pH. The pill burden is accordingly high to maintain serum phosphate in the normal range. It also binds biliary salts which could interfere with fat-soluble vitamin absorption and there is some risk with sevelamer HCl of inducing or increasing metabolic acidosis when using it.

While both aluminum and lanthanum carbonate are trivalent cations and bind trivalent anions including P, these two metals are quite different. Lanthanum carbonate is five times heavier than aluminum; intestinal absorption is 50 times less than aluminum and normal blood levels are also roughly 20 times less. The high affinity of lanthanum for phosphate ions means that drug-drug interactions are minimal. Unlike sevelamer, lanthanum carbonate binds phosphate equally well at all levels of pH in the GI tract.¹ Its main excretion route is through an endosomal-lysosomal hepatobiliary pathway and thus is not dependent on renal function; consequently, chronic kidney disease (CKD) patients with failing or non-functioning kidneys are not at risk of increasingly higher blood lanthanum levels as they are with aluminum-based phosphate binders.

As lanthanum carbonate is largely eliminated by the biliary route (80%), some 13% eliminated by direct transport across the gut wall into the lumen, levels do increase in the liver during the first few weeks of initiating treatment. However, with continued dosing, it reaches a steady state in the blood and in the liver, as equal amounts of the binder that enter the hepatocyte on the blood side are secreted into the biliary tract at the opposite side.

As reported by Hutchinson et al.², long-term treatment has not revealed any hepatotoxic effects, either clinically or in laboratory tests. A total of 93 patients, previously enrolled in earlier studies with variable exposure times to lanthanum carbonate of up to four years, were eligible for the two-year extension phase. No clinically relevant liver-related adverse events or changes in liver enzymes or bilirubin levels were observed at any time during a total of up to six years of exposure.

Although severe liver disease is relatively rare among dialysis patients, to use of lanthanum carbonate would be contra-indicated in patients with hyperphosphatemia and comorbid cirrhosis.

There is some limited diffuse and random accumulation of lanthanum carbonate in the bone, but it does not end up specifically at the calcification front, as does aluminum, where the latter interferes with bone mineralization and remodelling. Lanthanum is integrated in apatite lattice during mineralization similar to fluoride. It appears not to interfere with the bone mineralization process. In animal models, osteomalacia can be cured and renal osteodystrophy prevented, even with lanthanum being present in the bone.

In men, clinical studies with paired bone biopsies appear to support a therapeutic benefit of lanthanum carbonate on bone. D’Haese et al.³ have shown that, compared with patients treated with Ca carbonate, patients who received lanthanum carbonate for a year had significant improvement of bone histology, particularly less frequent evolution toward adynamic bone disease. Malluche et al.⁴ also found higher bone turnover after one year and higher bone volume after two years of treatment with lanthanum carbonate compared to standard treatment. This improvement occurred even in the presence of increased bone lanthanum content over time, approximately 1 µg/g/year.⁵

The up to six-year open-label extension trials reported by Hutchinson et al. showed that serum levels of Ca, parathyroid hormone (PTH), bone-specific alkaline phosphatase and osteocalcin generally remained stable across the treatment interval.

Adverse Events

No new adverse events (AEs) or unexpected increase in the incidence of AEs with increasing exposure to lanthanum carbonate were observed after up to six years of followup. AEs considered by investigators to be related to treatment occurred in approximately
up, the most common being nausea, diarrhea and flatulence. Gastrointestinal intolerance is not uncommon with all phosphate binders (Figure 1).

Figure 1.

<img2616|center>

Unlike aluminum-based binders, lanthanum carbonate does not cross the blood-brain barrier. It does not affect cognitive functions, based on results from a series of cognitive function tests by Altmann et al.⁶

Thus, it can be said that lanthanum carbonate has a good safety profile and appears to be reasonably well tolerated by the majority of patients requiring phosphate binder therapy.

Determinants of Adherence

Hyperphosphatemia is well documented for carrying risks of morbidity and mortality in dialysis and pre-dialysis patients. Hyperphosphatemia may be an inducer and/ or a marker of a more pathological state. It can favour ectopic calcifications in soft tissues, arteries and heart valves with potentially morbid or lethal consequences, as it can be a surrogate marker for non-compliance to various aspects of renal replacement therapy. In a systematic review of the prevalence and determinants of non-adherence to phosphate binder therapy in patients with end-stage renal disease (ESRD), Karamanidou et al.⁷ identified 34 studies meeting inclusion criteria and 13 reported rates of non-adherence of 22 to 74% (mean 51%), the difference in rates partially due to the way adherence was defined and measured in these studies.

