Memorandum

Janssen initiates CREDENCE study testing canagliflozin’s effect on diabetic nephropathy – February 27, 2014

Executive Highlights

  • Study to enroll ~3,700 type 2 diabetes patients with mild or moderate renal impairment; estimated primary completion in January 2019 (first patient enrolled last week).
  • Janssen hypothesizes that a nephropathy benefit could stem from a direct mechanistic effect on the kidneys, and could be an SGLT-2 inhibitor class effect.
  • Included below is a transcript of our recent interview with Dr. Norman Rosenthal, canagliflozin compound development team leader at Janssen.

Last week, Janssen announced the initiation of a renal outcomes study, CREDENCE, investigating the potential protective effects of the SGLT-2 inhibitor Invokana (canagliflozin) on the progression of diabetic nephropathy. Given the dearth of options that have been shown to affect long-term diabetes complications, especially diabetic nephropathy, we are very excited that Janssen has at least enough confidence in the possibility of a nephropathy benefit to invest in this trial. The study (formally titled Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Examination) is expected to enroll over 3,700 patients with type 2 diabetes and stage two or three chronic kidney disease (eGFR: 30-90 ml/min/1.73 m2), with a projected primary endpoint of January 2019. Patient enrollment began last week.

Some may be surprised by the initiation of this trial given the questions about Invokana’s safety and efficacy in people with renal impairment based on phase 3 data (discussed at Invokana’s FDA Advisory Committee meeting): Invokana’s glycemic efficacy, like all other SGLT-2 inhibitors, decreases with decreasing renal function; Invokana use was also associated with a temporary slight worsening of eGFR in patients with moderate renal impairment. However, as we learned this week from Dr. Norman Rosenthal, canagliflozin development team leader at Janssen, it is hypothesized that SGLT-2 inhibitors may act directly on the kidneys to exert a renal-protective effect independent of glycemic control. Indeed, in Invokana’s phase 3 program, the company saw a remarkable ~50% reduction in urinary albumin/creatinine ratios in the ~1,000 patients with renal impairment (to boot, 80% were already on an ACEi or ARB therapy). For comparison, the angiotensin receptor blocker (ARB) losartan (a blood pressure drug that is one of the current standard-of-care choices for chronic kidney disease patients) led to a 39% reduction in urinary albumin/creatinine ratios in clinical testing. Investigators have yet to conclude what the exact mechanism might be, but a leading hypothesis is that canagliflozin decreases intraglomerular pressure and hyperfiltration by increasing filtrate sodium levels and mediating vascular changes in glomerular efferents and afferents. Other possible mechanisms could involve Invokana’s effects on blood pressure, uric acid, or atrial natriuretic peptide (see Dr. Rosenthal’s detailed discussion below).

If CREDENCE’s results are positive then the implications could be profound: to date no other drug has definitively shown a benefit on improving long-term diabetes outcomes. If Invokana's CV profile comes out clean in CANVAS (and there is no reason it shouldn't), then a composite profile including weight loss, blood-pressure lowering, oral administration AND improved renal outcomes would be extremely compelling and, we imagine, may even (best case scenario) propel SGLT-2 inhibitors into the first-line position (or at least a higher position that current SGLT-2s have). It is surely too early to say if J&J would ever position Invokana as a CKD drug or if it would simply position it as the only diabetes drug to show renal protection, but either would offer a competitive advantage over the rather homogenous SGLT-2 inhibitor class. Lilly/BI are also studying renal outcomes (in the CARMELINA trial) for their phase 3 SGLT-2 inhibitor empagliflozin (expected to completely in 2018).

See below for a transcript of our interview with Dr. Norman Rosenthal, canagliflozin compound development team leader at Janssen — he discussed with us the possible implications of the CREDENCE trial, as well as thoughts and theories on canagliflozin’s possible effect on nephropathy.

