Laser Treatment for diabetic retinopathy for professionals

David Kinshuck



AnitVEGF injections

These are giving better results than laser for both diabetic maculopathy and proliferative retinopathy!


Laser ..types...Pascal, subthreshold

The commonest laser is Argon Green, wavelength 530nm, but other wavelengths can be used. Other types of light were used before laser was introduced. There are two new types of laser that seem very helpful.

First, the PASCAL laser, which can apply several hundred burns quickly, and seems ideal BJO 2010. It is also much less painful BJO 2010. Pascal macular grids. Pascal, fewer sessions for PRP BJO 2011 Nerve fibre Retina 3011.
Secondly, a subthreshold diode laser with a longer wavelength may be best for both macular and PRP laser. There are an increasing number of reports that subthreshold laser is safer for macular oedema.

The benefit is from the lower power used, not the different wavelength difference. Burns of shorter duration certainly produce less retinal damage are are probably more effective (2010), but need more burns

Less damage subthreshold Retina 2012. Subthreshold laser, even PRP, is recommended for milder retinopathy, but if the retinopathy is severe heavier laser burns will be needed.

Subthreshold is safe Retina14   Eye15  Eye10. Subthreshold PRP laser is unlikely to cause macular oedema or severe epiretinal membrane formation. Heavier burns (greyish white), which are needed for more aggressive proliferation, cause macular oedema in 20% of eyes.


Inadequate diabetic control (HbA1c & BP)

It is absolutely vital that diabetic control is addressed. If a patient is under the care of an very experienced diabetes specialist nurse/doctor, there may be no improvements that can be made. But almost by definition, patients must have had control that is less than perfect, as it is rare with good control to develop retinopathy.

These are the standard targets (HbA1c, blood pressure (BP), exercise, healthy diet, exercise, good mental health etc).


Why is HbA1c/BP control poor? Certain clinical situations are common.

For good results, as far as possible these issues have to be addressed.

Ideas that may help to improve control

  1. ASAP liaison with the diabetic team..whoever is providing the care
  2. For all patients using multiple insulin injections and who test their glucose regularly, the new glucose sensor is highly recommended  here.
  3. if HbA1c is high, a gradual drop may be best (current advice, limited evidence, this may change). However, it is more practical just to lower it the easiest possible way, accepting it may drop quicker than ideal for the eyes.
  4. 3-12 months, diabetes education courses
    • such as DAFNE for type 1 patients,
    • XPERT for type 2 patients,
    •  DESMOND for newly diagnosed type 2.
  5. 3-6 months after starting laser, a modern insulin regime, perhap sa pump with a sensor (but dont get the sugar too low until the retinopathy controlled: new-onset tight control with exacerbate the retinopathy).
  6. if the BP is higher than target, increase medication every 2-4 weeks until controlled
  7. psychologist/psychiatrist if problems severe
  8. obesity...expert help best; low self-esteem plays a significant role
  9. show patients their retinal photos (or another very similar)
  10. whilst people may not be receptive certain times, at other times, perhaps a year or two later, they may be
  11. address smoking
  12. nine steps for success
  13. patients need baby steps: perhaps 1-2 weekly visit to their diabetes doctor/nurse, making small changes each visit. This may need to be carried out over 6-12  months.
  14. Use stories to illustrate ways forward
  15. when a problem is found (e.g. poor control, smoking) compliance is likely to be ~3 x higher if an appointment is made there and then, and give the appointment to the patient. Often this is not possible, but merely mentioning the problem is much less effective. Thus we should be making appointments with the diabetes specialist nurse or the Stop Smoking clinic during our retinopathy consultation, is we want good compliance.

Planning treatment...taking account of diabetic control

We need to consider the rate of progression of retinopathy. With perfect control, there should be no/very little retinopathy progression. We can use UKPDS and DCCT data (and other papers) to calculate progression rate as opposite. But progression rate may change:

  1. if the HbA1c drops 3%, the retinopathy will progress much more quickly. This applies to the 3-4 years after starting good diabetic control, but after this time progression rate will reduce considerably.
  2. Pregnancy tends to lead to more rapid progression.
  3. sudden increases in blood pressure (usually due to decreased renal function), may also lead to more rapid progression.
  4. on the other hand, a type 1 patient diabetic for 40 years with reasonable control, should have more slowly progressive retinopathy
  5. anyone with good BP/HbA1c for 3-4 years: progression should be very slow and most patients with 4y of good glucose control, low BP, and not smoking will not have active retinopathy. It takes 3-4 years of such good control to stabilise the condition...such patients can be discharged from the laser clinic and just monitored by screening programs
  6. a certain degree of retinopathy may have developed in a patient
    • a long duration of diabetes (e.g. 40y) (will progress slowly)
    • with newly diagnosed diabetes (and recently high HbA1c) (will progress rapidly as HbA1c drops)
    • poor control for a few years
    • poor control for years, even is this was may years ago: 'legacy' effect
    • good control...but was poor in the last 3 years
  7. For the same degree of retinopathy, progression rates may be very different, and laser should be increased in eyes expected to deteriorate rapidly, as in the table below.
  8. Shorter pulses produce less retinal damage, Muqit 2010.

