skip to main content
Close Icon We use cookies to improve your website experience.  To learn about our use of cookies and how you can manage your cookie settings, please see our Cookie Policy.  By continuing to use the website, you consent to our use of cookies.
Global Search Configuration

AVEO's Fotivda made it to the US market only after the FDA resolved lingering concerns about the drug's impact on overall survival.

Questions about the impact of AVEO Pharmaceuticals, Inc.'s Fotivda (tivozanib) on overall survival in renal cell carcinoma continued to plague the kinase inhibitor in its second Phase III study and threatened to once again derail the drug’s regulatory path to market in the US.

After repeatedly advising AVEO not to submit for approval on the basis of interim overall survival data from TIVO-3 that suggested a negative trend, the US Food and Drug Administration ultimately concluded the final survival analysis did not suggest tivozanib had a detrimental effect, even though median overall survival was shorter than for the comparator agent, Bayer Health Care’s Nexavar (sorafenib).

The agency’s final conclusions on overall survival followed multiple changes in AVEO’s statistical analysis plan and exploratory analyses on the FDA’s part.

In granting approval, FDA reviewers said they were persuaded by the totality of the TIVO-3 data, including the statistically significant advantage in progression-free survival, supportive objective response rate data and an overall survival hazard ratio that, in the final analysis, did not demonstrate a decrement in RCC patients who received prior treatment with two lines of therapy.

The final conclusions on OS did not come easily but, rather, followed multiple changes in AVEO’s statistical analysis plan and exploratory analyses on the FDA’s part.

Initially, only one interim OS analysis was planned to be conducted at the time of the primary PFS analysis. However, this first OS analysis was unfavorable to tivozanib, and the FDA recommended against NDA submission at that time. The statistical analysis plan subsequently was amended several times, and the final version included two interim analyses and a new final analysis with longer follow-up.

Although the estimate of median OS was longer in the sorafenib arm on final analysis, “the median OS is thought to represent only one timepoint on the survival curve and not the overall curve,” the agency’s multidisciplinary review states. Citing the small difference observed between the two treatment arms in a post-hoc exploratory analysis, the overlapping survival curves, and a hazard ratio point estimate that favored tivozanib, “the review team concluded that overall, no decrement in overall survival was observed for patients on the tivozanib arm.”

The PFS and OS results also were interpreted in the setting of an active control, sorafenib, and consistent efficacy with tivozanib was demonstrated in the approximately 25% of patients who were pretreated with a checkpoint inhibitor. (See sidebar for story.)

However, the agency did not give AVEO everything it wanted. The sponsor had sought a broad indication that included all patients with advanced renal cell carcinoma, but the FDA limited the approved use to the same third-line setting studied in TIVO-3.

Complete Response Letter ...

The FDA approved tivozanib on 10 March for treatment of patients with relapsed or refractory advanced RCC following two or more prior systemic therapies. The US clearance came three-and-a-half years after the European Commission granted marketing authorization in August 2017, and eight-and-a-half years after AVEO submitted its first new drug application for the new molecular entity. (28 New Drug Approvals In EU Cancer Dominates But RA Skin Blood Disorders Well Served Too)

In September 2012, the company submitted an NDA for treatment of advanced RCC on the basis of a lone Phase III trial. The TIVO-1 study of patients with treatment-naïve advanced RCC met its PFS primary endpoint. However, the results for OS, a secondary endpoint, showed an unfavorable trend for tivozanib relative to sorafenib, the comparator agent. (Approval Or New Study Thats The Question For Tivozanib Panel Review)

AVEO attributed the negative OS trend to: the study’s one-way crossover design; the geographic distribution of patients, most of whom were from outside the US; and extensive use of subsequent targeted therapy by patients in the sorafenib arm.

The company submitted the original NDA against the advice of the agency, which had recommended conducting a second Phase III trial due to the negative OS trend in TIVO-1.

During a May 2013 review by the Oncologic Drugs Advisory Committee marked by blistering criticism of the TIVO-1 study design and execution, the agency’s external experts were nearly unanimous in voting that tivozanib failed to demonstrate a favorable benefit-risk profile in an adequate and well-controlled trial. (Aveos Tivozanib Falls On Survival Data Study Design Flaws)

The negative ODAC review was followed a month later by an FDA complete response letter. (Tivozanib Complete Response Shows Increased Communications Cant Always Save An NDA)

In the CRL, the agency said inconsistent PFS and OS results and an imbalance in post-study treatments make the TIVO-1 results uninterpretable and inconclusive for purposes of a benefit-risk assessment necessary for drug approval. The agency recommended AVEO conduct an additional trial using PFS as the primary endpoint and OS as a secondary endpoint. The trial should be powered to detect a difference in PFS and adequately sized to ensure that there is no adverse effect on OS, with the results applicable to the US population, the agency said.