  • CREDENCE trial details: The randomized, double-blind, placebo-controlled, parallel group trial (ClinicalTrials.gov Identifier: NCT02065791) will enroll ~3,700 type 2 diabetes patients (A1c from 6.5% to 10.5%) with diabetic nephropathy (eGFR between 30 and 90 ml/min/1.73 m2) and macroalbuminuria (urine albumin-to-creatinine ratio of 300-5,000 mg/g). Patients will not be excluded based on other anti-hyperglycemic agents except for other SGLT-2 inhibitors. The composite primary endpoint is the time to the first occurrence of either end-stage renal disease (ESRD), doubling of serum creatinine, renal death, or cardiovascular death. Secondary outcomes include sub-composites of time to first renal or first cardiovascular event, as well as all-cause mortality. Patients will be randomized to canagliflozin 100 mg (the lower of the two doses current available, for safety reasons) or placebo, and will receive standard of care treatment for diabetic nephropathy, including maximally tolerated doses of an ACE inhibitor or ARB. Study investigators will receive guidance on the management of volume-related adverse events and hyperkalemia. A1c levels will be measured during the study, but given that this is a treat-to-target study design, Janssen expects glycemic control to be relatively balanced between groups. The estimated primary completion date for CREDENCE is January 2019.
  • Janssen is conducting another long-term renal study (CANVAS-R) that will measure the progression of albuminuria as its primary endpoint — an important distinction is that CANVAS-R is a biomarker study, while CREDENCE is an outcomes study. Both CANVAS-R and CREDENCE will study the incidence of cardiovascular adverse events as non-primary endpoints. With the completion of CANVAS-R, CREDENCE, and the original CANVAS CVOT, Janssen will have a wealth of long-term data on the effect of Invokana on diabetes and its complications. We will be very interested to see how renal considerations in the SGLT-2 inhibitor arena are affected by SGLT-1/2 dual inhibitors such as Lexicon’s LX4211, which (in phase 2 results) appeared to show both safety and preserved glycemic efficacy in patients with renal impairment. As such, one would hypothesize that SGLT-1/2 dual inhibitors could pack the one-two punch of preventing progression of nephropathy outcomes through both reduced hyperglycemia and the direct renal effects of SGLT-2 inhibition.

 

Interview with Dr. Norman Rosenthal (Canagliflozin Compound Development Team Leader, Janssen Research & Development)

Q: Thank you so much for taking our call! This is very exciting news from Janssen today. We were pleasantly surprised to learn about CREDENCE, given that in the phase 3 data we have seen previously, canagliflozin (Invokana) demonstrated somewhat limited efficacy in the renal impairment population. What is different about this population or approach that leads you to believe that it will be effective enough to improve renal outcomes?

A: You’re right that the efficacy of canagliflozin and the SGLT-2 inhibitors, when it comes to glycemic control, is proportional to the underlying status of renal function. We have a label for use with patients who have an eGFR that is down to the stage III level of 45 ml/min/1.73m2 at the 100 mg dose. I think that what this story is, from our perspective, is the emerging science that’s been building and extending the early findings from Dr. Brenner and others, on the role of hyperfiltration at the single nephron and its impact on initiating and causing progression of diabetic nephropathy [see this 1996 publication from Dr. Barry Brenner detailing the hyperfiltration theory].

So the thinking here is based on the underlying observations we’ve had from the preclinical side, which we’ve seen in a variety of different SGLT-2 inhibiting agents as well as what we’ve seen from a very large clinical program that enrolled over 10,000 subjects. We’ve had a breadth of exposure across so many different types of patients with diabetes, including over 1,000 patients with renal impairment with eGFRs as low as 30 cc/min and we were able to get some keen insights as a result of having the ability to look at not just the glycemic response, but other biomarkers such as the pharmacodynamic impact on serum creatinine, eGFR, and urinary albumin/creatinine ratios.

CREDENCE builds from what we’re seeing in the emerging science, which has helped us learn how interesting this transporter is as it relates not just to glycemic control, but also to the hypothesis we’re targeting now: the hemodynamic hypothesis. There could be other mechanisms that could have that renal protection because of the benefits we’re seeing on some of these biomarker changes.

Q: So the hypothesis is currently that the benefit on nephropathy is not just through improved glycemic control, but something mechanistically that’s going on in terms of the filtration that’s happening at the nephron?

A: Yes, exactly. Certainly we’re not claiming that we understand the exact nature of the potential mechanisms, but that hypothesis is the one that’s leading based on a lot of the preclinical and laboratory work. The hyperfiltration model is certainly a plausible one, given that one of the hallmarks of the diabetic kidney is an increase in intraglomerular pressure. The increase in intraglomerular pressure can cause traumatic changes to the structure of the glomerulus in the filtering unit and lead to breakdown of the basement membrane and other changes like mesangial expansion. That’s why one of the biomarkers you get is the leakage of albumin, which is familiar to us for its clinical relevance.

So, in a manner that is complementary to the blockage of the renin-angiotensin system, there is evidence that SGLT-2 inhibition can reduce hyperfiltration in the kidneys. The proposal is that with the blockade of the transporters in the proximal tubule, you’re delivering more sodium downstream to the distal tubule. There, the juxtaglomerual apparatus and the macula densa will read the increase and signal the arterioles feeding the nephron to increase in tone and constrict, which would decrease pressure flow into the glomerulus. That would be complementary to the way that experimental data indicates that ACEi and ARB agents work, which is through a direct impact on vasoconstriction in the efferent side and increased dilation in the afferent side. 