Progression rates of microvascular disease

Eyes, kidneys, neuropathy etc

Progression rate varies despite patients having similar retinopathy appearance at any one time


duration of diabetes progression rate
40y reasonable control slow
newly diagnosed diabetes now well controlled HbA1c was 10, now 7 rapid
poor control for a few years (HbA1c  9%) intermediate
good control...but was poor in the last 3 years rapid
good control...but was poor in the last 3 years..started insulin recently rapid
blood pressure rise recently rapid
Good control in last 3-4 years HbA1c  7%, BP 130/75, non-smoker normally no progression whatsoever
pregnant (good control before conception) may be rapid
pregnant: poor control before conception and control better now extremely rapid


progression rate of diabetic retinopathy

Enlarge Calculating the amount of laser: increase if the retinopathy is progressing rapidly (Good HbA1c control is essential in the long term, but may increase the progression rate for 1-3 years. Retinopathy will generally stop progressing completely in non-smokers after 3-4 years of good control.)


Planning PRP laser (peripheral laser photocoagulation)

PRP laser treatment was the main treatment for proliferative retinopathy (new vessels) or moderately severe pre-proliferative retinopathy, although increasingly anti-VEGF injections are taking over.

It helps to think ahead....how much treatment is needed in the short (and long) term. Rather than carry out 1200 burns or anti-VEGF and see the result, best results are likely if the treatment is carried out early. Some patients need a lot of laser or anti-VEGF, some a little. This summarised in this paper, Eye 2011

"Following top-up PRP treatment, regression rates varied according to severity: 75% for mild PDR (n=6), 67% for moderate PDR (n=14), and 43% in severe PDR (n=3).

What is maximum laser?

In descriptions on this page, 5000 large or 18000 smaller lighter burns are given as the estimate for treatment of most of the peripheral retina (with laser carried out in multiple sessions as on this page). But this is a very approximate figure..some patients may have 5000 burns and a lot of space is left (and they may need more laser), some may have no space after fewer. Please take this into account when interpreting the examples below.

These figures are for the trans equator lens, the easiest to use for PRP laser. Each lens has different magnification (Scanlon, p121). The Mainster 165 is reported to be the ideal lens now.

To achieve complete disease regression, mild PDR required a mean of 2187 PRP burns and 264 mm ablation area, moderate PDR required 3998 PRP burns and area 456 mm), and severe PDR needed 6924 PRP laser burns (836 mm;). Conclusions Multiple 20-ms PRP treatments applied over time does not adversely affect visual outcomes, with favourable PDR regression rates and minimal laser burn expansion over 18 months. The average laser dosimetry and retinal ablation areas to achieve complete regression increased significantly with worsening PDR."   (DK now uses a lot more much smaller lighter burns. )

In an excellent review, essential reading for professionals, Eye 2011, the amount of laser can be calculated. The study was based on patients with an HbA1c current and recent <10%. Higher HbA1c patients may be expected to need a lot more laser. The paper describes laser in terms of retinal area that needs treatment, but this difficult to judge clinically. However, for significant nvd/e, most of the peripheral retina needs laser.

For the <10% HbA1c patients, in the Manchester study,

But remember, patients in higher risk groups e.g. recent HbA1c> 10, smokers will need relatively more. patients with very good control as discussed below need less.

Once there are retinal complications, such as haemorrhage and traction , vitrectomy is more likely to be needed Retina14.


Consider certain clinical situations

Patient 1

A patient with very active new vessels...generally needs ~15000 smaller lighter burns/eye, soon…1st 3 sessions in 1st week. Laser is needed quickly as it is very likely there will be a vitreous haemorrhage soon. (3000/eye/session, both eyes same session). Probably needs 18000 burns/eye (much of peripheral retinal lasered) over the next 2-4 weeks.

Examine monthly after this. Even this may not be enough and new vessels may continue to grow. If so, we believe anti-VEGF may be ideal for such patients. At present we laser 3-5 sessions and examine the result, but it is probably best to try and calculate how much is needed at the onset, and carry this out.