Factors that might explain these high rates of non-adherence were not widely assessed in these studies, but the large daily pill burden consumed by dialysis patients appears to be salient among them. As Hutchinson and colleagues8 pointed out, CKD stage 5 patients may be taking more than 10 different medications in a single day. Typically, the pill burden can be expected to be two- to threefold higher than the total number of prescribed medications per day and the cumulative effect of multiple dosing regimens may not only be confusing to patients, it is understandably overwhelming. This is particularly true for Ca-containing phosphate binders and sevelamer HCl, which require approximately three tablets per meal to achieve National Kidney Foundation Kidney Disease Outcomes Quality Initiative (K/DOQI) targets.⁷ This pill burden was determined as the mean value in clinical trials, but in several patients the pill burden is even higher.

Current formulation of lanthanum carbonate contains up to 1000 mg of active chelator per tablet and as such, most patients control elevated phosphate levels and achieve K/DOQI targets on treatment with a single tablet taken with each meal. This was shown by Hutchinson et al.⁸ who sought to determine what dose of lanthanum carbonate as monotherapy would produce the same serum phosphate control as their previous phosphate-binder therapy or therapies.

In the open-label trial, treatment was initiated at a dose of 1500 mg/day divided equally between meals and then adjusted each week to achieve optimal serum P control. Patients could receive a maximum dose of 3000 mg/day by the third week and 4500 mg/day by the fifth week.

The primary end point was the percentage of patients with controlled serum phosphate established in K/DOQI guidelines. At screening, before the wash-out period, approximately 35% of patients met K/DOQI serum phosphate targets. This percentage increased to 48% by week 12 with lanthanum carbonate monotherapy. Among those patients whose serum phosphate was between 1.13 and 1.78 mmol/ L by week 12, more than 77% were taking
n doses of =3000 mg/day. Of those patients whose serum phosphate was not controlled on previous phosphate binder(s), approximately 26% achieved control on lanthanum carbonate (Figure 2).

Figure 2.

<img2617|center>

As Hutchinson and co-authors suggested, many nephrologists tend to combine phosphate binders to try to improve efficacy but this may be counterproductive, only adding to the complexity of the regimen and making it more difficult for patients to adhere. While not assessed in this study, reducing the tablet burden may improve patient adherence.

In an effort to avoid well-documented toxicities associated with Ca overload, we currently use up considerable financial resources on alternatives to Ca-based phosphate binders. However, if an alternative costly phosphate binder has a low potency and a high pill burden, it incurs a risk of nonadherence in a significant number of patients. Such a binder would certainly not be costeffective in these patients who cannot comply with the regimen. Conversely, an equally costly but more potent phosphate binder—and thus one with a much lower pill burden— would increase the odds of compliance and ultimately be more cost-effective, in these circumstances.

Summary

In the management of hyperphosphatemia, we still overwhelmingly rely on Ca-based phosphate binders and/or sevelamer HCl due to formulary constraints. Yet we are largely unsuccessful at achieving K/DOQI targets, at least in part because of the demanding pill regimen associated with these phosphate binders. With a more potent phosphate binder that allows a lower pill burden, more patients may achieve K/DOQI targets for serum P. Increasing the armamentarium with a single tablet of lanthanum carbonate three times a day has the potential to enhance adherence and be more cost-effective in the end.

References

1. Autissier et al. J Pharmaceutical Sciences 2007; 96(10):2818-27.

2. Hutchinson et al. Nephron Clin Pract 2008;110(1): C15–C23.

3. D’Haese et al. Kidney Int 2003;63(suppl 85):S73-S78.

4. Malluche et al. Clin Nephrol 2008;70:284-95.

5. Bronner et al. Clin Pharmacokinet 2008;47(8):543-52.

6. Altmann et al. Kidney Int 2007;71:252-9.

7. Karamanidou et al. BMC Nephrol 2008;9:2.

8. Hutchinson et al. Nephrol Dial Transplant 2008; 23(11):3677-84.

PRACTICAL LESSONS FOR THE PRACTICING NEPHROLOGIST

Editorial Overview:

Robert Horne, PhD

Professor of Behavioural Medicine, Director, Centre for Behavioural Medicine, Department of Practice and Policy, The School of Pharmacy, University of London, London, UK

A recent study¹ has confirmed speculation that consistent control of mineral and bone disorders in hemodialysis patients results in markedly improved outcomes, as discussed previously in this report. Effective therapies can improve bone and mineral status in chronic kidney disease (CKD) patients. However, low adherence compromises benefit. This article will examine the reasons for non-adherence and discuss implications for interventions to help patients get the best from medicines.