When we were looking at our 1,000 patients who had an eGFR between 30 and 60 in the stage three kidney disease area as well as in the broader program, we made a couple observations. We saw that eGFR falls quickly with exposure to the SGLT-2 inhibitor, but it happens with all of them. There was about a 2-4 ml/min/1.73 m2 reduction, which seems to be dose-dependent and reversible, and tends to stabilize of maybe even trend a bit back towards normal over time.

Concurrent with that, we also observed in our more focused review of the 1,000 patients with renal impairment that there was about a 50% reduction in urinary albumin/creatinine ratio, or excretion. In these 1,000 patients, over 80% were already on ACE or ARB therapy. By the way, when you look at the losartan study that Brenner published, the reduction in albumin/creatinine ratio seen there was only around 39%, and so we thought the effect size we were seeing looked very compelling. That data really helped us put our story together to take to regulators, and both the EMA and FDA were supportive of this as a scientific hypothesis, although it needs to be proven, which is the purpose of the work we’re doing now.

Q: It seems that this potential effect could be very exciting for patients. We haven’t heard of anything like this that could have a really positive effect in this area to make us believe that long-term complications like nephropathy could be prevented. What you are working on here for the field could be quite extraordinary.

A: I completely agree, this is all about the patient. Nephropathy is a complication of diabetes that can reach at least 30% of patients who have had type 2 diabetes, and it can lead to stage three or even more advanced kidney disease. We want to do whatever we can to look into what our SGLT-2 inhibitor can do beyond glycemic control. So we are very excited to have the support to take on this very large, event-driven trial. We plan to recruit 3,700 patients in 25 countries, at 600 sites. We know it’s going to take a couple of years to recruit the patients we’re looking for, but we’re committed to do whatever we can to fully show the extent of what this molecule can do, and our mission is to bring meaningful change to patients with diabetes.

Q: We’d love to help communicate news of this trial to patients and we’re sure other diabetes advocates would love to help as well! Big picture, could you talk, even broadly, about the bearing that this study could have on the cardiovascular outcomes trials being run for canagliflozin, and how the results from CREDENCE could affect the context in which CV data on canagliflozin is viewed?

A: We do have a post-marketing safety requirement to complete a second trial to demonstrate that there is no cardiovascular harm with canagliflozin. That study will complement the first one that we started pre-marketing called CANVAS. The second one recently started, and it’s called CANVAS-R. The two studies will represent 10,000 patients, and when we do our analysis, our expectation is that it will show no cardiovascular harm. It’s not powered to show a benefit, but they will fulfill our requirement to show that it’s safe from the FDA’s perspective with regards to cardiovascular safety.

However, the primary reason we called it CANVAS-R is because we’re measuring albuminuria as the primary efficacy measure, to investigate the progression or regression of albumin excretion. Patients are not being selected with a certain degree of albumin excretion, but we will be looking across a broad range at the progression of regression of albuminuria with canagliflozin compared to placebo. CANVAS-R is expected to read out in 2017. It’s a biomarker study, not an outcomes study like CREDENCE, but I think it will contribute as well to the understanding we’ve build from the 1,000-patient study from our earlier clinical program.

I also wanted to mention that we had our first patient enrolled in CREDENCE last week… so one down, 3,699 to go. I think in the marketplace, providers, caregivers, and patient advocates are very excited by this study, and I think there will be a lot of excitement that there is a new therapeutic opportunity that is being tested.

Q: This is so interesting; thank you for taking time to explain it all. A question on a related front: could you discuss whether you think that an effect on nephropathy could be a class effect?

A: As far as there being something unique about canagliflozin versus other agents, I don’t know. The clinical science and animal model work that others have done suggest that the whole class might have the attribute of reducing hyperfiltration. There might be other variables at play, such as differences in pharmacokinetics, efficacy related to glycemic control, or dosing, that are just unknown to us at the moment. My suspicion is that it’s probably more of a class effect, but that is only my best scientific guess at this point.

Q: And one last question – we know how busy you are! How do you see SGLT-1/2 dual inhibition fitting into the picture?

A: I’m not sure I’ve seen any evidence to suggest that SGLT-1 inhibition would be playing a role, if it is in fact the hemodynamic mechanism is at play here. There are other potential questions to consider, such as whether the effect is directly related to sodium or glucose. It could also be related to blood pressure or uric acid; there is evidence that SGLT-2 inhibitors reduce uric acid by around 10%. It could be related to atrial natriuretic factor. I think we need to make certain that we commit to understanding more broadly the other potential mechanisms before we assume it’s the hemodynamic hypothesis, even though we think that’s the leading hypothesis at the moment.

Q: Dr. Rosenthal, thank you so much for all you are working on, on this important front, and thank you to Janssen and J&J for taking this big bet.

 

-- by Manu Venkat, Jessica Dong, and Kelly Close