Several recent studies of novel potassium binders were recently published in the New England Journal of Medicine (NEJM) and Journal of the American Medical Association (JAMA). Some of the authors of the HARMONIZE trial (Drs. Mikhail Kosiborod, David Packham, Henrik Rasmussen, Alex Yang, and Bhupinder Singh: HMAuthors) discuss their publications about the novel compound ZS-9 (sodium zirconium cyclosilicate), and how it might change the current clinical approach to hyperkalemia management, with Dr. Edgar V. Lerma (eAJKD), eAJKD Advisory Board Member.

Disclosures: Dr. Lerma is an author of the JAMA article; Dr. Mark Rocco, medical writer for ZS Pharma, helped prepare the interview.

eAJKD: You recently published two Phase 3 ZS-9 studies. Please tell us your major findings and conclusions?

HMAuthors: ZS-9 has been evaluated in three prospective, randomized, double-blind, placebo-controlled studies with over 1100 patients to date. From these studies, daily administration of ZS-9 has been observed to rapidly and predictably lower and maintain normal serum potassium levels, while demonstrating a favorable safety and tolerability profile.

The acute phase (first 48 hours) of the first pivotal Phase 3 trial (ZS-003), which was recently published in the NEJM, showed statistically significant and clinically meaningful reductions in serum potassium in patients receiving 2.5g, 5g or 10g of ZS-9 three times daily, compared with placebo, with 99 percent of subjects becoming normokalemic after receiving 10 grams of ZS-9 thrice daily for 48 hours. Additionally, the maintenance phase of this trial demonstrated that the reductions achieved in the acute phase could be maintained for up to 12 days using 5 grams or 10 grams of ZS-9 administered once daily. These results were statistically significant when compared to patients randomized back to placebo, who reverted to hyperkalemia.

In the second pivotal Phase 3 trial, known as HARMONIZE (HyperkAlemia RandoMized interventiON multI-dose ZS-9 maintEnance), all three doses (5, 10 and 15 grams) of once daily ZS-9 met the primary endpoint, producing dose-dependent lower potassium levels and higher proportions of patients with normal potassium levels for up to 28 days as compared to placebo. Specifically, the proportion of patients with normal potassium levels during the randomized maintenance phase was 94%, 90%, and 80% in patients receiving 15g, 10g, and 5g once daily of ZS-9 versus only 46% in patients receiving placebo. Similar to the first, larger Phase 3 trial, the HARMONIZE study showed rapid, statistically significant and clinically meaningful reductions in serum potassium in the acute 48-hour phase, with a median time to potassium normalization of 2.2 hours. Ninety-eight percent of patients became normokalemic after receiving 10 grams of ZS-9 three times a day for 48 hours. Importantly, the degree of potassium reduction was greater in patients with more severe hyperkalemia at baseline. Furthermore, ZS was very well tolerated and safety, tolerability and adverse events were generally consistent with previous ZS-9 clinical studies.

Several additional points to highlight, when comparing the two studies include:

1. The main difference in the acute (initial 48 hours) phases of ZS-003 and HARMONIZE (ZS-004) was that one was double-blind, randomized, and placebo-controlled, while the other was open-label. Specifically, ZS-003 had a double-blind, randomized, and placebo controlled acute phase with multiple ZS-9 dose groups (1.25g, 2.5g, 5g, 10g) compared to placebo, whereas HARMONIZE was open-label with everyone receiving 10g TID of ZS-9 (patients entered the double-blind, randomized, placebo-controlled phase after the initial 48 hours).
2. The primary endpoint was acute efficacy (48 hrs) in ZS-003 and extended efficacy (28 days) in HARMONIZE (ZS-004).
3. Mean baseline potassium was higher in HARMONIZE than that in ZS-003 (5.6 and 5.3 mEq/L, respectively), because there was no upper limit on potassium at entry in the HARMONIZE study. For example, patients with potassium as high as 7.2 mEq/L were included in HARMONIZE, whereas patients with potassium levels above 6.5 mEq/L were excluded in ZS-003.
4. The two overlapping dose groups in ZS-003 and HARMONIZE, namely the 5g and 10g dose groups, demonstrated very similar and consistent safety and efficacy.

eAJKD: How does this novel drug work, and how is it different from sodium polystyrene sulfonate (SPS). Why was it compared to placebo as opposed to SPS?

HMAuthors: Sodium zirconium cyclosilicate (ZS-9) is a non-absorbed cation exchanger with a unique, inorganic crystalline structure that was specifically engineered to entrap excess potassium ions (K⁺) throughout the entire gastrointestinal tract. ZS-9 was designed to be highly selective for potassium and have high capacity for potassium absorption by engineering the crystalline pore size to ~3 Å, mimicking the selectivity filter of physiologic potassium ion channels. Due to its selectivity, ZS-9 binds potassium throughout the GI tract, including potassium in food when given in conjunction with meals. This feature is likely responsible for the very rapid onset of action with a clinically meaningful and statistically significant reduction in potassium levels occurring as early as 1 hour after administration. Conversely, organic polymer resins, such as sodium polystyrene sulfonate, SPS (Kayexalate^®), only begin working once they reach the colon, where potassium concentrations are higher.

In contrast to organic polymer resins already on the market (i.e., SPS) or those currently in development, ZS-9 is insoluble, non-absorbed, highly stable, administered without a purgative agent (i.e., sorbitol), and does not swell in contact with water (i.e., is non-hygroscopic) – features that likely explain its favorable tolerability profile.

SPS was approved in the late 1960s without demonstration of efficacy in randomized, placebo-controlled trials. Additionally, SPS has been associated with poor tolerability and multiple reports of serious gastrointestinal toxicity. This makes its use questionable in the acute setting and unsuitable for chronic administration. Hence, it was deemed an inappropriate comparator, with recommendation directly from the FDA to conduct placebo-controlled studies instead.

eAJKD: What are the side effects/drug interactions of ZS-9?

HMauthors: In all three clinical trials to date, ZS-9 has demonstrated a favorable safety profile. It has been well-tolerated across all pre-defined subgroups, including CKD, HF, and diabetes mellitus patients, as well as those patients receiving renin-angiotensin-aldosterone system (RAAS) therapy. Moreover, treatment-related adverse events, including gastrointestinal (GI) side effects, were comparable to placebo.

In the HARMONIZE study, peripheral edema was observed more frequently with the 15 gram ZS-9 dose than with the other ZS-9 doses or with placebo. However, this observation was most likely due to imbalances between patients receiving 15g of ZS-9 versus placebo, with higher rates of heart failure, more patients with an eGFR <60, and higher levels of brain natriuretic peptide (BNP) at baseline in patients receiving 15g of ZS-9 versus those receiving placebo (e.g., 45% with heart failure on 15g ZS-9 dose versus 31% on placebo). As expected, based on ZS-9’s mechanism of action, the rate of mild hypokalemia (potassium levels between 3.0 – 3.4 mEq/L) was higher in patients receiving the two highest doses of ZS-9 (10g and 15g daily) in the HARMONIZE study; all of these transient episodes of mild hypokalemia resolved after protocol-directed dose adjustments. There were no cases of severe hypokalemia (potassium levels < 3.0 mEq/L) in any of ZS-9 trials to date, which includes over 1100 patients exposed to the compound.

In contrast, the product labels of approved potassium-binding treatment options, such as SPS, contain warnings of severe, potentially fatal GI side effects, including colonic necrosis, bleeding and perforation, especially when co-administered with sorbitol.

eAJKD: How does ZS-9 compare with SPS? How does it compare with Patiromer?

HMAuthors: Like SPS and Patiromer, ZS-9 is a non-absorbed potassium binder. However, unlike these two organic polymer compounds that have non-specific cation binding and are co-administered with sorbitol, ZS-9 is an inorganic crystal designed to be highly selective for potassium and does not require, nor is administered with, a purgative agent. This results in a potassium-binding agent with higher potency, faster onset of action, better GI tolerability, and no impact on other electrolytes (e.g., magnesium and calcium ions).

eAJKD: In which populations have you tested ZS-9?

HMAuthors: The ZS-9 clinical program was designed to investigate treatment of acute and chronic hyperkalemia in ambulatory patients. Therefore, we expect ZS-9 to help all hyperkalemic patients regardless of the underlying cause. In our three prospective, randomized, double-blind, placebo-controlled clinical trials and our open-label continuation trials, studying over 1200 patients to date, we prospectively specified CKD, diabetes, heart failure, and patients on RAASi therapy as important subgroup populations. ZS-9 has been shown to be equally efficacious and safe in all of these pre-defined patient subgroups. In addition, in the HARMONIZE and ZS-005 studies, there was no upper limit in serum K⁺ levels and no lower limit of eGFR; so even patients with stage 5 CKD (not on dialysis) were enrolled.

eAJKD: What are the effects of ZS-9 on Ca⁺², Mg⁺², HCO₃^–?

HMAuthors: Unlike non-specific organic polymer exchange resins such as SPS, and the investigational agent Patiromer, sodium zirconium cyclosilicate (ZS-9) is a selective, non-absorbed, cation exchanger with a unique crystalline lattice that traps excess K⁺ in the gut with >25-fold selectivity for K⁺ over Ca⁺² and Mg⁺². In contrast, non-specific polymers like SPS have reported hypocalcemia and hypomagnesemia. It is hypothesized that the high K⁺ specificity of ZS-9 is attributable to the chemical composition and diameter of the micropores, which possibly act in an analogous manner to the selectivity filter utilized by physiologic K⁺ channels. As a likely result of this selectivity, in both Phase 3 trials to date, we observed no difference in either magnesium or calcium levels between any of the ZS-9 dose groups and placebo.

Interestingly, ammonium ions (NH₄⁺) are of very similar size to K⁺, and thus, are also bound by ZS-9. In our first two trials (ZS-002 and ZS-003), we observed an increase in serum bicarbonate, potentially contributing to an improvement in metabolic acidosis, which may be explained by the removal of ammonium by ZS-9. In particular, a plausible explanation for this improvement may be due to ammonia and bicarbonate combining in the liver to form urea, and therefore, by removing ammonia, bicarbonate is spared.

eAJKD: What does the future hold for ZS-9, in terms of FDA approval?

HMAuthors: In conjunction with our recently completed Phase 3 trial, the HARMONIZE study, we are conducting an extension study that is currently in its eighth month of investigation, and will generate longer-term open-label safety and tolerability data in patients who participated in HARMONIZE. Furthermore, we are conducting an additional long-term study to evaluate the extended safety and tolerability of ZS-9, as well as its efficacy in restoring and maintaining normal serum potassium levels in patients with hyperkalemia over at least one year of drug exposure.

In parallel, we intend to file a new drug application (NDA) with the U.S. Food and Drug Administration, and concomitantly with the European Medicines Agency (EMA), during the first half of 2015 with an anticipated approval in the first half of 2016.