'25 at 25': Individual participant data (IPD) meta-analysis

15 Apr 2024

For more than 25 years, the Meta-analysis Programme at the MRC Clinical Trials Unit at UCL (and formerly the MRC Cancer Trials Office) has had a strong international reputation for the conduct of large, collaborative systematic reviews and meta-analyses based on individual participant data (IPD).

As part of our ‘25 at 25’ series to celebrate the Unit’s 25th anniversary, this piece explores how our IPD meta-analyses have influenced the treatment of patients across a broad range of conditions, and how our methodology research has pushed the science of meta-analysis forward.

Systematic reviews and meta-analysis are essential tools for rigorously evaluating the effects of treatments, informing clinical practice and subsequent research. Where individual randomised controlled trials (RCTs) are too small to provide convincing evidence about treatment effects, or their results are inconsistent or unclear, systematic reviews and meta-analyses bring together evidence from similar trials to estimate treatment effects more precisely.

These systematic reviews and meta-analyses are usually based on aggregate data extracted from trial publications, but data can be limited and meta-analysis findings prone to reporting biases. Instead, meta-analyses based on IPD involve the collection of the original trial data on each individual participant from each relevant trial. This gives better and more consistent data, greater scope in the analyses and more reliable and detailed findings.

The Meta-analysis Group at the MRC CTU at UCL was one of the first groups worldwide to regularly use IPD in meta-analyses. They have led or collaborated on 22 IPD meta-analyses of 325 RCTs involving around 93,500 participants in total. To date, these have influenced more than 530 national and international practice guidelines, impacting on the treatment of patients worldwide.

Should people with muscle-invasive bladder cancer receive chemotherapy?

Invasive bladder cancer is where the tumour has invaded the bladder’s muscle wall but has not spread elsewhere in the body. Previously, the standard treatment was surgery to remove the bladder and surrounding tissues and/or radiotherapy directed at the tumour. However, doctors disagreed over whether patients should also receive chemotherapy before these other treatments. Several RCTs had attempted to answer this question, but gave inconclusive results.

In 2005, the MRC CTU at UCL published a meta-analysis based on IPD from 11 RCTs and 3,005 participants. It showed that patients who received chemotherapy before surgery and/or radiotherapy lived longer than those who had surgery and/or radiotherapy only. These results have since informed 55 treatment guidelines and the UK’s National Institute for Health and Care Excellence (NICE) recommends that anyone diagnosed with invasive bladder cancer is offered chemotherapy before local treatments.

During the same year, the Unit published another IPD meta-analysis, looking at the effect of giving chemotherapy after surgery and local treatments for invasive bladder cancer. While the results were promising, more trial evidence was still needed.

In 2022, the Meta-analysis Group returned to the question of post-operative chemotherapy, publishing an update of this IPD meta-analysis. With 10 RCTs and more than double the number of participants (1,183 in total) now included, the findings provided clear evidence that giving chemotherapy after other treatments also helps patients live longer, compared with surgery and/or radiotherapy alone.

The results of these two IPD meta-analyses give doctors and patients a greater choice of when to give chemotherapy, depending on their circumstances and preferences.

Do antiplatelet agents prevent pre-eclampsia during pregnancy?

Pre-eclampsia is a condition that can occur during pregnancy, characterised by high blood pressure and often protein in the urine. For some women, it can lead to serious complications and danger for themselves and their baby. Women who develop pre-eclampsia in pregnancy also have an increased risk of high blood pressure, stroke or heart disease in later life.

Antiplatelet agents reduce the body’s ability to form blood clots and can therefore help lower blood pressure. The ability of antiplatelet agents, for example low-dose aspirin, to prevent or delay pre-eclampsia had been widely investigated in trials, but it was not clear how effective they were. Therefore, a collaboration between the MRC CTU at UCL and the Universities of Sydney and Oxford undertook an IPD meta-analysis to assess the use of antiplatelet agents and explore if some women might benefit from them more than others.

The team collected IPD on 32,217 participants and their 32,819 babies, from 31 RCTs across 25 countries worldwide. They found that women who were assigned to an antiplatelet agent were 10% less likely to develop pre-eclampsia or deliver their baby before 34 weeks than those who were not assigned this treatment. Antiplatelets were beneficial to all women, irrespective of, for example, their age or whether it was their first or subsequent pregnancy.

These results went on to inform the UK NICE treatment guidelines for hypertension in pregnancy.

Does chemotherapy help people with non-small cell lung cancer live longer?

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, comprising around 85% of all cases. For patients with disease that was only present in the lung, surgery to remove the tumour was the main treatment. Those with disease that had spread near the lung were usually given radiotherapy. For disease that had spread further around the body, patients were only given treatment for their symptoms, known as supportive care. Despite decades of research, outcomes for patients with NSCLC remained poor.

Our prior IPD meta-analyses, published in 1995, included 52 trials and nearly 10,000 patients. These studies showed the potential of chemotherapy to help patients with more advanced stages of NSCLC live longer, but the effects in those with earlier stages of the disease were less clear.

Many large trials looking at the effects of chemotherapy followed as a result of these meta-analyses. Therefore, the MRC CTU at UCL initiated a series of four updates and three new IPD meta-analyses with collaborators at the Institut Gustave-Roussy in France. Including more than 100 trials and 23,000 patients and published between 2008 and 2014, these studies have comprehensively defined the role of chemotherapy in treating NSCLC.

The series of meta-analyses found that adding chemotherapy to the best supportive care improves survival of patients with advanced disease, and that giving chemotherapy before or after surgery improves survival in patients with operable NSCLC. They also showed that administering chemotherapy alongside radiotherapy is better than giving it before or after.

These projects have influenced numerous treatment guidelines internationally, such that chemotherapy is now part of the ‘backbone’ of treatment for the majority of patients with NSCLC.

Which prostate cancer patients benefit most from docetaxel?

Several RCTs, including the MRC CTU at UCL’s STAMPEDE trial, have shown that adding the chemotherapy drug docetaxel to standard hormone therapy helps people with advanced prostate cancer live longer. However, uncertainty remained about precisely who gets the biggest benefit from docetaxel. A big advantage of IPD meta-analysis is being able to work out if some groups benefit more or less from a treatment.

STOPCAP M1 is a worldwide collaboration, led by MRC CTU at UCL, which aims to identify which treatments work best for people with hormone-sensitive prostate cancer which has spread elsewhere in the body. The results of its first IPD meta-analysis were published in 2023, showing that docetaxel was most effective in those whose cancer had spread further throughout their body, including their liver or lungs, or at least four locations in the bones. Potentially, there was an even bigger effect if they also had a larger tumour in their prostate.

On the other hand, participants whose cancer had not spread very far did not benefit at all from docetaxel, and were equally likely to live for at least five years whether they were treated with docetaxel or not.

This STOPCAP M1 meta-analysis is an important step towards making existing prostate cancer treatments more personalised, allowing doctors and patients make informed decisions about when to use docetaxel. This means patients who are unlikely to benefit from docetaxel can be spared its side effects and be given an alternative treatment instead.

Advancing the science of IPD meta-analysis:

Despite being widely regarded as the gold standard for meta-analysis, IPD is still less commonly used than other types of systematic review. Using IPD improves the quality of the data and reliability of the analysis, but it relies on extensive collaboration between researchers, who must persuade trial teams to share their data. It also takes more time and resources to prepare and check the original data.

As well as conducting IPD meta-analyses, our researchers provide advice, training and publish on the practical and statistical methodology underlying IPD meta-analyses. This includes co-editing and authoring the first handbook on IPD meta-analysis.

Further information: