Is orthokeratology useful for control of low myopia?

There's a common clinical belief that Orthokeratology (OK) doesn't work as well in lower myopes for myopia control. I've even seen this included in conference presentations as prescribing advice. This might be presumed in comparison to higher myopia, or in advocating for multifocal soft contact lenses as a better treatment choice. Is orthokeratology useful for control of low myopia? Here's what's fact and what's fiction.

The comments below from a clinical case discussion in our Myopia Profile Facebook group concerned a child with an OU -1.00 refraction, and capture the clinical assumption nicely.

Heng et al² undertook a retrospective evaluation of one year of OK wear, studying 141 OK wearers and comparing them to 130 matched controls. They reported that the axial elongation in low (up to 3D), moderate (3-5D) and high (5-6D) myope groups in their study were 26.3%, 34.2% and 57.1% slower than the control group respectively. On the surface this would indicate that OK works better for higher myopes, but it was not shown if the axial elongation in the low and moderate myope OK groups were significantly different from each other, and the amount of high myopic OK wearers for comparison was very small. They could have also been different ages - having more older children in the higher myopia group, for example, could have influenced the appearance of less progression in the higher myopes.

Overall there was no relationship found between baseline myopia and the amount of axial elongation which occurred. Instead, the main relationship found was that between progression and age. Younger children (less than the mean age of 9.4 years) had a greater percentage of faster progression - more than 0.36mm or 1D in the year - and hence were concluded to benefit more from OrthoK reducing axial elongation in absolute terms.

In looking at further OrthoK studies, if a correlation was reported between baseline myopia and axial elongation, this could indicate an effect of better myopia control (less axial elongation) with higher myopia. I looked over several two-year OrthoK studies to see if this correlation was examined, and where it was, here's what I found.

• Kakita et al, 2011: OK worked to slow myopia as shown in numerous studies. A significant correlation was found in the OK group (but not the control group) where lower baseline myopia was correlated with faster axial elongation. This could mean the OK-wearing lower myopes were progressing faster. However the effect was small - R-squared value of the correlation was 0.08, meaning only 8% of the variation in axial elongation was explained by the factor of baseline myopia. Perhaps the lower myopes were younger, but the relationship between refraction and age wasn't evaluated. The relationship between age and progression also wasn't evaluated.
• Cho et al, 2012 (ROMIO study): Axial elongation was not significantly correlated with initial myopia but was correlated with initial age - younger age meant more progression. On the upside, younger age (7-8 years) also meant a bigger difference between OrthoK and control groups in the percentage of fast progressors (more than 0.36mm/yr) - 65% in the control group but only 20% in the OrthoK group. The larger absolute difference in progression speed happened for the younger children wearing OK - they got the most benefit.
• Hiraoka et al, 2012 (five year study): Again, a correlation was found between younger age and more axial elongation, but the correlation between baseline myopia and axial elongation was not statistically significant.
• Chen et al, 2013 (Toric OK or TO-SEE study): Younger children progressed faster; and again no relationship was found between axial elongation and initial myopia.
• Santodomingo et al, 2013: Analysis of predictive factors of slower axial elongation found that later age of myopia onset meant less progression. There was no significant relationship found between level of myopia and myopia progression.
• Zhu et al, 2014: This retrospective study found the myopia control effect of OK was significant after one year and two years in low and moderate myopes, but only after one year in high myopes. As above, the two-year axial elongation was associated with baseline age but not with baseline myopia.

The sum total of this analysis is that OrthoK is likely to be 'most effective' for reducing absolute axial elongation and fast progression in younger children (up to and including 9 years of age) and that younger age is the key relationship to efficacy. There is an absence of any systematic evidence that OrthoK works better for higher than lower myopes from a myopia control perspective - it works well, on average, for any myope.

There is a clinical assertion that a low myopic OrthoK treatment won't provide as much of an optical change to the periphery as what a soft multifocal contact lens (MFCL) could provide, and this could lead to less efficacy.

This assumption is likely built on the comparative optics of each. OrthoK is known to change peripheral refraction in an approximate 1-to-1 relationship to central refractive change.² This means you can think of a -1.00D OrthoK treatment as having a +1.00 'peripheral add'. When compared to the labelled add power on a MFCL, it may seem we could do more with the latter, given that commercially available MFCLs could include +2.50 Adds or more for this -1.00 myopic patient. Surely more add is better?

In fact, the direct comparison of OrthoK treatment power to MFCL labelled 'add' power is not that simple. MFCL's have been shown to induce a smaller and less consistent change in peripheral refraction than the equivalent powered OrthoK lens.³ This means a +1.50 Add MFCL, as suggested in the image above, isn't necessarily better than a -1.00 OrthoK treatment, which will produce the +1.00 effective peripheral add, and in fact the MFCL induced change may be much less. Note that this comparison was utilizing a commercially available, continuous aspheric design MFCL so perhaps newer MFCL designs would be different, but this is the only comparison available in the literature.

There is one study which directly compared OrthoK to a novel MFCL design (called 'soft radial refractive gradient' or SRRG) over two years. Paune et al⁴ found that both contact lenses reduced axial myopia progression compared to the single vision spectacle control group - OrthoK by 38% and SRRG by 27%. That makes OrthoK look better, and indications were that it did win the race - while the axial elongation was not statistically different between the OrthoK and SRRG groups, the SRRG group wasn't significantly different from the SV group meaning it fell somewhere in between - SRRG approached a significant difference from SV but didn't quite make it. There are no other head-to-head comparisons of OrthoK and MFCLs in the literature for efficacy.

What does meta-analysis data tell us? We have a longer duration of meta-analysis data available for OrthoK (two year data - open access here and here) showing -0.26mm less progression over two years compared to -0.10mm less progression in a one year meta-analysis of various MFCL designs. This MFCL meta-analysis, though, was published in 2016 and newer designs which have emerged since may fare closer to OrthoK results.

This blog is a long-winded way to give you a short answer - OrthoK works well for management of low myopia. OrthoK works, MFCLs work too, and it doesn't need to be more complicated than that, so pick the option which suits your patient best.

• There is currently no supportive evidence for OK being less effective in lower than higher myopia. It should work similarly for myopia control regardless of baseline myopia.
• There is no evidence suggesting that MFCLs are comparatively more effective than OK for low myopia. In fact, the one direct comparison may indicate the opposite, and their optics are different. But let's keep things simple - both work. Pick what CL option suits your patient best, as current meta-analyses indicate that while OK has the greater volume of myopia control evidence, the overall efficacy is likely similar as for MFCLs.
• The chief determinant of myopia progression is age - the sooner you start, the better the overall impact of myopia control will be on final axial length.

1. He M, Du Y, Liu Q, Ren C, Liu J, Wang Q, Li L, Yu J. Effects of orthokeratology on the progression of low to moderate myopia in Chinese children. BMC ophthalmology. 2016;16(1):126 (link)
2. Queirós A, González-Méijome JM, Jorge J, Villa-Collar C, Gutiérrez AR. Peripheral refraction in myopic patients after orthokeratology. Optom Vis Sci. 2010;87(5):323‐329. (link)
3. Ticak A, Walline JJ. Peripheral optics with bifocal soft and corneal reshaping contact lenses. Optom Vis Sci. 2013;90(1):3‐8. doi:10.1097/OPX.0b013e3182781868 (link)
4. Pauné J, Morales H, Armengol J, Quevedo L, Faria-Ribeiro M, González-Méijome JM. Myopia Control with a Novel Peripheral Gradient Soft Lens and Orthokeratology: A 2-Year Clinical Trial. Biomed Res Int. 2015;2015:507572.  (link)