... Leads To TIVO-3

In that new trial, TIVO-3, 350 advanced RCC patients who had received at least two prior lines of systemic therapy were randomized to tivozanib or sorafenib. Tivozanib once again demonstrated a statistically significant improvement on the PFS primary endpoint (HR 0.73, [95% CI: 0.56, 0.95]). The median PFS estimates for the tivozanib and sorafenib arms were 5.59 months and 3.88 months, respectively.

Consistent PFS result were observed across prespecified subgroups. Objective response rates were 18% and 8% for tivozanib and sorafenib, respectively.

“Progression-free survival is considered by FDA to be an acceptable endpoint to demonstrate the effectiveness of new therapeutics across many oncology indications and has been the primary basis for approval of therapies indicated for the treatment of advanced renal cell carcinoma,” the agency’s multi-disciplinary review states. (See list of reviewers at end of story.)

“Although the overall difference in median PFS between arms was 1.71 months, which is shorter than the eight-week scanning interval on the trial, a post-hoc interval censoring analysis conducted by FDA to explore the robustness of these results with respect to the imaging assessment interval was consistent with the primary analysis,” the review states. “Because the study was a replacement trial design compared to standard of care therapy, the benefit in PFS was deemed appropriate to support substantial evidence of effectiveness.”

Changes To Survival Analyses Plans

However, the OS results proved more problematic to interpret.

During a January 2019 pre-NDA meeting, the agency recommended against submission of an application at that time due to an unfavorable trend in the interim OS analysis of TIVO-3.

The story continues after the table of OS analyses results....

TIVO-3 Overall Survival Results



First Interim Analysis (183 death events)

Median OS (months)



Hazard Ratio (95% CI)

1.12 (0.84, 1.51)

Second Interim Analysis (227 death events)

Median OS (months)



HR (95% CI)

0.99 (95% CI: 0.76, 1.29; p=0.95)

Final Analysis (251 death events as of 1 May 2020)

Median OS (months)



HR (95% CI)

0.97 (0.75, 1.24)

Source: FDA review documents


“The results of TIVO-3 do not allay concerns about the potential detriment in OS outlined in the complete response letter dated June 6, 2013,” the agency said in a written response ahead of the meeting. “Because of the potential decrement in OS demonstrated in two trials, patients currently receiving tivozanib on clinical trials (both sponsor- and investigator-initiated) should be re-consented and information provided concerning OS for both trials.”

This initial unfavorable OS analysis prompted the European Medicines Agency’s Committee for Medicinal Products for Human Use to suggest it could take regulatory action if the negative survival trend continued in subsequent analyses. (Aveos Tivozanib Under CHMP Spotlight If Negative OS Trend Continues In TIVO3)

Drug Review Profiles

This two-part analysis of Fotivda’s review by the US FDA is the latest Pink Sheet Drug Review Profile. Our exclusive profiles take a deep-dive look at the data and dynamics behind an interesting recent US drug approval, with lessons learned for other sponsors.

In a February 2019 amendment to the statistical analysis plan, AVEO changed the planned final OS analysis to a second interim analysis, to be conducted when all patients had been on study for two years. An additional final OS analysis was added and planned to occur when 75% of patients would experience an event, for a total of 263 OS events.

Another pre-NDA meeting was held in October 2019 to discuss the results of the second interim OS analysis. In preliminary comments, the agency once again recommended the sponsor not submit an application at this time.

“We remain concerned about the results of TIVO3 in the context of the overall development of the drug,” the agency said, according to meeting minutes. “Your current interim OS results do not abrogate concerns over detriment and may worsen with final analysis at 263 events. The median overall survival for tivozanib is worse than that of sorafenib.”

In response, AVEO said that based on statistical modeling assumptions and durability of the PFS advantage over sorafenib, the company remains confident that the final OS analysis will not worsen with additional follow-up.

The sponsor sought the FDA’s agreement on a plan to submit the NDA “as planned” in December 2019, with updated OS data to be submitted in June 2020 to confirm no adverse effect on OS.