Patient 2

Vitreous haemorrhage with new vessels..similar treatment to patient 1. Vitrectomy if you cannot laser through the haemorrhage (vitrectomy with endo laser).

Patient 3

Disc new vessels, poor control. Retinopathy will progress reasonably quickly, laser needed over ~4weeks..similar treatment to patient 1.

Patient 4

Disc new vessels, poor control. But will start to improve diabetic control. Retinopathy will progress even quicker with the improving control, laser needed over ~4weeks, similar treatment to patient 1. Good control helps in the long term, sometimes not in the short term.

Patient 5

Disc new vessels, good control for the last 4 years. Retinopathy will progress slower, may manage with 10000 burns/eye, over ~8weeks.  Examine in 2-4 months to see benefit, may need more laser over next 2 years.

Patient 6

Severe pre-proliferative, poor control. But will start to improve diabetic control. Retinopathy will progress quickly, may manage with 12000 burns/eye, over ~8weeks.  Examine in 2-4 months to see benefit, may need more laser over next 2 years. Difficult to predict outcome. Many type 2 patients starting insulin are in this position. Best to laser early in my opinion.

Patient 7

Severe pre-proliferative, poor control. Won't improve HbA1c, blood pressure 130 systolic. Difficult to predict, controversial. I prefer laser: 9000 burns/eye, over ~8weeks.  Examine in 2-4 months to see benefit, may need more laser over next 2 years. If no laser, examine every 3-4 months.

Patient 8

Reasonable control, already had new vessels treated with 12000 burns, but new vessels remain. Need ~6000 more.

Patient 9

Reasonable control, already had 12000 burns, not much room for more laser, but new vessels persist...? Avastin. Check the peripheral retina carefully...certainly more laser if there is lots of space.

Patient 10

Poor control (or control poor in the last 1-3 years), active new vessels, vitreous haemorrhage, only had 9000 burns. Needs more laser. ?Avastin then laser. ?Vitrectomy   ?Avastin then vitrectomy the next week. Laser other eye.

Patient 11

New vessels 2002, laser PRP, 3000 larger burns both eyes (equivalent to 12000 smaller lighter burns), vessels regressed. Mediocre control at this time.   2008 presents with vitreous haemorrhage. Hard to see new vessels, but there are a number of blot haemorrhages. Also, lots of unlasered peripheral retina.

Blot haemorrhages suggest continuing ischaemia and probable new vessels..therefore laser remaining peripheral retinal until most lasered.

Patient 12

Alternatively..same patient as patient 11 but no blot/retinal haemorrhages...could be a simple PVD pulling some 'older' new vessels...this is more likely if there are no retinal haemorrhages at all. Probably safe to watch in this case.

Patient 13

Age 80y, had macula laser 2 years ago. Poor control, very overweight. New vessels now....continue with peripheral laser..until most periphery lasered.

Patient 14

New vessels elsewhere just in one area of retina, not many haemorrhages. May manage with laser to the appropriate sector e.g. superior temporal quadrant. But all sectors with ischaemic blot haemorrhages should be lasered.

Patient 15

Same patient as patient 14 with good control/non-smoker for the last 20 years..generally not much laser needed.

Patient 16

A patient with type 2 diabetes using insulin HbA1c 9% / 75 mmols/mol with sugar levels about 9 at work. Very severe pre-proliferative retinopathy requiring laser, about to develop macular oedema.

He lifts heavy fridges and does not want to get a hypo and drop the fridge. A colleague got a severe injury in such a situation. A diabetic consultant recommended a sensor that would alarm if the glucose dropped.

Using this he could aim for sugars of 6-7 at work, achieving much better control. But patients with type 2 diabetes have to pay for thier own sensors in the UK.

Patient 17

Age 28, type 1 diabetes, years of mediocre control, presented with proifertive retinopathy, lots of laser. Stable.  6 months later started a closed loop insulin pupmp sensor system targeting a HbA1c of 6.7 mmol/mol.

Then developed severe macular ischaemia in one eye, so the diabetic team were asked to aim for a higher sugar for 6 months, and anti-VEGF commenced then drop gradually to 6.7. Outome awaited (April 2020).


Laser PRP..how much, some ideas

Laser settings for PRP...proliferative retinopathy

PRP laser..preventing macular oedema

If there is macular oedema before starting laser, this is likely to increase with PRP laser, which is one of the main reason anti-VEGF injections are taking over as the main treatment. This paper is important, indicating that oedema of para-foveal areas makes post-PRP laser loss of vision much more likely, and sometimes this is permanent. This may mean we should carry out a macular grid first, and wait for the oedema to reduce...but very often there is no time to do this (active proliferation requires laser soon). This paper, to my mind, implies that it is important to grid laser anyone with para-foveal macular oedema, in the pre-proliferative stage, so that the macular oedema has time to settle by the time full PRP laser is essential.

