Target trial emulation enables treatment comparisons based on real-world data

Target trial emulation (TTE) offers opportunities for comparing treatments using real world data. Its successful application requires a well defined research question, the expertise of an experienced data science team, and access to sufficiently large, high quality data sources.

Target trial emulation, or TTE, has recently attracted increasing interest among pharmaceutical companies. While the method has been discussed at industry congresses for several years, interest has accelerated as regulatory authorities in many countries have indicated an increased willingness to consider evidence generated through real‑world evidence (RWE) approaches in the decision‑making. RWE is expected to provide locally relevant insights to support regulatory and reimbursement decisions and to complement evidence from clinical trials.

According to Iiro Toppila, Data Analysis Lead at Medaffcon, TTE is particularly valuable when there is a need to directly compare two treatments using real‑world data, especially in cases where such comparisons have not been evaluated in randomised controlled trials (RCTs).

“Traditionally, such novel treatment comparisons have often relied on indirect methods, such as network meta‑analyses that combine data from multiple randomised clinical trials. TTE offers an alternative by enabling a direct comparison between two treatments using existing registry data and an analytical framework that emulates a randomised controlled trial,” Toppila explains.

TTE Requires Treatments to Be in Clinical Use

Another key application of TTE, according to Toppila, is the evaluation of different endpoints. The method can be used to assess for example treatment effectiveness and adverse events, compare treatments, and analyse their safety profiles. In addition, real‑world data typically enables substantially longer follow‑up periods than are feasible in conventional clinical trials.

A fundamental limitation across these use cases is that TTE requires the treatments to be already in clinical use. Without existing data, meaningful comparisons cannot be conducted.

“TTE is well suited for comparing treatments that are already available on the market. It cannot be applied in situations where no patients have yet received the treatment,” says Johanna Vikkula, Senior Data Scientist at Medaffcon.

New Opportunities in Rare Diseases

TTE also opens new possibilities for studying treatment effects in patient populations that have traditionally been difficult or impossible to investigate in clinical trials, such as patients with rare diseases.

“For example, we conducted a study comparing treatments in a specific liver indication among patients undergoing liver transplantation. Organ transplant requires always immunosuppressive treatment, and there are several cheap standard-of-care options available.  A dedicated clinical trial would never be feasible for such a rare patient population, studying practically free treatment options and long-term outcomes. By applying TTE principles, however, we were able to leverage data from long term European wide and US wide registers and generated meaningful evidence on outcomes in this specific subgroup in head-to-head comparison setting,” Toppila says.

Finland and the other Nordic countries provide an exceptionally strong environment for the application of TTE methodologies. The region benefits from comprehensive, high‑quality health registries and advanced analytical expertise.

“High‑quality Nordic registry data enables the construction of robust TTE models and virtual control arms. Although population sizes are relatively small in European or world-wide scale, the volume of data is still substantial, and its quality is exceptionally high,” Toppila notes.

The Research Question Should Drive Method Selection

Both Toppila and Vikkula emphasise that methodological choices should always be driven by the research question rather than by a particular analytical tool.

“TTE is one tool within a broader methodological toolbox. Once the research question is clearly defined, we can determine which method or combination of methods is best suited to address it,” Toppila explains.

“If the research question is poorly defined, there is a real risk that the analysis will focus on the wrong patient population, use an inappropriate study design, or apply unsuitable methods,” Vikkula adds.

The effective use of TTE requires close collaboration with an experienced data science team.

“Our role often also involves setting realistic boundaries on what conclusions can be drawn from different study designs and data sources. With retrospective RWE data, it is easy to overinterpret findings, and comparisons can be misleading if confounding factors are not carefully addressed. If the objective is to make genuinely causal inferences about cause‑and‑effect relationships, strong methodological expertise and the judgment of experienced data scientists are essential,” Toppila concludes.