“Ultimately the choice to submit the data is the sponsor’s, however an ODAC discussion will likely be required,” the agency responded. “Should the sponsor wish to proceed with a revised OS analysis in June 2020, they should submit an updated SAP with a planned OS update based on the projected number of events at that time.”

The statistical analysis plan was amended again in November 2019, and the timing of the final OS analysis was changed from an event-driven analysis to a specific date – 1 May 2020 – based on an estimate of approximately 263 events by that time.

Rather than resubmitting the drug under the prior NDA, the company submitted an entirely new application on 31 March 2020 for treatment of patients with relapsed or refractory RCC.

Restricted Mean Survival Time Analysis

In the final OS analysis, the hazard ratio point estimate is close to 1 and the median estimate of the tivozanib arm is approximately three months shorter than that of the sorafenib arm, the FDA's multidisciplinary review states. “Therefore, possible detrimental effect of tivozanib on OS is a concern.”

Given that the median is only one timepoint on the survival curve and the difference at one timepoint may not reflect the overall curves, the review team conducted a post-hoc exploratory analysis using the restricted mean survival time (RMST) method to estimate the treatment effect. “Using a cutoff time at month 44 for RMST estimation, the difference of RMST estimate between tivozanib and sorafenib is -0.4 months (95% CI: -3.7, 2.9),” the agency said.

“Given that the Kaplan-Meier plot shows overlapping survival curves, a hazard ratio less than one in favor of the tivozanib arm, and a small difference of OS between the two arms observed using the RMST approach, it appears that tivozanib does not have an apparent detrimental effect on OS.” – FDA

Reviewers also explored RMST estimates at different cutoff timepoints. “Overall, the point estimates of RMST differences using various cutoffs are between -0.51 and 0.02 months (negative values favor sorafenib) and the confidence intervals are wide,” the review states. “Given that the Kaplan-Meier plot shows overlapping survival curves, a hazard ratio less than one in favor of the tivozanib arm, and a small difference of OS between the two arms observed using the RMST approach, it appears that tivozanib does not have an apparent detrimental effect on OS.”

The review team also performed additional sensitivity analyses on final OS. The unstratified OS analysis had a hazard ratio point estimate of 1.003 (95% CI: 0.78, 1.28) and the analysis stratified by stratification factors data collected on electronic case report forms had a hazard ratio point estimate of 0.97 (95% CI: 0.75, 1.25). “Results are similar to the primary findings of OS,” the agency said.

The agency reviewers’ interpretation of the OS results also was influenced by the efficacy of the active comparator used in TIVO-3.

“The FDA also notes that the overall survival analysis of tivozanib are vs. sorafenib, which represents an active therapeutic option in this setting,” the review states. Sorafenib’s approval in advanced RCC was based on a placebo-controlled Phase III trial in which the later-reported OS hazard ratio estimate was 0.88 (95% CI: 0.74, 1.04) favoring sorafenib despite crossover. “Thus, these results are interpreted through that perspective as well.”

‘Improved Tolerability’

The FDA review team found tivozanib’s safety profile to be “within range” for the RCC patient population and similar to that of other kinase inhibitors.

Overall, patients receiving tivozanib in TIVO-3 had lower rates of dose reduction, interruption or permanent discontinuation relative to those in the sorafenib arm.

No patient-reported outcomes data were collected in TIVO-3. However, PRO data from the TIVO-1 study in the front-line setting suggested an improvement in the analysis of overall treatment bother results, the FDA said. “This suggests an overall improved tolerability over an accepted standard-of-care according to several clinically relevant metrics.”

The FDA also found tivozanib’s toxicity profile to be consistent with other drugs affecting the VEGF pathway, “which increases the familiarity of providers with risk management practices in this patient population,” the review states.


Read also


Next steps

Whether you’re a small biotech start-up, research firm, generic manufacturer or a global pharmaceutical giant, you need focused, independent insight and opinion on market developments.

Our team is ready to hear from you for a particular request or area of interest. Please do not hesitate to reach out and discuss.

Contact us for product technical and account support.

  • US Toll-Free   : +1 888 670 8900 
  • US Toll             : +1 212-600-3520
  • UK & Europe : +44 (0) 208 052 0700

Have an immediate and specific information need?

Browse and buy from 1000s of analysis and research reports now: