Spectrum Pharmaceutecals Inc (SPPI)

[antihama]

Talk about apropos. A paper from New England Journal of Medicine today titled “The Changing Face of Clinical Trials: The Primary Outcome Fails — What Next? (http://www.nejm.org/doi/full/10.1056/NEJMra1510064?query=TOC)” fits nicely with the long's belief that the 2 apaziquone trials which will be reviewed at the ADHOC meeting on Sept 14th will be given strong consideration for approval. From the intro

Nevertheless, an unreasonable yet widespread practice is the labeling of all randomized trials as either positive or negative on the basis of whether the P value for the primary outcome is less than 0.05. This view is overly simplistic.

The paper has a list of 12 key questions when the primary outcome fails (i.e. 0.05 p value). Some of the Qs are:

Is There Some Indication of Potential Benefit?
Whether a signal of treatment benefit (a “trend”) should be inferred from a P value greater than 0.05 requires thoughtful consideration...
Was the Trial Underpowered?
The inclusion of too few patients in a study increases the risk that a significant treatment benefit will not be shown, even if such an effect exists (a type 2 error). For instance, in a trial of bisoprolol versus placebo in patients with systolic heart failure,6 the hazard ratio for the primary outcome, death from any cause, was 0.80 (95% CI, 0.56 to 1.15; P=0.22). However, with only 621 patients, the trial was underpowered. Fortunately, the sponsors persisted, and the subsequent CIBIS II trial,7 which included 2647 patients, showed that mortality was lower among those who received bisoprolol than among those who received placebo (hazard ratio, 0.66; 95% CI, 0.54 to 0.81; P<0.0001). Note that the estimated 34% lower mortality with bisoprolol in this larger trial was within the 95% confidence interval reported in the first study.
In general, when a trial is too small to detect modest treatment effects, it is appropriate to describe the findings as inconclusive rather than negative. An adequately powered study requires the accrual of a sufficient number of primary-outcome events, which can be achieved by recruiting more patients, enrolling patients at higher risk, prolonging follow-up, specifying an outcome that occurs more frequently (including the use of composite outcomes), or a combination thereof.

With regards to Qapzola, Spectrum ran 2 double-blind, placebo-controlled, randomized multi-site P3 clinical studies
• Patients with primary or recurrent Ta, G1-G2 NMIBC
• 562 evaluable patients per trial
• Central Pathology read as Ta, G1-G2
•Primary endpoint: Recurrence rate at 2 years

Individually, the trials did not reach statistical significance; SPI-611 p-value was 0.13 and SPI-612 p-value was 0.082. Even though it turns out there was a reason why the individual trials failed i.e. blood enzyme, DT-diaporase, in the bladder after resection, combining the 2 almost identical randomized placebo controlled multi-site studies, you get strong statistical significance (SS) with a p-value of 0.022. If one could do a reboot, one would just add several hundred pts to each trial in order to get SS from the get go. Another Q to answer

Is There a Strong Biologic Rationale That Favors the Treatment?
One needs to be wary of arguments regarding biologic rationale. Almost any new treatment in a phase 3 trial has a plethora of supportive scientific data from animal studies and early-phase trials. Nonetheless, history is filled with records of many large pivotal trials that failed to exhibit any signs of efficacy (or that revealed heretofore unanticipated safety issues). For instance, the hypothesis that raising high-density lipoprotein cholesterol levels could be a new means of reducing cardiovascular events looked promising, but no trial of inhibitors of cholesteryl ester transfer protein has fulfilled that promise.36 Nature often overcomes our best efforts to interrupt the order of things. Thus, if methodologic flaws in a trial are not the cause of treatment failure, it is usually time to “move on,” while trying to understand the biologic reasons for the failure [my bold].

Here in fact, we know that flaws in treatment methodology was the cause of failure. When 1 accounts for the presence of blood enzyme, DT-diaporase, by looking at the 30 - 90 min after resection dosed patients. the individual studies become statistically SS w p-values of 0.04 and 0.022. When you combine the 30 - 90 min pts from the 2 studies you get a p value of 0.001.

DISCUSSION
The 12 points explained above can be used to provide assistance in deciding what to do next when a trial fails to produce a positive finding for its primary outcome. Certainly one needs to be circumspect. Researchers may opt to move in one of three directions.
Declare That the Trial Is Positive
Remarkable circumstances are usually required to report that a trial is positive even though the results of the primary outcome were not statistically significant at the prespecified level….

CONCLUSIONS
When the primary outcome of a trial fails to achieve statistical significance, we propose that researchers ask a series of searching questions that will help them clarify whether the new treatment may still have value. The options are to claim “success” anyway on the basis of the total evidence (an option that is rarely used), to plan a future trial with design improvements (a costly option), or to accept that the new treatment is likely to be ineffective (a frustrating option).

So how will the Advisory committee members look at the data presented? Will they look at it strictly from the classical view of the 2 P3 trials failing and vote it down or will they look at the data as a pretty convincing argument that if you don’t have apaziquone deactivated by the blood enzyme you get another tool in the war chest to fight NMIBC? I think the title of the paper says it all The Changing Face of Clinical Trials: The Primary Outcome Fails — What Next?