Avastin in ophthalmology Briefing for CCGs

I keep getting asked
Here’s my view 

No referencing – email me if u want refs

NB my COI statement at bottom

1) Introduction

Treatment with a drug known as Lucentis® (Ranibizumab) or Eylea (aflibercept) is now recommended best practice in many ophthalmic diseases. This treatment prevents or delays sight loss in over 90% of recipients when given as injections for periods of up to two years. 

Some discussion has taken place on the use of avastin in ophthalmic conditions. This paper sets out some considerations. 

There are many and complex legal and regulatory issues surrounding this issue. These are set out in the recent BMJ investigation. NHS Clinical Commissioners has recently entered into this debate calling for a number of parties to work together to address this problem. NHSCC have already set out a set of Q&A on this matter. There were a few additional issues that may arise. 
2) Licence and regulatory framework

Avastin has market authorisation in this indication, licensed in cancer.

It doesn’t have market authorisation in ophthalmology. We all know this is a commercial decision by Roche and their subsidiary, Genentech. Were this a drug that Roche had commercial interest in getting licence it would have done so

IVAN and CATT were set up as trial to mimic a PIII licensing study. And more than adequately passed the bar – against an active comparator, not a placebo

Similar study now in DMO (NEJM late 2014) and LEAVO will do similar in Vein occlusion 

Given the good quality evidence of efficacy and safety simple cost minimisation would suggest that Avastin is the drug of choice in ophthalmic indications where aVEGF is used. 

In the UK, the MHRA consider Avastin unlicensed in ophthalmic use. The FDA and EMA consider it to be an off label use of a licensed medicine. There is valid debate to be had whether the EMA position is actually the “official” position given that the licence is Europe wide. EMA have expressed an opinion that the term “unlicensed” has no meaning in law.
Prescribing of unsafe unlicensed drugs undermines the regulatory system”.

Many have made accusations about commissioners seeking to give preferential weight to “unsafe unlicensed drugs”. There are many that would level an accusation that this is about NHS commissioners “trying to do things on the cheap” and save money by using “unsafe” and unlicensed drugs.

All commissioners are conscious of the history of the licensing process, why it exists and the safeguards it brings for patient safety. This is not in dispute. It would seem that Bev has not been brought to market for commercial reasons, not safety reasons. This represents significant lost value for the broader population who will bear the opportunity cost of the investment in Lucentis and Eylea. One must ask whether it is efficient, fair or equitable that patients with other diseases bear this opportunity cost. 

Turning the initial accusation about “doing things on the cheap” on its head to frame it as a question of “how many sight years do you want to buy for a very limited budget”.

Finally on this issue, it is worth reflecting that the most likely reason Roche has not sought Market Authorisation for Avastin is a commercial not a clinical one. The safety argument per se was settled with the Cochrane review. 

The views and likely positions of a number of influential bodies are set out in the BMJ Investigation. It is unknown whether any of these views will change in the wake of the NHSCC pressure and the BMJ investigation. 

GMC position

Slight change to GMC guidance on px. Less sabre rattling in my view. http://www.gmc-uk.org/guidance/28349.asp

Seems down toned compared to previous guidance to me

I wonder whether the GMC position about threat rather than action. Can GMC credibly take an individual doctor to task, if she has institutional backing, especially in circumstances where avastin is the drug of choice in private practice.

In summary:

  • Avastin has an EMA licence, thus license is European 
  • EMA view = off label use of licensed indication
  • MHRA view = unlicensed. 
  • Whether MHRA view has primacy is (highly) debatable
  • Till we actually leave the EU, EMA remains the regulator.

3) Legal

The issue that it is illegal to make the determination of “special need” a priori for a population- it must be done individual by individual. 

There are multiple interpretations of this case law. David Lock QC has published his view in the BMJ. This updates a previous opinion published as a BMJ Blog. Poland v ECJ was about importation not prescribing. 

Novartis and JR threat

Novartis had previously indicated they would take the JR route.

