FDA Rejects Apaziquone Application for Bladder Cancer

The new drug application (NDA) for apaziquone was for intravesical instillation immediately following transurethral resection in patients with non-muscle invasive bladder cancer. Spectrum has halted its ongoing phase III apaziquone program, and following discussions with the FDA, is considering a new smaller study of the treatment.

When ODAC convened in September, the panel considered an NDA for apaziquone based on findings from 2 phase III trials (identified as 611 and 612), which both missed the primary endpoint of a statistically significant reduction in disease recurrence at 2 years with a single dose of apaziquone versus placebo.

Subsequent to these findings, Spectrum further examined the drug through pooling analyses of the 2 trials that were not part of the initial study designs. The company reported to the panel that these analyses revealed that the impact of apaziquone may be increased based on the timing of the treatment relative to resection.

Prior to the complete response letter, an ongoing phase III trial designed through a special protocol assessment with the FDA had been examining the impact of the timing of treatment, as well as the effect of administering a second dose of apaziquone.

The multicenter, placebo-controlled, double-blind studies 611 and 612 had the same eligibility criteria and basic study design. The trials primarily included patients with Ta grade 1 or 2 nonmuscle invasive bladder cancer by central pathology review.

Patients were randomized to 4 mg of apaziquone in 40 mL diluent or placebo in 40 mL diluent. Treatment was administered within 6 hours of transurethral resection and retained in the bladder for 1 hour.

Patient demographics were well balanced between the treatment arms in both trials. The median age was 68 in study 611 and 67 in study 612. Across both trials, over 95% of patients were white and approximately 70% were male. Over 90% of patients in study 611 were from the United States, while 80% of patients in study 612 were from other countries.

At baseline, approximately two-thirds of patients with TaG1-2 disease in both arms of study 611 had 1 lesion, with the remaining patients having 2 to 4 lesions. Around eighty percent of patients in each arm had lesions that were all <3 cm. In study 612, approximately 60% of patients with TaG1-2 disease in both arms had 1 lesion, with the remainder of patients having 2 to 4 lesions. Around 87% of patients across the study had lesions that were all <3 cm.

In study 611, 38% (112/295) of patients with TaG1-2 disease in the apaziquone arm had disease recurrence compared with 44.6% (121/271) of patients in the placebo arm. The 6.6% difference favoring apaziquone was not statistically significant (odds ratio [OR], 0.76; 95% CI, 0.54-1.06;P= .11).

In study 612, 40.4% (114/282) of patients with TaG1-2 disease in the apaziquone arm had disease recurrence compared with 46.6% (139/298) of patients in the placebo arm. The 6.2% difference favoring apaziquone was not statistically significant (OR, 0.78; 95% CI, 0.56-1.08;P= .13).

The safety analyses included patients enrolled in the trials across all disease stages, not just TaG1-2. During study 611, there were 11 deaths in the apaziquone arm (N = 406) and 13 deaths in the placebo arm (N = 396). In its FDA briefing, Spectrum reported that none of the deaths seemed to be related to apaziquone treatment.

Discontinuations related to adverse events (AEs) occurred in 4 patients in the apaziquone cohort and 3 patients in the placebo group. The incidence of all-grade AEs was 80% and 75% in the apaziquone and control arms, respectively. This included grade 3/4 AE rates of 19% and 21%, respectively.

All-grade AEs occurring at higher rates in the apaziquone arm included dysuria (18%), urinary tract infection (18%), bladder pain/discomfort (9%), procedural pain (8%), and bladder spasm (7%). Serious AEs occurring more frequently in the apaziquone arm included atrial fibrillation (1.5%), chest pain (1.2%), COPD (1.2%), and cellulitis (1.2%).

In study 612, there were 14 deaths in both the apaziquone arm (N = 402) and the placebo arm (N = 411). As in the 611 trial, the deaths did not appear to be related to apaziquone treatment.

AE-related discontinuations occurred in 4 patients in the apaziquone cohort and 3 patients in the placebo group. The incidence of all-grade AEs was 80% and 81% in the apaziquone and control arms, respectively. This included grade 3/4 AE rates of 17% and 20%, respectively.