Ovature
The 14 April 2009 announcement concerning the halting of recruitment in the OVATURE study clearly is of considerable significance for the immediate future of the Company.
Despite the fact that the Company’s future is based on a substantial technology platform and is a whole lot more than one drug or one clinical trial, there is no escaping the fact that a negative outcome for OVATURE will have a significant impact on the Company’s ability to raise the capital required for ongoing clinical R&D.
“So what is the significance of the April announcement to the future of the OVATURE study?”
To answer this, we need to look at the announcement, first in context of what OVATURE was trying to achieve, and second what the Company’s options might be in the event that the trial data is positive.
1. Purpose and design of OVATURE
Two key questions often asked are (i) why was late-stage ovarian cancer chosen as the preferred avenue for registration of phenoxodiol, and (ii) why was the study designed in a way that was always going to make recruitment challenging?
The answer to both questions lies in the FDA’s particular view of the world.
For drugs such as phenoxodiol that target cancer cells directly and either kill them or stop them growing, the FDA traditionally has had a very clear definition of a beneficial anti-cancer effect, and that is that the new therapy needs to increase the life-span of a cancer patient. It hasn’t been enough in the past to show that the new drug might shrink tumors or even just slow down the overall rate of tumor growth when either outcome would not occur normally with available therapies. There have been some celebrated instances in the past where an experimental drug has slowed down initial rates of tumor growth, only to see that growth return in an accelerated manner, ultimately providing no eventual survival benefit to the patient. With this experience in mind, the FDA’s view is that the only real test of efficacy is when the overall survival rate of a group of cancer patients receiving a test drug is increased compared to a control group receiving the available standard therapy. Certainly other factors such as drug safety, practicality of use, cost and the quality of life being delivered to the patient, all come into the mix, but extension of life remains a crucial part of the equation for the FDA.
This is why the great majority of clinical trials of new anti-cancer drugs are carried out in patients with end-stage disease. First, because the time required to show a survival benefit in patients with end-stage disease is a lot shorter and that minimizes trial length and study costs; second, because the FDA requires that you show a survival benefit over ‘best practice’ therapy, and for end-stage disease that generally means no other therapy; third, because the toxicity of most anti-cancer drugs leads to an understandable reluctance to expose patients with early-stage disease to serious, perhaps even life-threatening, side-effects. But even without the side-effect issue, the harsh reality is that it is much quicker and substantially cheaper to prove that a new drug prolongs the survival of a cancer patient with end-stage disease by a matter of months or even weeks, than it does to show that survival of patients with early-stage disease is extended by a matter of years.
This regulatory reality has always been a particular challenge for phenoxodiol, because all the available information on the mechanism of action of phenoxodiol suggests that its effect is greatest the earlier in the disease process that it is used. And if this sounds an obvious statement that would apply to all anti-cancer drugs, it isn’t. The phenoxodiol situation is the opposite of the situation with almost every other anti-cancer drug.
Almost all other chemotoxic drugs work through a non-selective poisoning action of the cell to the point where the cell is unable to divide. That means that the effectiveness of these drugs is directly proportional to the rate at which they divide ….. the more the cancer is dividing, the more effective is the drug. But normal cells undergoing division also are susceptible to this poisoning effect, producing particularly toxic side-effects on tissues that have a relatively high rate of turnover such as bone marrow (anaemia), hair follicles (hair loss) and gut (vomiting, diarrhea). The challenge therefore with most chemotoxic drugs is to balance the positive with the negative effects. In late-stage disease, the cancer is likely to be the fastest growing tissue in the body and the effect of the drug on the cancer in that situation is proportionally greater than on normal tissues. But in early-stage disease, the cancer may be growing substantially slower than normal tissues such as bone marrow and the lining of the gut, so that the toxic effects of the drug on these tissues outweigh the effect on the cancer.
Phenoxodiol is in the opposite situation. There is no need with this drug to think of balancing negative and positive effects. Its highly selective action against cancer cells means that it is without any of the usual side-effects associated with anti-cancer drugs, meaning that it could be used in early-stage disease without any apprehension. Furthermore, it is more likely to be effective where the overall tumor load in the body is still relatively small and the individual tumor volume is more conducive to drug penetration.
For those with even a passing understanding of phenoxodiol and how it works, it is obvious that this is a drug that could revolutionise the face of cancer therapy through its ability to be used in early-stage disease.
