Dear Editor,

When reading the article we were a bit surprised by the extremely positive declaration of the clinical results after subthreshold diode micropulse photocoagulation. The authors stated that subthreshold diode micropulse laser photocoagulation minimises chorioretinal damage in the management of CSMO and demonstrates a beneficial effect on visual acuity and CSMO resolution (in 96% of all treated eyes, n=95). However, we have the strong feeling that the presented results may not support this conclusion. Moreover, the basic mechanism of the proposed laser interaction is unclear. It is most likely that nothing than marginal thermal side effects occurred in the retina during treatment.

Conventional laser therapy is regarded as effective in treating CSMO. Unwanted side effects as e.g. induction of CNV or subretinal fibrosis principally do not appear if laser parameters for threshold exposition are carefully used. Generally a new laser method must measure against this gold standard, so that at least comparable results are obtained. This seems not to be the case in the present article.

Conclusions such as "visual acuity was stable or improved in 85% of treated eyes" are questionable if Tab.2 shows stabilization in only 76.8% of eyes (+ < 3 ETDRS lines) and only 8.4% achieved significant better visual acuity (also 14.7% lost more than 3 lines). Regarding Tab.1 and Tab.3 overall visual acuity became worse and not better and nearly none of the p-values showed significance (if p-value was significant visual acuity was worse).

As stated by the authors, the validity of this pilot study is limited by its small size and retrospective nature. In fact no uniform postoperative patient follow-up was performed and only the "last available visual acuity measure" was taken for outcome assessment. This value was obtained between 3 and 29 (mean 12.2 months), which underlines a high variability. In other words, the results might only reflect the spontaneous untreated course of CSMO. Also questionable is the postoperative gradation of CSMO as "worse, better and resolved". There were no OCT scans taken either pre- or post-operatively to verify macular thickness. Also, angiography seems to be performed only in patients who appeared to need additional treatment. Thus from none of the presented results could it be objectively concluded that CSMO improved in 96% of eyes.

Finally it is stated in the results section that in "79% of eyes exhibiting complete resolution of CSMO postoperatively had significantly better visual outcomes compared to 17% of eyes with persistent and 4% of eyes with worsening macular oedema (table 8)". However, Tab.8 clearly demonstrates that only 8% (n=6) with resolved oedema (n=75) gained 3 or more lines, which is significantly better, but that 84% (n=63) had only stable visual acuity within + 3 lines visual acuity change, whereas also additional 8% had significant vision loss (Tab.8). Thus the proposed positive clinical result in terms of CSMO resolution and visual acuity improvement could not been followed.

Independent of clinical results, the mechanism of the micropulse laser method is unclear. Since there were no ophthalmoscopically nor angiographically visible laser damage in the tissue, one has to ask, what happens to the fundus during treatment? As proven in many experimental studies conventional laser photocoagulation leads to primary destruction of the RPE since it absorbs about 60% of the energy from a green laser beam. The RPE damage is repaired within 7 days by migration and proliferation of neighbour cells and this seems to lead to an enhanced pump-function of the new RPE cells leading to resolution of CSMO. Bruch´s membrane usually stays intact, thus no potential CNV induction is expected. Because of the long laser exposition times of about 100ms during irradiation, thermal damage to photoreceptors leads to irreversible laser scotoma.

In both laser treatments (thermal laser and SRT) the primary RPE damage can be demonstrated by angiography revealing leakage from the damaged RPE site, thus the mode of action of the laser treatment can clearly be proven and is comprehensible. This is not the case in the article by Luttrull et al. who used repetitive laser pulses of 100µs pulse duration (which are - at required energies for RPE damage - too long to spare photoreceptors) within an envelope of 300ms. Temperature calculations for the laser parameters set in this article reveal an increase of tissue temperature of 1.8°C per pulse within the laser spot (taken into account that there is just a 20% energy absorption within the RPE/choroid complex at 810nm wavelength). The mean temperature increase in the centre of the laser spot is - due to heat accumulation at the high repetition rate of 500Hz - about only 11°C after 300ms. Neither thermal nor thermomechanical based tissue alterations are expected for this low short time temperature increase. Consequently it might be not remarkable that - as also stated in the discussion section - the angiographical visible diabetic leakage after therapy mostly persisted. The discussion of possible mechanisms of this kind of micropulse laser irradiation as e.g. up- and down- regulation of different growth factors or heat-shock proteins is speculative.

Carsten Framme, MD; Veit-Peter Gabel, MD