Macular oedema is very common after PRP, even after macular grid laser. Sometimes this is because the PRP was too heavy, as Bandello above. (Low blood pressure will help a little...even <130mmHg systolic. Lighter burns as above have further reduced this problem substantially clinically (no hard evidence for this..and as above OCT does demonstrate oedema/increased retinal thickness in many patients).

As above, in all except young type 1 patients with healthy maculae many ophthalmologists prefer to carry out a light grid laser before the PRP, to prevent macular oedema. Carry out the grid and PRP laser at the same session in very severe cases or if you think the patient is not likely to come back again.

Eyes with considerable ischaemia are far more prone to macular oedema. If macular oedema is present anti-VEGF injections are really essential.


Laser PRP and visual field loss

Laser burns enlarge. If there is active proliferation, there is absolutely no choice but to laser. But once much of the retina is lasered, the patient will have lost a fair amount of visual field.

At this stage, further laser will cause much more significant field loss: if 60% of the retina is lasered, lasering 10% of the the remaining unlasered retina will cause 25% more field loss.

For this reason, a number of our patients have had to stop driving, and have great difficulty with their side vision. To prevent this, we are starting to use Avastin. We hope 3 injections (after much of the retina has been lasered) will stop proliferation indefinitely. We are waiting to confirm this! All our Avastin patients get new vessel regression for a while, but unless they are well lasered the proliferation occurs later...the Avastin only delays progression.

Everyone is hoping subthreshold laser PRP will enable lighter burns and hence maintain a better visual field. The PASCAL laser produces quicker burns...may be this will be helpful. PASCAL laser seems to be a big step forward, but the laser costs 3 times as much as a regular laser, and inadvertent macular laser is a possibility with the current model.


Lasering through a cataract or vitreous haemorrhage

When lasering through a vitreous haemorrhage or cataract


Indirect laser

Retinopathy is symmetrical

When carrying out PRP laser it is very important to treat the fellow eye, this important paper suggests. Unless the retinopathy is highly asymmetrical, which is most unusual, laser is important. Furthermore, the same paper suggests that 4-6000 3-500μ burns are needed to prevent retinopathy developing to maintain vision in the long term. This is because the worst visual results were in the second eye to be treated.

If the retinopathy is very asymmetric suspect carotid artery stenosis.


Planning treatment for diabetic maculopathy

Remember to address diabetic control.   Anti-VEGF injections are the main treatment, but if not available laser as here (as previously regularly used):


Laser for diabetic maculopathy: settings

Remember to address diabetic control.   Anti-VEGF injections are the main treatment, but if not available laser as here (as previously regularly used):

A treatment plan for diabetic maculopathy modified after Simon Harding, 2007 and Olk, (Now replaced  with anti-VEGF injections):

give information sheet and record this in notes


or diffuse maculopathy?

circinate, small area of oedema

diffuse/substantial oedema

focal laser

Anti-VEGF main treatment, laser may be needed additionally: grid laser

generally include laser shots to microaneurysms if >1 disc diameter from fovea & scatter laser to non-foveal areas of non-perfusion/IRMAs (usually temporal retina)

reassess ~6 months



Reassess 2 months. If poor laser response as seen clinically, repeat (i.e. if laser burns not visible and maculopathy no better;) do not repeat if plenty of burns visible. Anti-VEGF main treatment always needed at this stage.

Reassess 4 months later, consider repeating laser if oedema/thickening remains or not reduced with OCT, otherwise FFA.

oedema persists, FFA

treat leaky areas as identified with FFA; consider anti-VEGF drugs; very ischaemic macula with very large avascular foveal zone...very  difficult to achieve improvement..consider PRP

Monitor all patients for deterioration/proliferation.


Maculopathy grid laser technique

new vessles grow in proliferative retinopathy

Create boundaries for the laser....must keep away from the fovea.
Some exceptions..may need a little laser to microaneurysms slightly within this are..but only a few shots of laser , and not near the fovea.


Extend the boundary a little horizontally


Laser always moving from the fovea (again, this is to prevent foveal laser). Subthreshold laser is invisible so you need to remember which area has been treated.


Standard grid laser, with extra shots further out, more laser in ischaemic area ... areas of microaneurysms and haemorrhages



PRP Laser is traumatic and painful

Having laser is a very traumatic experience, especially PRP laser. Trento and others investigated this (EASDec 2003) and found


Read this book

Retinopathy and poorer quality of life

Anti-VEGF treatment

Can be very helpful