And this was always going to be the challenge facing phenoxodiol. How to get marketing approval for a drug offering therapy at an early stage of the cancer process? The FDA has never approved (and probably never even considered) such a request, so there was no track record to rely on. The current Phase 2 study at Yale and Harvard universities using phenoxodiol to treat prostate cancer prior to it reaching an advanced state is a strategy based on this concept. But 6 years ago when a course of action for regulatory approval of phenoxodiol was being considered, there were enough unknowns in the whole process without facing something that no pharmaceutical company before had ever faced.
The decision about the preferred regulatory pathway to pursue was made easier by the unexpected discovery that one of the biological effects of phenoxodiol was its ability to restore chemo-sensitivity in cancer cells. This discovery suddenly expanded the possible therapeutic applications of phenoxodiol from that of a monotherapy where one would be relying solely on its ability to kill cancer cells in its own right, to that of a drug that could be used to restore sensitivity to far more potent cytotoxic drugs. This restorative effect ranged across most commonly-used anti-cancer drugs, but was particularly the case with the platinum-based drugs, cisplatin and carboplatin.
This finding subsequently led to considerable interest by a team at Yale University in the ability of phenoxodiol to restore sensitivity to platinum drugs in ovarian cancer cells. The platinum drugs are the most effective drugs in the treatment of ovarian cancer, but almost all ovarian cancers eventually go on to become resistant to them, and in so doing, become resistant also to all other drugs. The Yale team proposed that a combination of phenoxodiol and platinum drugs offered the best available therapeutic prospect for patients with late-stage ovarian cancer, and made a major commitment to pursuing this opportunity clinically.
The Yale proposal became an obvious route to pursue from a regulatory point of view. It provided an opportunity to use phenoxodiol in an end-stage situation where the use of phenoxodiol on its own might have been inappropriate. Women with late-stage ovarian cancer that has become resistant to platinum therapy have few, if any, therapeutic options remaining to them, and their survival consequently is very limited. The ability of phenoxodiol to restore sensitivity in these cancers to platinum drugs to the extent that survival would be extended, if only by a matter of only a few months, was an attractive therapeutic option to pursue for the Company as a means of bringing the drug to market.
Having opted for the concept of OVATURE as arguably the quickest and most cost-effective way of bringing phenoxodiol to market, the design of the study nevertheless was challenging, as the Company has noted on a number of occasions. The main challenge was what the drug was to be compared to in order to show a clinical benefit.
Normally the FDA would require a Sponsor to show that its experimental treatment is superior to existing best-practice. In the case of patients with ovarian cancer that finally becomes unresponsive to platinum therapy (known as platinum-refractory disease), best practice at the time that OVATURE was being designed was to try a range of anti-cancer drugs such as doxorubicin, docetaxol and gemcitabine used either singly or in combination. The likelihood of seeing any response to such drugs remained very low since (a) they were substantially less effective therapies than platinums in the case of ovarian cancer in the first place, and (b) the development of resistance to platinums normally meant that the cancer also became resistant to all other drugs in any case. Nevertheless, this remained an option of hope for a proportion of patients, that hope being that by slowing the growth of the cancer sufficiently to allow the patient to survive a further 12 months or so, then that passage of time might allow some restoration of sensitivity to platinum drugs.
If phenoxodiol was being assessed as a stand-alone monotherapy, then it could quite legitimately have been compared to a cocktail of various standard drugs as representing best practice. That approach would have slotted very well into current medical practice for patients with end-stage ovarian cancer, and given the very low success rate of ‘best practice’, a trial involving phenoxodiol as an alternative option no doubt would have been very attractive to oncologists and patients alike.
However, the strict interpretation of what was being tested in OVATURE was whether phenoxodiol could restore sensitivity to platinum in platinum-refractory cancer. And that meant that the only way you could test whether this was occurring or not was to compare the effect to that of platinum drugs alone. In made no difference that nobody, from regulators to highly experienced oncologists, had any expectation that re-treating a patient with a platinum within 6 months of showing no effect whatsoever with that same drug would produce any further effect. The harsh reality was that a direct comparison between phenoxodiol + platinum versus platinum therapy alone was unavoidable and non-negotiable from the FDA point of view.
And here was the challenge. For most oncologists, the notion of re-treating a patient with a toxic drug which just a few months earlier had shown no effect on tumor growth would be a no-no. Yet, this is how OVATURE had to be designed, with doctors knowing that they were committing half of their patients to a toxic course of therapy that was highly unlikely to yield any benefit.
To meet the definition of having a platinum-refractory cancer, OVATURE was required to go to a level of stringency that had not been used in any other study. Patients could only be admitted to the study where they had showed no response to a course of platinum therapy no more than 6 months prior to admission to the study. If that platinum-free gap was as much as 12 months, then the general experience among oncologists is that enough platinum-sensitivity might return simply through the passage of time to see some response in some patients to a repeat course of platinum therapy. But 6 months is universally regarded by oncologists as far too short to see any meaningful restoration of sensitivity.
The issue all along for the Company would have been whether the downside concerning the 50% of recruits that would end being randomized to the control group (platinum alone), would be offset sufficiently by the prospect that the remaining 50% receiving phenoxodiol would have the potential for some upside anti-cancer benefit.
The original study size was 470 patients (subsequently reduced because of slow recruitment to 340). Split between the 60 sites worldwide that the Company intended to recruit, that would have meant that the recruitment rate per hospital was not particularly onerous given the normal throughput of patients with ovarian cancer in those hospitals. With sites cooperative about putting patients onto the trial, it would have been reasonable to expect that there would have been full recruitment within 12-18 months.
However, recruitment very clearly has been difficult, suggesting that the downside of the study was too much of a challenge. Or perhaps as the Company has hinted at, ‘best practice’ care has shifted. When OVATURE was being planned, there was an emerging, but very embryonic, interest in so-called ‘dose dense platinum therapy’. Essentially, this means increasing the dose and the frequency of platinum therapy as the cancer progressively develops platinum-resistance. Emerging data suggests that this approach may be offering a small benefit, albeit at the cost of considerable risk of toxicity, but perhaps it has been of sufficient benefit to persuade patients to opt for it compared to the unknown prospective benefit of OVATURE.
This difficulty in recruitment clearly is the major reason behind the Company’s recent decision to suspend the study after only 141 patients have been recruited.
Finally, one further comment about the detail of the design of the study. The primary outcome being sought from the use of phenoxodiol is an increase in progression-free survival. This simply refers to the time that patients live without showing evidence of disease progression, something that would happen if the cancer either stopped growing or shrank. The size and number of tumors is being monitored every 2 months or more frequently if disease progression is suspected. Disease progression is defined as the development of new tumors or an increase in the diameter of the largest tumors by 25%.
For patients in the control group receiving carboplatin alone, most oncologists would accept that it would be quite extraordinary if any of those patients experienced any slowing in tumor growth. Once a tumor develops a level of resistance that results in it showing no response to a particular dose of platinum therapy, it is highly unlikely to show any further response to that same course of therapy within a period as short as 6 months later. For the patients receiving carboplatin alone, clinical experience would suggest that the disease would continue to grow inexorably in the face of carboplatin therapy, leading to a diagnosis of disease progression in some patients by the time of the first (+ 2 months) radiological examination, and in almost all remaining patients by the time of the second (+ 4 months) examination. There is no published data of which I am aware that details what the median progression-free survival time would be for such a situation, but a reasonable guess would be between about 10-12 weeks.
The hoped for outcome of the study would be that by the second (+ 4 months) radiological examination, a proportion of patients in the phenoxodiol + carboplatin group would be without disease progression, leading to a median progression-free survival time greater than 10-12 weeks.
2. The Accelerated Approval program
The Accelerated Approval program was introduced in 1992 in response to concerns over the long time many experimental life-saving drugs were taking to come to market as a result of long Phase 3 clinical trials. The argument was that for life-threatening diseases with no effective therapies, the normal burden of proof of benefit for experimental drugs should be relaxed providing that there is no increased jeopardy to patients.
In the words of the FDA, Accelerated Approval “is intended to make promising products for life threatening diseases available on the market on the basis of preliminary evidence prior to formal demonstration of patient benefit.” As discussed earlier, the FDA’s traditional definition of ‘patient benefit’ has been increased patient survival or relief of symptoms leading to improved well-being, and ‘preliminary evidence’ normally has been a surrogate marker, such as progression-free survival or tumour shrinkage which is considered likely to indicate clinical benefit.
