Risk of Progression or Death Reduced With Ivosidenib in Advanced Cholangiocarcinoma

The risk of progression or death was reduced by 63% with ivosidenib as treatment of pretreated patients with IDH1-mutant advanced cholangiocarcinoma compared with placebo, according to data from the phase III ClarIDHy study that was presented at the 2019 ESMO Congress.

Ghassan K. Abou-Alfa, MD

The risk of progression or death was reduced by 63% with ivosidenib (Tibsovo) as treatment of pretreated patients withIDH1-mutant advanced cholangiocarcinoma compared with placebo, according to data from the phase III ClarIDHy study that was presented at the 2019 ESMO Congress.

The median progression-free survival (PFS) was 2.7 versus 1.4 months, for ivosidenib and placebo, respectively (HR, 0.37; 95% CI, 0.25-0.54;P<.001). In addition to the primary endpoint of PFS, the IDH1 inhibitor also showed a non-statistically significant improvement in overall survival (OS). Despite more than half of patients crossing over (57%), there was a 31% trend toward reduction in the risk of death with ivosidenib (HR, 0.69; 95% CI, 0.44-1.10; P= .06).

"The significant improvement in PFS by central review, the favorable OS trend, and the tolerable safety profile all support the clinical benefit of ivosidenib in patients withIDH1-mutated cholangiocarcinoma," said lead investigator Ghassan K. Abou-Alfa, MD, from the Memorial Sloan Kettering Cancer Center. "This pivotal trial represents a practice-changing result where genomic testing would become the standard of care and of course, other than this drug itself, it would open the door to more target-specific trials in this rare indication, where limited progress has been made over the past decade."

In the ClarIDHy study, 185 patients withIDH1-mutant cholangiocarcinoma were randomized in a 2:1 ratio to oral ivosidenib at 500 mg daily (n = 124) or matched placebo (n = 61). Crossover from the placebo arm to ivosidenib was permitted following signs of radiographic progression. The primary endpoint of the study was PFS by blinded independent review, with OS, PFS by local review, safety, and response as key secondary endpoints.

Prior to study entry, patients had received 1 to 2 prior therapies, including 1 gemcitabine- or 5-FU—containing regimen. The median age of patients was 62 years, 91% had intrahepatic disease, and 92% had metastatic disease. Nearly half of patients (46%) had received 2 prior therapies, and the remainder received 1. All patients had confirmedIDH1mutations by next-generation sequencing and an ECOG performance score of 0 or 1. Slightly more patients in the placebo group had an ECOG score of 1 (67.2%) compared with the ivosidenib group (59.7%).

The partial response rate was 2.4% with the IDH1 inhibitor compared with 0% for placebo. Stable disease (SD) was experienced by 50.8% versus 27.9% of patients, for an overall disease control rate (PR plus SD) of 53% versus 28%, for ivosidenib and placebo, respectively.

At 6 months, 32% of patients in the ivosidenib arm remained alive and progression-free, compared with no progression-free patients in the placebo group. By month 12, the PFS rate was 22% in the investigational arm.

"The progression-free survival is only measured when you have a CT scan. In the curve I presented, the first scan happened at 6 weeks, which is where the separation of the curves begins to occur," said Abou-Alfa. "For that reason, the focus should not be the center of the median, even though it was positive, but more on the spread of the curve and the continuation of the spread and how it continued on at 22% at 12 months."

The median OS was 10.8 months with ivosidenib and 9.7 months for placebo. At 6 months, 67% of patients in the ivosidenib arm remained alive compared with 59% in the placebo group. By month 12, the OS rate was 48% with ivosidenib compared with 38% for placebo.

To adjust for the high rate of crossover between the arms, investigators performed an analysis looking at rank-preserving structural failure time (RPSFT) for placebo. In this analysis, the 6-month OS rate fell to 46% in the placebo group, and the median was 6 months. The 12-month OS rate was unavailable, as no more patients remained in the placebo arm at this point. When using the RPSFT data, there was a 54% reduction in the risk of death with ivosidenib, with medians of 6 versus 10.8 months, respectively (HR, 0.46; 95% CI, 0.28-0.75; P <.001).

“When looking at the medians the impact may not seem clinically meaningful but from my point of view this trial is practice-changing," said ESMO discussant Angela Lamarca, MD, PhD, MSc, from Christie NHS Foundation Trust. "When adjusting for crossover, going from 6 months to 10.8 months [for OS] is definitely clinically meaningful for patients with cholangiocarcinoma, who unfortunately have a very, very poor prognosis. We need to make sure all of our patients have access to the molecular profiling, which at the moment is a challenge in some of the countries.”

The most common treatment-emergent adverse events (AEs) in the ivosidenib and placebo groups, respectively, were nausea (35.5% vs 25.4%), diarrhea (30.6% vs 15.3%), fatigue (26.4% vs 16.9%), cough (20.7% vs 8.5%), abdominal pain (21.5% vs 13.6%), ascites (20.7% vs 15.3%), decreased appetite (19% vs 18.6%), anemia (14.9% vs 5.1%), and vomiting (19% vs 16.9%). Treatment-emergent AEs led to discontinuation for 8.5% of those in the placebo arm compared with 5.8% of those in the ivosidenib group.

Grade ≥3 treatment-emergent AEs were experienced 46.2% of patents in the ivosidenib group compared with 35.6% for placebo. The most common grade ≥3 treatment-emergent AEs in the ivosidenib and placebo groups, respectively, were ascites (7.7% vs 6.8%), bilirubin increase (5.8% vs 1.7%), anemia (5.1% vs 0%), and AST increase (5.1% vs 1.7%). “In my personal experience with the drug, it is very tolerable," said Abou-Alfa.

Agios Pharmaceuticals, the developer of ivosidenib, plans to submit a supplemental new drug application by the end of the year for ivosidenib in patients withIDH1-mutant previously treated cholangiocarcinoma. Ivosidenib is currently approved for the treatment of patients withIDH1-mutant acute myeloid leukemia.


Abou-Alfa GK, Maraculla TM, Javle M, et al. ClarIDHy: A global, phase 3, randomized, double-blind study of ivosidenib (IVO) vs placebo in patients with advanced cholangiocarcinoma (CC) with an isocitrate dehydrogenase 1 (IDH1) mutation. Presented at: 2019 ESMO Congress; September 27 to October 1, 2019; Barcelona, Spain. Abstract LBA10_PR.