Study of clinical applications and safety for Pascal® laser photocoagulation in retinal vascular disorders

Authors


Paulo E. Stanga
Consultant Ophthalmologist and Vitreoretinal Surgeon
Honorary Reader in Ophthalmology
Manchester Royal Eye Hospital
University of Manchester
Oxford Road
Manchester M139WH
UK
Tel: + 441612761234
Fax: + 441612765555
Email: retinaspecialist@btinternet.com

Abstract.

Purpose:  To establish safe laser parameter standards for 10–30 ms Pascal® laser in clinical practice and to evaluate clinical and visual outcomes using this 532-nm multi-spot photocoagulation system.

Methods:  Retrospective observational case series of 313 patients treated between 2006 and 2008. Evaluation of eight groups: A – panretinal photocoagulation (PRP) for proliferative diabetic retinopathy (PDR); B – focal laser treatment for clinically significant diabetic macular oedema; C – grid laser for diffuse diabetic macular oedema; D – sector PRP for ischaemic branch retinal vein occlusions (I-BRVO); E – full PRP for ischaemic central retinal vein occlusions (I-CRVO); F – macular laser treatment for macular oedema secondary to non-ischaemic BRVO; G – full PRP for rubeosis iridis and/or neovascular glaucoma (NVG) secondary to I-BRVO, I – CRVO or PDR; H – laser retinopexy for retinal breaks/degenerations.

Results:  Mean LogMAR visual acuity for all procedures improved postlaser (p = 0.065), and laser prevented visual loss in 85% eyes. Topical anaesthesia was only required. At mean follow-up of 5 months, 72% procedures had a successful clinical outcome. Significantly higher powers were required for PRP using Pascal® compared to conventional laser (p = 0.001) in PDR, I-BRVO, I-CRVO and NVG. Sixty-seven per cent of patients (15/20) were successfully treated with single-session 20-ms PRP using a mean 1952 burns. There were no laser-associated adverse effects or ocular complications associated with multi-spot PRP or macular Pascal® arrays.

Conclusions:  The clinical efficacy using 10- to 30-ms pulse duration Pascal® laser is comparable to conventional standard protocols used for the treatment of vascular retinal disorders. Higher power, 10- to 30-ms pulse duration laser may be safely and effectively used in clinical practice.

Introduction

Retinal vascular disorders, notably diabetic retinopathy, constitute a leading cause of legal blindness. Laser photocoagulation is the gold standard first-line treatment, and current work is ongoing to improve efficacy and visual outcome combined with precise and safe laser techniques. Since the 1950s, a number of laser systems have been used by ophthalmologists including xenon arc, ruby and krypton (Beetham et al. 1969; Irvine & Norton 1972; Schulenberg et al. 1979). Zweng et al. (1970) from Stanford University co-developed the first commercially available argon laser slit-lamp photocoagulator.. Early treatments of clinically significant macular oedema (CSMO) involved long pulse duration (100–200 ms) focal or grid laser, and patients with proliferative diabetic retinopathy (PDR) received scatter panretinal photocoagulation (PRP) using 100- to 1000-ms pulses.

In 1976, a landmark trial in laser photocoagulation treatment of PDR was undertaken by The Diabetic Retinopathy Study (DRS) Research Group (1976) in association with the National Eye Institute. They reported a 60% reduction in the risk of severe visual loss over a 2 year follow-up period in both treatment arms for PDR. The Early Treatment in Diabetic Retinopathy Study (ETDRS) (1991) demonstrated the effectiveness of photocoagulation for macular oedema and the use of early PRP at the severe non-proliferative stage in type II diabetes. Argon laser photocoagulation parameters were applied to subsequent retinal vein occlusion trials, and scatter PRP and macular laser were both shown to have positive effects in the prevention of visual loss using conventional laser (Branch Vein Occlusion Study Group 1986, BVOSG; Rehak & Rehak 2008).

The Pattern Scanning Laser (Pascal®, 532 nm) photocoagulator was introduced in 2005 (Blumenkranz et al. 2006). The burns are applied in a rapid predetermined sequence in the form of a multi-spot pattern array produced by a scanner under programmed hardware control. The pulse duration is reduced to 10–20 ms for macular photocoagulation and 20–30 ms for PRP. Initial pilot work demonstrated potential benefits for patients while remaining consistent with standard protocols (Sanghvi et al. 2008). As yet, there are no large clinical trials to report on clinical outcomes and complications using Pascal® laser.

Retinal photocoagulation using 10- to 30-ms pulse duration may be less destructive and initiate healing responses within the retina (Jain et al. 2008). The localization of medium pulse duration burns has recently been reported in humans (Muqit et al. 2008). It may be worthwhile for clinicians to consider the application of localized burns with better predictability for ischaemic retinopathy and macular oedema, as laser retreatments may be required with risks of laser burn expansion in the long term. However, currently, there are limited published data for 10- to 30-ms laser regarding safe therapeutic windows for photocoagulation in vascular retinal disorders.

This study aims to investigate the effectiveness of the Pascal® 532-nm laser system in sight-threatening diabetic retinopathy, retinal vein occlusions, macular oedema and retinal tears. We aim to establish safe benchmark Pascal® laser parameters for clinical practice and to evaluate clinical and visual outcomes using medium pulse laser photocoagulation.

Methods

Design

This is a retrospective observational case series of consecutive patients who underwent Pascal® laser photocoagulation at Manchester Royal Eye Hospital between 29 November 2006 and 31 January 2008. Information collected included age, sex, indication for treatment, grade of surgeon, procedural anaesthesia, best-corrected visual acuity (BCVA), clinical efficacy and outcome, and complications following laser. Data were also collected on the parameters used for the treatment including power, pulse duration, number of burns per treatment session and retinal spot size. Laser parameters were analysed for both the Pascal® system and also for the same clinical disease entity that may have previously undergone conventional single-spot laser treatment (HGM, Salt Lake City, UT and Litechnica, Manchester, UK). The presence of ocular co-morbidities such as age-related macular degeneration, glaucoma and cataract was recorded.

Treatment groups

We identified eight groups that underwent laser photocoagulation using the Pascal® system. Fluorescein angiography was performed prelaser in all patients to classify the disease stage, except for the retinal tears/degenerations group. Optical coherence tomography was performed at the discretion of the surgeon; however, this test was not used routinely for patients with maculopathy. Patients undergoing PRP (Fig. 1) for proliferative diabetic retinopathy (PDR) were defined by the presence of retinal neovascularization according to ETDRS guidelines and comprised group A. Patients with ETDRS-defined CSMO underwent focal laser (group B), and diffuse diabetic macular oedema (DMO) was treated with macular grid laser (group C, Fig. 2). As per BVOSG criteria, patients in group D underwent sector PRP for ischaemic branch retinal vein occlusions (I-BRVO). Patients in group E received full PRP for ischaemic central retinal vein occlusions (I-CRVO). Patients in group F with macular oedema secondary to non-ischaemic BRVO underwent BVOSG-defined macular grid laser. Group G consisted of patients with rubeosis iridis and neovascular glaucoma (NVG) secondary to I-BRVO, I-CRVO or PDR and underwent full PRP. Patients with retinal breaks/degenerations comprised the final group H and underwent retinopexy (Fig. 3).

Figure 1.

 Patient with proliferative diabetic retinopathy and area of retinal neovascularization. Colour fundus photograph of 20-ms laser panretinal photocoagulation burns at 1 hr after treatment. Pascal® 5*5 arrays are visible, using 1.5 burn-width spot-spacing and 400-micron laser spot size. Laser burns are seen to overlap the venules and arterioles, with no signs of any direct damage to retinal blood vessels from treatment arrays.

Figure 2.

 Clinically significant diabetic macular oedema treated with full grid macular laser. Colour fundus photograph of 10-ms laser burns at 1-hr posttreatment. The A plus B pattern consists of four concentric rings with 112 spots encircling the fovea. The diameter of the innermost ring is 2000 μm (‘safety zone’), and this inner ring is made up of 12 spots and is delivered in 132 ms that is theoretically below a patient’s reaction time. Threshold grey–white laser burns are seen to overlap the macular venules and arterioles, with no signs of any direct damage to blood vessels from laser burns. There is satisfactory laser burn uptake within the treatment arrays.

Figure 3.

 Clinically significant diabetic macular oedema treated with focal macular laser. Colour fundus photograph of 10-ms laser burns at 1-hr posttreatment. Pascal® pattern A+B arrays are visible with single and double octants of laser spots within treatment areas. Laser burns are seen to overlap the macular venules and arterioles, with no signs of any direct damage to blood vessels from laser burns.