Accelerated Approval status simply means that an experimental drug can be approved conditionally by the FDA for marketing before the drug has completed its final clinical trial and before the primary end-point is confirmed. The FDA still requires the Sponsor to go on to complete the study, but receiving Accelerated Approval status at an early stage means that the drug can be brought to market considerably sooner than would be the case if the clinical trial was being conducted under normal conditions.
Under the Accelerated Approval program, the FDA and the Sponsor agree on the design and the conduct of the study before it commences. In particular, both parties agree upfront on when the interim analysis will be performed, what the surrogate end-points are, and what level of statistical significance needs to be achieved. If the interim analysis data meets the pre-agreed objectives, then the FDA is obligated to grant Accelerated Approval.
The Accelerated Approval program has been something of a work in progress for the FDA since its inception 12 years ago, with two matters leading to adjustments to the program over that time. The first matter is that of Sponsors prematurely ending studies after receiving Accelerated Approval. In the first few years of the program, there were a number of instances of drugs receiving approval based on interim analysis data, but with the Sponsors then failing to go on to complete the study. This was perhaps not surprising when you consider that the Sponsor, with an approval in the drawer, had nothing to gain and everything to lose by continuing with the study, all at considerable extra cost. The FDA response to this was to insist that Accelerated Approval would only be granted when the study was fully enrolled and the Sponsor thereby committed to finishing the study.
The second matter concerns the fact that in a significant proportion of studies under the Accelerated Approval program over the past 15 years, the final analysis on completion of the study has failed to confirm the predictability of the surrogate end-points. That is, secondary end-points such as progression-free survival and tumour response rates have ultimately failed to translate into increased survival benefit. Despite this, many of these drugs remain on the market despite their approval not being converted into full traditional approval.
3. Significance of April 14 announcement
Returning now to our original question … what is the significance of halting OVATURE prematurely?
At the current rate of recruitment, the Company clearly had little option but to stop the study. A total of 141 patients recruited in almost 3 years made the task of recruiting a further 200 patients look fairly forbidding. This should not be seen as reflecting badly on the Company…rather it is the reality of the situation stemming from a study design they were obligated to follow by virtue of the FDA approval process. Such a step is not uncommon in the FDA approval process.
The Company is in the process of unblinding the data and has indicated that it hopes to have that data available about this coming July-August. I am not sure how realistic that target is given that the task of retrieving and then verifying large amounts of data is fairly daunting, but the Company must be reasonably confident of doing so by nominating a target date.
The fate of the trial and the data in the event of a negative outcome speaks for itself and there is little point in dwelling that situation. So let’s assume for the sake of discussion that the unblinded data proves to be positive. What are the options for the Company in that situation?
Attracting the interest of a major pharmaceutical player seems one likely option. A positive trial outcome would do what hasn’t been possible before, and that is to provide unequivocal proof of efficacy in a blinded, double-arm study. It is difficult to pinpoint why there has been so little apparent interest from major pharma in phenoxodiol to date when those companies often in-licence drugs with far less clinical data than phenoxodiol. A mindset that has trouble accepting that a drug can be potently cytotoxic against cancer cells but completely spare normal cells, or a drug that works by a mechanism (tNOX inhibition) that has little cache in cancer circles, are the sorts of cynicism that Novogen almost certainly would face. But nothing speaks like success, and an ability to reverse a high level of resistance to carboplatin where nothing previously has worked should go a long way to dissolving those prejudices.
Another option is to continue to pursue the Accelerated Approval application with the FDA. For this, we need to look at where the trial is now in relation to the milestones set by the FDA.
At the 2008 Annual General Meeting, the Company explained that the number of events that would form the basis of the final analysis was 190, and exactly half of that (95) for the interim analysis. The purpose of a total recruitment of 340 patients was to allow a healthy margin for patient drop-outs and matters such as incorrect patient management that would make a patient’s data ineligible. A total of 340 patients provided some certainty in getting to 190 analysable events.
So the first milestone to hit is 95 events. Is this likely from 141 patients? Probably, is the best guess.
From there, the arguments in favour of the Company would probably be along the lines of:
- that completion of the study was impractical;
- that the usual situation where a study was halted after the interim analysis without any meaningful data on the primary endpoint does not apply here. In OVATURE, the data being analysed at the interim analysis is the primary endpoint;
- that the drug was without significant safety issues.