SHIP PCT cluster settled out of court. Terms undisclosed. I’m led to believe the judge wanted the case to be heard in open court as there was a legitimate public interest to be heard. Novartis didn’t want this to happen. Judge was of opinion this was a case about right to make a profit not a safety issue. 

Threat of legal action remains – though may be lower as Novartis now have other drugs and their profit not so dominated by a single medicine. Bayer also have a wider product line. Difficult to judge this one.
Letter from Freeman to CCGs saying use of avastin is illegal.

This seems to be disputed by the advice published by David Lock QC in the BMJ, at same time as the Cohen investigation.

There is no case law in this respect.
Novartis v Poland

Much is made of the Poland case. Can be argued to be a red herring as that was about importation not prescribing or injecting.

My understanding  is that the Polish case involved the use of the same drug made generically while the original drug was still patented.   This breaks IP law.

Avastin is a different drug and is not a boot-legged generic.

I am no lawyer but I do not see how the Polish case is relevant – smokes and mirrors from usual suspects?
Italy – govt going to ECHR to challenge Novartis. 

Will watch that one with interest.
4) Evidence of efficacy

For wAMD, five trials are currently comparing the efficacy of Lucentis® and Avastin® to answer the question of whether Avastin® is not inferior to Lucentis® with respect to visual acuity. Two of these have now reported , concluding there is limited if any difference in the efficacy or the safety profiles of Lucentis or Avastin. There is now RCT evidence demonstrating equivalent efficacy in DMO, particularly when visual acuity at commencement was mild. There is also substantial evidence highlighting efficacy of Avastin in RVO. 


There was a Cochrane Review published in late 2014 that considered the available RCT and observational evidence; it concluded that there were no differences in the safety profiles of the available anti VEGF medicines. 

Manufacturing standards important – produced in a specials lab with appropriate accreditation – Liverpool and Moorfields.
Proving a patient has a systemic AE that is causally attributable to a specific drug is likely impossible.
5) NICE 

NICE want to appraise avastin, either as 1) proprietary agent or 2) as comparator.

Prevented by doing so in 1) as nobody will refer it, and in 2) as judgement made not in routine use in NHS (cop out and technicality I will explain when we meet)

AMD – avastin is included in the scope. 

Uncertain what the eventual recommendation will be – but if the committee avoids a positive recommendation it will almost certainly be on grounds of institutional risk management.

Uncertain question of whether we want to go against NICE TA recommendation and disallow first line use. Legal issues – against SoS funding directive on CCG, not necessarily same issue for prescriber, but can be considered against NHS constitution. 
Lack of central guidance

Many, including NHSCC and RC Opth, have called for “central guidance” and for government to remove the hurdles to Avastin use; it is doubtful whether this guidance will ever be produced. 

Thus calling for it sounds like long grass to me. RCOpth have within their power and expertise to produce said central guidance. It is considered unlikely that NICE will produce further advice on this matter, in AMD or other indications. A Clinical Guideline on AMD is currently being developed. Watching how avastin is handled in this guideline will be awesome. 
6) Patient preferences & views of patient groups

Should be tested before any definitive position is taken.

Empirical evidence (Foss, Nottingham) suggests that patients care about outcomes not the means or specific drugs used.

Rotherham conducted public survey – 2012/2013?, broadly concluded same.

Patient group support is important. It is unknown whether there are local appropriate patient groups. 

RNIB been vociferously opposed to any move to use Avastin. Their corporate conflict of interest is well documented. The Macular Society have historically been more supportive. Some discussion with them would be warranted. 

7) Funding and money – how much cash will it save

The funds freed up by such a proposition may depend on the precise nature of the proposition. 

Here I will assume that we are talking about new starts – ie the incident population – in macular clinic across the three main indications (AMD, DMO, RVO). 

Incident population

Estimate that 5,5 and 3 patients start in macular clinic with AMD, DMO and RVO per week in Sheffield, thus 15, 15, and 9 across SYB. For each I have estimated a mean of 7 injections. My estimate is that the cost of treating 100% of the new incident population with lucentis / eylea across SYB is £6.3m per year, or £709k with avastin (difference = £5.7m), and over five years the difference is £28.4m.