Laser technique

Threshold grey–white intensity burns (ETDRS grade 2–3) were used for all laser treatments in groups A-H. Mainster wide-field and PRP lenses were used for PRP and sector laser with 200- to 400-μm spot size, and Mainster focal contact lens for macular treatments using 100-μm spot size. Laser treatments were performed by retinal specialists and experienced ophthalmologists-in-training. PRP was undertaken in single session or multiple sessions (2–3 sessions) dependent on clinician preference. A full PRP technique was applied to PDR, I-CRVO and NVG cases using 1–1.5 spot spacing in Pascal® mode. A sectoral, scatter retinal photocoagulation technique with 1–1.5 spot spacing was used for I-BRVO cases. All macular laser was performed in a single session using 10–20 ms, and microaneurysms were not directly photocoagulated. In Pascal® mode, 5*5, 4*4, 3*3 arrays were used at the discretion of the laser operator for PRP laser (Fig. 1). The A plus B mode was used for macular laser, as well as single-spot laser (Fig. 2). In retinopexy, up to four rows of laser with 0.5 spot spacing were applied as a standard barrage treatment (Fig. 3).

Clinical outcomes

The clinical outcomes were evaluated at 3–5 months postlaser for all groups. The final visit was determined by patient compliance with hospital attendance, hence the variation in final follow-up duration. In groups A, D, E and G, complete regression of retinal and/or iris neovascularization was designated a successful outcome. In cases of laser retreatments, an unsuccessful outcome was recorded. In groups B, C and F, the absence of retinal thickening on clinical biomicroscopy was designated a successful outcome. Persistence of macular thickening was designated a treatment failure. Successful laser retinopexy (group H) was defined by a well-established laser surround scar. Patients requiring repeated laser retinopexy were designated treatment failures.

Statistical methods

Data were analysed using spss (Statistical Package for Social Sciences V.13). Descriptive statistics were used to summarize data and explore groups. Visual acuities (VA) were converted from Snellen to logMAR to explore changes in vision pre to postlaser using the Sign and Kruskal–Wallis tests. Chi-Square and one-way anova tests were used to assess any differences in gender or age between groups. For those patients in groups A, who had both previous conventional laser and Pascal®, differences between laser parameters were explored, particularly the power using T and Mann–Whitney U-tests. A two-tailed p value of <0.05 was considered significant (0.95 level of confidence).

Results

A total of 313 procedures were performed (152 left, 161 right) in 265 patients. Forty-eight patients received bilateral treatments. There were 118 (44.5%) women, 147 (55.5%) men, and the mean age was 59.3 years (SD 13.3, range 14–93, 95% CI 57.7–60.9). Age outliers (3) were 14, 16 and 93 years old. Gender did not vary significantly between groups A to H (chi-square 8.52, df = 7, p = 0.289).

Mean LogMAR visual acuity (LogMAR VA) for all procedures improved postlaser (pre-LogMAR 0.34; post-LogMAR 0.33; p = 0.065). Sixty-four (20.4%) procedures had an improvement in vision, and 202 (64.6%) had no change in vision. Laser prevented visual loss in 85% of eyes. Forty-seven eyes (15.0%) had deterioration in vision, with 28/47 as a result of ocular co-morbidity including cataract, epiretinal membrane, macular ischaemia, persistent cystoid macular oedema and macular degeneration. Visual and clinical outcomes are presented in Table 1.

Table 1.   Clinical and visual outcomes after laser.
GroupEyes treated
n (%)
Successful outcome
n (%)
Pre-LogMAR visual acuity
Mean (SD) Snellen VA
Post-LogMAR visual acuity
Mean (SD) Snellen VA
Follow-Up Period
Months (SD, range)
A
PDR
121 (38.9%)85 (70.5%)0.34 (0.29) 6/120.37 (0.36) 6/124.6 (2.0, 3–15)
B
CSMO
56 (17.9%)50 (89.3%)0.28 (0.26) 6/120.27 (0.25) 6/125.7 (3.2, 3–15)
C
Diffuse CSMO
61 (19.5%)38 (62.3%)0.41 (0.27) 6/180.38 (0.30) 6/125.1 (3.0, 3–16)
D
I-BRVO
13 (4.1%)9 (69.2%)0.38 (0.30) 6/120.32 (0.36) 6/126.2 (3.4, 3–14)
E
I-CRVO
6 (1.9%)5 (83.3%)0.49 (0.44) 6/180.47 (0.43) 6/184.7 (1.2, 3–6)
F
CMO
16 (5.1%)10 (62.5%)0.53 (0.26) 6/180.55 (0.39) 6/245.4 (2.2, 3–9)
G
NVG
9 (2.9%)2 (22.2%)0.86 (0.31) 6/36(0.0) 6/606.0 (2.6, 3–11)
H
Retinal Tear
31 (9.9%)26 (83.9%)0.08 (0.26) 6/7.50.05 (0.23) 6/64.0 (1.9, 3–12)
Totals313225 (72%)   