According to a World Health Organization report “Adherence to long-term therapies, evidence for action.” Geneva: WHO 2003), approximately 50% of medications prescribed for long-term illnesses are not taken as directed. Chronic kidney disease (CKD) is no exception. A recent review of adherence to phosphate-binding medication identified a mean non-adherence rate of 51% across 13 studies.² However, there was quite a wide variation across studies (22% to 74%) in part attributable to the way in which nonadherence was measured and defined. If the prescription was appropriate, then this level of non-adherence is a concern for those providing, receiving or funding healthcare because it not only entails a waste of resources but also a missed opportunity for therapeutic benefit. Unfortunately, despite the high individual and societal costs of non-adherence, effective interventions remain elusive,³ due to limitations in the way interventions have been designed and tested.⁴ However, research over the last decade has improved our understanding of the reasons for non-adherence and how we might develop more intelligent and effective interventions to help patients obtain the best from medicines through optimal adherence to appropriate prescriptions.⁵

The Causes of Non-Adherence: Dispelling Common Myths

Non-adherence is not significantly related to the type or severity of disease with rates of 25% to 30% noted across 17 disease conditions.⁶ Furthermore, providing clear information, although essential, is not enough to guarantee adherence. Likewise, a plethora of studies have failed to identify clear and consistent relations between adherence and socio-demographic variables such as gender and age in adults.⁷ Adherence is positively correlated with income when the patient is paying for treatment^8,9,10,11 but not with general socio-economic status.⁷ Adherence is often lower for more complex regimens, but reducing the frequency of dosage administrations does not always solve the problem.¹² There is little evidence that adherence behaviours can be explained in terms of trait personality characteristics. Even if stable associations existed between socio-demographic or trait characteristics, they would serve to identify certain “at risk” groups so that interventions could be targeted but could do little to inform the type or content of these interventions. The challenge of developing interventions that support adherence is to identify causes that can be modified rather than fixed characteristics which cannot. In summary, the notion of the “non-adherent patient” is a myth: most of us are non-adherent some of the time. Nonadherence is therefore best seen as a variable behaviour.

Non-Adherence as a Variable Behaviour with Intentional and Unintentional Causes

There are many causes of non-adherence but they fall into two overlapping categories: intentional and unintentional (Figure 1). Unintentional non-adherence occurs when the patient’s intentions to take the medication are thwarted by barriers beyond their control such as poor recall or comprehension of instructions, difficulties in administering the treatment or simply forgetting. Deliberate or intentional non-adherence arises when the patient decides not to follow the treatment recommendations. It follows that we need to consider two issues: resources and motivation. Unintentional nonadherence is linked to problems of resources. To understand intentional non-adherence we need to consider the processes influencing motivation to start and continue with treatment such as the beliefs and preferences of the patient.

This simple model explains the limited efficacy of interventions to improve adherence through information provision or ‘education.’

In order for information to change behaviour, it must be consistent with or change our underlying beliefs. But what are the key beliefs influencing uptake and adherence to medication? Over the l
eagues and I have researched this question. One of our goals was to identify an accessible framework summarizing the key beliefs influencing medication uptake and adherence across clinical, socio-demographic and cultural categories.

Figure 1.

<img2618|center>

Identifying the Salient Beliefs About Medicines Influencing Adherence to Medicines: the Necessity Concerns Framework

The Necessity Concerns Framework (NCF) is a theoretical model of adherence-related beliefs to be specifically developed for medication.^{1
posits that the motivation to start and persist with prescribed medication is influenced by the way in which the individual judges the necessity of the medication for maintaining current and future health relative to their concerns about potential adverse effects (Figure 2).}

Figure 2.

