BBC Panorama – 6 September 2010

Issued: London UK

Patient safety is our first priority. We strongly refute any allegation that our actions have put patients at risk.

We have carried out an extensive research programme, involving more than 50,000 patients to analyse the safety and benefits of Avandia.

All the results from our research programme have been given to the regulatory authorities worldwide and we are fully committed to being open about the results of all our clinical research.

In 2006, the company conducted its own meta-analysis on Avandia and we provided the results to regulators and posted them on our website for researchers and medical professionals to see.

Dr Nissen’s own meta-analysis was published in 2007. The results of this analysis were consistent with our own findings, which were publicly disclosed 6 months earlier.

We regret that comments made by GSK during a meeting with Dr Nissen might be misinterpreted as seeking to stifle an independent view of the science around Avandia.

Our view remains that whilst a meta-analysis of this type can be useful for raising scientific questions, it cannot definitively determine risk.

Controlled clinical trials are the most rigorous form of scientific evaluation that can be used to assess the benefits and risks of medicines.

The data from these types of trials have shown that Avandia does not increase the overall risk of heart attack, stroke or death.

We continue to believe that Avandia is safe and effective when it is prescribed appropriately.

GlaxoSmithKline – one of the world’s leading research-based pharmaceutical and healthcare companies – is committed to improving the quality of human life by enabling people to do more, feel better and live longer. For further information please visit www.gsk.com.

Allegations that GSK attempted to suppress Dr Nissen’s findings

In 2007, Dr. Nissen contacted GSK during the development of his meta-analysis to request information. We directed Dr Nissen to the wealth of information contained within our Clinical Study Register. The site includes study protocols and result summaries, including the results of our own meta-analysis of 42 double-blind, controlled clinical trials of Avandia posted in 2006.

Prior to the publication of Dr Nissen’s findings, GSK received an unsolicited copy of the manuscript. Upon reviewing the manuscript, GSK disagreed, and continues to disagree with Dr. Nissen that a meta-analysis of this type can definitively determine risk.

Inadequate warnings for patients with congestive heart failure

Congestive heart failure (CHF) is a well characterised side-effect associated with both of the drugs in the thiazolidinedione class, Avandia and Actos, and this is clearly indicated in the approved labelling for these products. The absence of head-to-head clinical data precludes an accurate estimate of the relative difference in the incidence of CHF between these two medicines.

Evidence of additional risk to patients receiving Avandia compared to Actos

In the absence of large randomised head-to-head results that directly compare Avandia to Actos, definitive statements on the relative risk between Avandia and Actos cannot be made. Two separate studies of Avandia and Actos, known as RECORD and PROActive, demonstrated no increase in the overall risk of heart attack, stroke or death with either medicine.

Criticism of the RECORD study

The RECORD study is the largest and longest prospective clinical trial examining Avandia’s cardiovascular safety. RECORD was initiated in 2001 and, in consultation with European regulators, was designed to compare cardiovascular (CV) outcomes of patients on Avandia added to two commonly used medicines for type 2 diabetes.

An inspection of the RECORD study by the US regulator concluded that there was no evidence “of systemic or pervasive findings that would undermine the reliability of the RECORD data”. However, due to the size and scope of the study, the inspectors acknowledged that there were limitations to what could be evaluated.

They reported:

• Procedures, resources and training for the conduct of the trial were adequate;
• No evidence of tampering with or falsification of data;
• No evidence of serious violations;
• No evidence that publication of the interim CV analysis in June 2007 influenced the conduct of the trial. However, the endorsement to conduct an interim CV analysis by the two members of the Steering committee and two members of the data safety monitoring board did not fulfil the requirements of the charters of these committees.

Internal GSK emails suggesting that GSK was hiding negative data regarding Avandia

The internal emails referenced were selectively disclosed by lawyers seeking damages against the company from 14 million pages of documents. They include drafts and other documents taken out of context, which therefore are incomplete and misleading. The documents do not offer new clinically relevant scientific information, and do not provide the full extent of GSK’s studies and evaluation of Avandia.

The study cited within the internal emails is known as Study 175 and was conducted in 1999. It is a small, short term study designed to examine the effect of Actos, the other medicine in the same class as Avandia, on lipids. This study did not involve Avandia. There were no data on outcomes, such as heart attacks, in this study and the assertion that this study informed GSK's views about heart attacks and Avandia is completely unfounded.

The findings of this study were no different from the available published data on lipids that were known and included in the Actos label at that time. The effect of Avandia on lipids was discussed at the 1999 Advisory Committee Meeting and was included in the label at approval.

With respect to the two other studies referenced (079 and 096), GSK submitted the data from these studies to the FDA in 1999. The data from these studies was made publicly available on GSK’s Clinical Trial Registry.

Disclosure by GSK of payments to healthcare professionals

Last year, GSK began publishing payments made in the US to healthcare professionals on an individual, named basis for the provision of scientific and medical expertise. As a company we are not required to disclose this information, but have taken this proactive measure as part of ongoing efforts to increase transparency.

We have committed to publishing further information during 2010 on payments made in the UK and other regions, however initially the same level of detailed reporting will not be possible and we will publish aggregate information per country. We are developing our systems to support this in the future, and earlier this year, GSK wrote to all the healthcare professionals with whom we work, to explain that we are revising their contracts, as required by law, in order to reserve the right to publish this payment information.

Healthcare professionals bring expert knowledge and perspective from their clinical and management experience which they share with us and other healthcare professionals to support improvements in patient care. These services are valuable to GSK and we believe these professionals should be fairly compensated for the services and expertise they provide.

Allegations that Avandia provides little benefit to patients and could cause an additional burden to the NHS

Within the UK there are over 2 million people living with type 2 diabetes. Diabetes patients often have difficulty controlling their disease and physicians need a range of options to enable them to provide the best management possible, often prescribing two or three medicines together to help patients control and maintain their blood sugar levels. If left uncontrolled, type 2 diabetes can lead to a variety of serious health problems, including heart disease, stroke, kidney failure and eye damage.

The management of diabetes in the UK is estimated to cost £286 a second; representing 10% of the overall NHS budget spend.

Avandia has been shown to be effective at helping control blood sugar in patients with type 2 diabetes and is the only oral anti-diabetic that has been shown to sustain blood sugar control for up to five years – and this is compared to the most widely used diabetes medicines today.