COBALT Trial and PBC

The **COBALT Trial** was a significant clinical study aimed at evaluating the long-term clinical benefits of **obeticholic acid (OCA)** in the treatment of **primary biliary cholangitis (PBC)**, a chronic autoimmune liver disease. Below is a detailed overview of the COBALT trial and its implications for PBC management:

---

### **What is PBC?**

• **Primary biliary cholangitis (PBC)** is a chronic autoimmune liver disease characterized by the gradual destruction of bile ducts in the liver. This leads to bile accumulation, inflammation, and progressive liver damage, potentially resulting in cirrhosis, liver failure, or the need for a liver transplant.
• PBC disproportionately affects **middle-aged women** and has no known cure. The primary treatment options aim to slow disease progression and manage symptoms.

---

### **What is Obeticholic Acid (OCA)?**

• OCA is a synthetic bile acid analog that activates the **farnesoid X receptor (FXR)**, a key regulator of bile acid production and inflammation in the liver.
• It was **conditionally approved for PBC** under the accelerated approval pathway because it demonstrated biochemical improvements in **alkaline phosphatase (ALP)** levels, a surrogate marker of disease activity. However, confirmatory evidence of long-term clinical benefits was required.

---

### **Purpose of the COBALT Trial**

The COBALT trial was designed to:

1. Assess whether OCA provides **long-term clinical benefits** for PBC patients.

2. Confirm its ability to improve survival and reduce severe liver-related outcomes, such as liver transplantation, liver failure, or death.

---

### **COBALT Trial Design**

1. **Type of Study**:

• A **global, randomized, double-blind, placebo-controlled trial**.
• Included a **prespecified external control (EC) analysis** using real-world data to supplement the randomized trial.

2. **Participants**:

• 334 PBC patients from **27 countries** were enrolled.
• The majority were **middle-aged women** (mean age: 53 years).
• About **88% of participants** were already taking **ursodeoxycholic acid (UDCA)**, the standard first-line therapy for PBC.
• Patients with decompensated cirrhosis or other major liver conditions were excluded.

3. **Treatment Groups**:

• Patients were randomized to receive either **OCA (5–10 mg daily)** or a **placebo**.
• A **real-world external control group** was created using data from the Komodo Healthcare Map database, which included 1,051 matched PBC patients.

4. **Primary Endpoint**:

• The composite endpoint included time to death, liver transplantation, **MELD score ≥15**, uncontrolled ascites, or hospitalization due to hepatic decompensation events (e.g., variceal bleeding or encephalopathy).

---

### **Challenges Faced by the COBALT Trial**

1. **Recruitment and Retention Difficulties**:

• Once OCA became commercially available, it became challenging to recruit and retain patients in a long-term placebo-controlled study.
• Many patients in the placebo group switched to commercially available OCA, compromising trial blinding and integrity.

2. **Functional Unblinding**:

• Since **ALP levels** were routinely monitored, physicians and patients could infer treatment assignments. This led to **bias** and an increased likelihood of patients in the placebo group discontinuing or switching to open-label OCA.

3. **Trial Termination**:

• The trial was terminated early in 2021 due to **futility** and the practical challenges of maintaining a blinded, placebo-controlled design.

---

### **Key Results of the COBALT Trial**

1. **Randomized Arm Results**:

• In the **intention-to-treat analysis**, endpoint events occurred in **28.6% of OCA patients** and **28.9% of placebo patients** (hazard ratio [HR] = 1.01).
• This initially suggested no significant difference between OCA and placebo.

2. **Adjustments for Crossover and Censoring**:

• Adjustments using **inverse probability of censoring weighting (IPCW)** were applied to account for crossover and informative censoring.
• After these corrections, the hazard ratio shifted to **0.77**, suggesting a favorable benefit for OCA.

3. **External Control (EC) Analysis**:

• The EC analysis compared OCA-treated patients with 1,051 matched real-world patients.
• Results showed a **60% reduction in the risk of adverse clinical outcomes** (HR = 0.39; 95% CI: 0.22–0.69; P = 0.001), including death, liver transplantation, or hepatic decompensation.

4. **Safety and Tolerability**:

• OCA was generally well tolerated.
• The most common side effect was **pruritus (itching)**, reported in **78.6% of OCA patients** compared to **51.2% of placebo patients**.
• Other side effects included mild **edema**, **abdominal pain**, **nausea**, and **headache**.
• Serious adverse events occurred at similar rates in both groups (~32%).

5. **Hepatic Safety Profile**:

• Hepatic adverse events (e.g., elevated bilirubin, ascites, varices) were less frequent in the OCA group compared to the placebo group, alleviating earlier concerns about potential hepatotoxicity.

---

### **Implications of the COBALT Trial**

1. **Efficacy of OCA**:

• The trial demonstrated that OCA reduces the risk of serious liver-related events, supporting its long-term benefit in managing PBC.
• While the randomized portion of the trial did not show a clear difference due to biases, the adjusted and external control analyses consistently indicated significant benefits.

2. **Real-World Evidence (RWE)**:

• The use of real-world data in the EC analysis was pivotal in confirming OCA's efficacy.
• This highlights the potential of **hybrid trial designs** that combine randomized and real-world evidence, especially for rare diseases like PBC.

3. **Regulatory and Ethical Considerations**:

• The COBALT trial underscores the ethical challenges of maintaining placebo controls when effective commercial therapies exist.
• Future confirmatory trials may increasingly rely on real-world evidence to validate accelerated approvals.

4. **Lessons on Data Bias**:

• The trial revealed how **functional unblinding** and **differential dropout** can distort results in randomized trials.
• Advanced statistical corrections and external controls can help mitigate these biases and provide more accurate estimates of treatment effects.

---

### **Conclusion**

The COBALT trial reaffirmed the clinical value of OCA in PBC by demonstrating its ability to reduce the risk of serious liver events. Despite challenges with functional unblinding and trial termination, the use of real-world evidence and advanced statistical methods provided robust support for OCA's long-term benefits. The trial also highlighted the evolving role of hybrid trial designs in rare diseases, paving the way for more innovative approaches to confirmatory studies in the future.