<img2619|center>

Studies involving patients from a wide range of illness groups including asthma,^14,15 renal disease,^16,2 renal transplantation¹⁷ diabetes, cancer and coronary heart disease,¹⁸ bipolar disorder,¹⁹ hypertension,²⁰ depression,^21,22 HIV/AIDS,^23,24 rheumatoid arthritis²⁵ and hemophilia²⁶ have consistently found that low rates of adherence are related to doubts about personal need for medication and concerns about potential adverse effects. A meta-analysis of over 30 peer-reviewed studies confirms the utility of the NCF in operationalizing the salient medication beliefs influencing adherence across a range of therapeutic categories, cultures, sociodemographic groups and countries.²⁷

The Common-sense Origins of Necessity Beliefs and Concerns

Doubts about necessity and concerns about potential adverse effects are modifiable determinants of adherence. However, in order to challenge doubts about personal need for medication and allay concerns, we need to understand their origins. Research has shown that necessity-concerns evaluations maybe related to misconceptions about the disease and the likely benefits and risks of treatment.^15,24

Medication Necessity Beliefs

People do not blindly follow treatment advice even from respected clinicians. Rather, we evaluate the advice and decide whether it is a good idea for us, based on our understanding of the illness and treatment. Research shows that this is where an adherence problem often begins. In order to be convinced of a personal need for ongoing medication, we must first perceive a good fit between the problem (the illness or condition) and the solution (the medication).

Patients’ common-sense perceptions of illness influence their beliefs about the necessity of medication.¹³ Symptom perceptions relative to expectations are key. Until we experience a chronic condition most of our experience of illness is symptomatic and acute. We know we are ill because we experience symptoms. We know when we are better because the symptoms go away. We carry these expectations of symptoms and illness with us when we encounter long-term conditions. For many long-term conditions the medical rationale for maintenance treatment is based on a prophylaxis model. The benefits of treatment are often silent and long-term. This may be in stark contrast to our intuitive common-sense model of “no symptoms, no problem.”²⁸ Moreover, missing doses may not lead to an immediate deterioration in symptoms, so reinforcing the (erroneous) perception that high adherence to the medication may not be necessary.¹⁵ Related to this is the fact that people often stop taking treatment when they judge that the condition has improved. These judgments are often based on potentially misleading symptom perceptions rather than on objective clinical indicators of disease severity.²⁴

Medication concerns are often related to “background beliefs” about the nature of pharmaceuticals as a class of treatments.29 Surveys of lay beliefs about medicines have shown that many people are suspicious of pharmaceuticals and the pharmaceutical industry. They tend to view all medications as having common properties.¹³ Their benefits are often taken for granted with a focus on the potential negative effects. In this view medications are often seen as intrinsically harmful addictive substances that are overused by physicians and the healthcare system.²⁹

Implications for the Development of Effective Interventions

The NCF provides a model for the development of interventions. The observed, consistent relations between adherence and medication beliefs (necessity beliefs and concerns) and their origin in misconceptions about illness— and the likely benefits and risks of treatment— creates an opportunity for educational interventions. I have outlined a Perceptions and Practicalities Approach (PPA) to facilitating informed adherence which uses the NCF as a means of identifying and addressing the key perceptual barriers to adherence (Figure 3). Efforts to support patient adherence are likely to be more effective if they take account of the perceptual factors and concerns e.g. necessity beliefs in influencing patients’ motivation to start and continue with treatment, as well as the practical barriers (e,g limitations in capacity and resources) influencing their ability to take the medication as recommended. The first step is to provide the patient with a ‘common-sense’ rationale for why the treatment is needed. This is particularly true for phosphate binder medication as many patients do not have a clear understanding of why phosphate control might be important to them or how phosphate bin
onstrated by Riley et al.³⁰ This is necessary to help patients see a close fit between the problem (uncontrolled phosphate) and the situation (adherence to phosphate binder medication), and to provide a convincing “story” for why medication is still necessary, even when it does not improve symptoms.

Figure 3.

<img2620|center>

Secondly, clinicians should elicit and address patient concerns about the medication, help the patient to make treatment choices that are informed by an understanding of the likely risks and benefits, rather than by potentially erroneous beliefs or misconceptions about the condition and treatment.

Finally, it is important to assess and address potential practical barriers and to make the regimen as convenient and easy to take as possible.

