Cyramza OS Clinical 9.6 months vs. RWE 7.9 months

HIRA discloses study result on RWE platform

Kadcyla data insufficient, PFS analysis finds 0.4 months shorter

Apparently a study in South Korea comparing the clinical trial result and real world data has found a targeted therapy for stomach cancer Cyramza (ramucirumab) and a targeted therapy for breast cancer Kadcyla (trastuzumab) have shorter overall survival in the real world.

 

The clinical trial report on Cyramza stated its overall survival (OS) reached 9.6 months (8.5-10.8), but a statistic claim billing analysis by the Health Insurance Review and Assessment Service (HIRA) found the figure to be at 7.9 months (6.3-8.4), and it marked 10.0 months (9.3-10.7) in an analysis on medical record archived by research participating hospitals.

 

As for Kadcyla with insufficient period of monitoring, patients’ progression-free survival (PFS) was compared and found the initial clinical trial result was at 9.6 months, which was 1.6 months longer than the actual medical record analysis at 8.0 months (6.8-9.4).

 

The said gaps between clinical trial result and real-world data (RWD) was verified in the Retrospective Study on Designing the Real-world Evidence (RWE) Platform for Improved Pharmaceutical Reimbursement Management (Principle investigator: Dr.

 

Kim Dong-suk and Dr.

 

Byun Ji-hye) published by HIRA on Nov.

 

12.

 

HIRA selected Lilly’s Cyramza (ramucirumab) and Roche’s Kadcyla (trastuzumab) as subjects for RWD-basis outcome study to generate RWE.

 

The research was conducted as randomized controlled trial (RCT), the same method the pharmaceutical companies have used, when listed the drugs, by collecting and analyzing medical records from 242 healthcare providers (68 hospitals) associated with Korean Cancer Study Group.

 

◆Cyramza: The RCT evidence used when listing for the healthcare reimbursement compared the paclitaxel single therapy against Cyramza in combination with paclitaxel, which confirmed the combination therapy’s efficacy, statistically.

 

Based on the billing data collected from May 2018—the point of Cyramza listing—through December 2018, the fatality rate in patients with a record of Cyramza prescription, from the point of first stomach cancer diagnosis to death, was reviewed.

 

As a result, 1,023 out of 1,419 patients have died.

 

The median OS was approximately 240 days (7.9 months), which was shorter than 9.6 months from the original RCT.

 

The majority of the patients, or around 55% of them, were administered with the drug four times or less, followed by the number of patients who used the drug five to six times and over 10 times.

 

The average number of administration marked 5.3 times among all patients.

 

Analyzing the impact of variables affecting the OS with Cox proportional hazards model, the OS was statistically better when the diagnosis was given over a year ago, administration period was over 12 weeks, the number of administration was over seven times, and when it was at tertiary hospitals.

 

The outcome study on electronic medical record (EMR) analyzed patient survival after using Cyramza with Kaplan-Meier estimator.

 

The study concluded the time to progression (TTP) to be at 4.2 months, PFS at 4.03 months and OS at 10.03 months.

 

The RWE based on these RWD confirms Cyramza in combination with paclitaxel showed shorter OS, compared to the original RCT.

 

◆ Kadcyla: Apparently, 331 out of 1,009 patients have died among the patients who were administered with Kadcyla at least once for year since the point of listing in August 2017 through July 2018.

 

The median OS was approximately 449 days (15 months), which was shorter than both 29.9 months (Dieras et al, 2017) and 22.7 months (Krop et al, 2017) demonstrated in the company’s RCTs.

 

But HIRA said there is a limitation with the insufficient time of monitoring as the drug received the reimbursement only from August 2017.

 

Even with EMR data, the PFS in patients who used Kadcyla was at 6.8 months (6.2-7.4), but again the monitoring period was too short to confirm median OS.