This needs robust testing, and it will be complicated by use of steroid treatments in DMO / RVO on account of clinical situation and to ease pressure in macular clinics.

See embedded spreadsheet. 

Prevalent population (ie the current spend on aVEGF)

For now I have not considered the prevalent population and or the impact of switching from lucentis / eylea to avastin in those with either stable vision or those with unstable / non or poor responders. See note above re switching however. 

It’s likely I’m wrong
Other points on this issue 

a) It’s worth counting the opportunity cost over the years and making this public

b) Loss of R&D income may be a concern to trust. 

c) May be an issue around NHSLA costs?

d) QC costs for court will be needed.
8) Views of local clinicians.

The views of local clinicians on this matter are not well known. AVASTIN IS STANDARD OF CARE IN PRIVATE PRACTICE (AND ACROSS THE WORLD)

9) Switching and sequential use

Sequential use and and STOPPING There is little evidence on which to base criteria for stopping treatment and so there is considerable uncertainty about precisely how much treatment may be needed in the longer term. There is recent data indicating that there are considerably fewer patients stopping treatment than had originally been anticipated when NICE first appraised Lucentis and Eylea. 

There is a legitimate argument to be made that if a patient is not responding to aVEGF x, there is little to be gained from switch to aVEGF y or z as they are essentially the same molecule. In addition there is little good evidence that supports switching in non responders to first treatment. Thus there might be a legitimate case for a policy of NOT supporting aVEGF switching. There is an equally powerful argument (esp from opportunity cost basis) for stopping.

There are further complex issues with regard to switching and sequential use, NICE is silent on these matters in technology appraisals (TAs). The London Medicines Evaluation Network Review recently concluded there was limited evidence to guide practice here. 
10) what are the potential policy propositions

There are a number of options available to any CCG. Some considerations are set out below each. These are not mutually exclusive. 

Suggested as follows:

a) To set out a commissioning policy that Avastin will be funded in ophthalmic indications in the same clinical cohorts as NICE have currently approved Eylea or Lucentis. 

  • Novartis have previously indicated they may take legal action. This would follow their previous action against SHIP PCT cluster. 
  • This step across many CCGs would be more difficult to fight.

b) To set out a commissioning policy of only funding Lucentis or Eylea as a second line treatment following loss of response to Avastin.

  • This would require legal advice. Secretary of State funding directive for TA recommended treatments states that CCG should make funding available should a clinician choose to use that treatment. 
  • The Mandate states that patients are entitled to NICE approved treatments. 
  • This policy would essentially the same as saying that Eylea and Lucentis will not be routinely funded first line, this might be construed as in direct contravention of the TA funding directive.  

c) To commission an alternative provider for AMD intra vitreal injections who uses Avastin as a first line treatment. 

  • This was the route taken by Stockport PCT and a number of Greater Manchester PCTs went through this route.
  • This would involve entering into a contract with a provider other than the current provider and offering patients choice of providers. There may be a range of procurement issues to contend with. 

Whichever way you go -see you in court.

The CCG should also consider the extent to which it is prepared to support a 50:50 gain sharing agreement with any provider. And how this would be structured.


Given all the above, the opportunity cost and the broader financial position of the NHS, I can’t see it as ethically or morally defensible to continue with eyela or lucentis as first line in the indications they are currently used.
There is a need to discuss and agree the potential policy proposition.
It is recommended legal advice is sought. 

See you in court.


Aug 2016
Conflict statement

I have participated in Ad Boards for Bayer opthal, my (now ex) employer was reimbursed for my time. I personally received no pharma shilling.

I am also chair of the TANDEM steering committee, and have been a member (not current) of a NICE TA committee that has considered aVEGFs on many occasions. 


One thought on “Avastin in ophthalmology Briefing for CCGs

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s