This process of adherence support should occur not just at the start of treatment but also during treatment review as beliefs, adherence and support needs may change over time. Although this may take more time in the short term, this is likely to be justified by improved treatment outcomes and greater efficiency in the long term. Moreover, other members of the clinical team, particularly pharmacists, could contribute to the provision of adherence support, so saving physician time.

Summary

Non-adherence is an impediment to optimum treatment outcomes in CKD. It is a big problem but it is should not be seen as the patient’s problem. Rather it represents a fundamental limitation in the delivery of healthcare, often due to a failure to fully agree on the prescription in the first place or to identify and provide the support that patients need later on. Practitioners have a duty to help patients make informed decisions about treatment and use appropriately prescribed medicines to best effect. Addressing non-adherence requires a “no-blame” approach that encourages patients to discuss their use of medicines and any doubts or concerns they may have about the treatment. Efforts to support adherence will be more effective if they are tailored to the needs of individuals and take account of the perceptual factors (e.g. beliefs and preferences) influencing patients’ motivation to start and continue with treatment as well as the practical factors influencing their ability to adhere to the agreed treatment. Supporting adherence requires a team approach from clinicians, pharmacists and nurses, and needs to happen not just at prescribing, but also during regular review, as perceptions, abilities and adherence may change over time.

References

1. Danese et al. Clin J Am Soc Nephrol 2008;3(5): 1423-9.

2. Karamanidou et al. BMC Nephrol 2008;9:2.

3. Haynes et al. Cochrane Database Syst Rev 2005;4:CD000011.

4. Horne et al. Concordance, Adherence and Compliance in Medicine Taking: A conceptual map and research priorities. London: National Institute for Health Research (NIHR) Service Delivery and Organisation (SDO) Programme. http://www.sdo.nihr.ac.uk/ sdo762004.html, 2005.

5. Horne R, Kellar I. Chapter 6: Interventions to Facilitate Adherence. In: Horne R, Weinman J, Barber N, Elliott RA, Morgan M, eds. Concordance, Adherence and Compliance in Medicine Taking: A conceptual map and research priorities. London: National Institute for Health Research (NIHR) Service Delivery and Organisation (SDO) Programme. http://www.sdo.nihr. ac.uk/sdo762004.html, 2005.

6. DiMatteo et al. Health Psychol 1993;12:93-102.

7. DiMatteo et al. Med Care 1980;18:376-87.

8. Piette JD, Heisler M, Wagner TH. Arch Intern Med 2004;164(16):1749-55.

9. Piette JD, Heisler M, Wagner.TH. Am J Public Health 2004;94(10)1782-7.

10. Piette JD, Heisler M, Wagner.T.H. Diabetes Care 2004;27 (2):384-91.

11. Piette et al. Med Care 2004;42(2):102-9.

12. Claxton AJ, Cramer J, Pierce C. Clin Ther 2001;23(8):1295-310.

13. Horne R. Treatment perceptions and self regulation. In: Cameron LD, Leventhal H, eds. The self-regulation of health and illness behaviour. London: Routledge, 2003:138-53.

14. Menckeberg et al. J Psychsom Res 2008;64:47-54.

15. Horne R, Weinman J. Psychol Health 2002;17(1): 17-32.

16. Karamanidou C, Weinman J, Horne R. Br J Health Psychology 2008;13(2):205-14.

17. Butler et al. Nephrol Dial Transplant 2004;19(12): 3144-9.

18. Horne R, Weinman J. J Psychosom Res 1999;47(6): 555-67.

19. Jónsdóttir et al. Acta Psychiat Scand 2008; In press.

20. Ross S, Walker A, MacLeod MJ. J Hum Hyperten 2004;18(9):607-13.

21. Aikens et al. Ann Fam Med 2005;3(1):23-30.

22. Hunot et al. Prim Care Companion Clin Psychiatry 2007;9(2):91-9.

23. Gonzalez et al. Ann Behav Med 2007;34(1):46-55.

24. Horne et al. JAIDS 2007;45(3):334-41.

25. Neame R, Hammond A. Rheumatol 2005;44(6): 762-7.

26. Llewellyn et al. Psychology and Health 2003;18(2): 85-200.

27. Horne et al. Data on file, manuscript in preparation.

28. Halm et al. Chest 2006;129(3):573-80.

29. Horne R, Weinman J, Hankins M. Psychol Health 1999;14:1-24.

30. Riley et al. Br J Med 2007;12(1):19-22.