A successful clinical outcome was found in 225 procedures (72%), and 88 (28%) had unsuccessful outcomes that were defined by repeated laser treatment. Final clinical outcomes were not affected by age, sex or ocular co-morbidity. The mean follow-up time was 5 months (SD 2.6, range 3–16).

The laser parameters are outlined in Table 2. Multi-spot Pascal® PRP for PDR was performed using a mean 1374 burns at 30 ms. In DMO, average number of burns used was 99–159 at pulse duration 20 ms. A higher number of burns were used for I-CRVO and NVG, mean 1715 laser burns between groups E and G.

Table 2.   Pascal® laser parameters.
GroupnParameterMean valueSDRange
A
PDR
121Burns1374495.9200–2992
Power358.5 mW111.1150–800
Duration30 ms3020–200
Spot size216.1 μm0100–400
B
CSMO
56Burns9962.28–329
Power133.3 mW40.3100–350
Duration20 ms2010–100
Spot size100 μm0100
C
Diffuse CSMO
61Burns1599935–450
Power133.5 mW36.570–240
Duration20 ms3010–100
Spot size100 μm0100
D
I-BRVO/NVE
13Burns722360.3263–1320
Power321.2 mW72.2175–410
Duration50 ms7020–200
Spot size246.2 μm0200–400
E
I-CRVO/NVE
6Burns1720368.61328–2251
Power262.5 mW41.1200–300
Duration60 ms7020–200
Spot size183.3 μm0100–200
F
CMO
16Burns122117.733–459
Power177.5 mW76.9100–350
Duration10 ms510–20
Spot size100 μm0100
G
NVG
9Burns1710622.6879–3061
Power525 mW182300–750
Duration20 ms020
Spot size200 μm0200
H
Retinal Tear
31Burns232156.434–727
Power358.3 mW121.7175–575
Duration80 ms10010–400
Spot size212.9 μm0200–400

Subgroup analysis of PDR included 15 patients who underwent single-session 20-ms Pascal® PRP. After a single session with 0.06 logMAR visual gain, 67% (10/15) had a successful outcome. Treatment was uncomplicated in all eyes using a mean 1952 burns (SD 425.7, range 1502–2992) and power 333 mW (SD 73.1, range 180–450). In group A, 60% of eyes (72/121) had received an initial conventional, 100-ms argon laser PRP treatment followed by 20-ms Pascal® PRP. The differences in laser parameters were explored for those using Mann–Whitney-U tests. There was a significant difference in powers used between the two laser groups, with 370.8 mW for Pascal® PRP compared to 300.4 mW for argon PRP (p = 0.001).

In the macular oedema groups, 89% (50/56) in group B, 62% (38/61) in group C and 63% (10/16) in group F had successful outcomes following 10- to 20-ms laser. A primary macular treatment used 133.3 mW (group B), 133.5 mW (group C) and 177.5 mW (group F) to achieve resolution of retinal thickening on biomicroscopy postlaser.

Topical anaesthesia was used for all Pascal® laser treatments, and there was no requirement for subtenon’s or peribulbar anaesthesia in any case. In the multi-spot Pascal® mode using PRP arrays (5*5, 4*4, 3*3) and pattern A+B macular arrays, we did not observe direct damage to retinal blood vessels or any signs of choroidal/Bruch’s rupture (‘pops’), retinal or choroidal haemorrhage (Figs 1–3). At 3 months postPRP, 20-ms laser burns remained localized with minimal tissue carbonization and scar expansion (Fig. 4).

Figure 4.

 Patient with proliferative diabetic retinopathy treated with 20-ms Pascal® laser panretinal photocoagulation burns, with 5 × 5 array and 1.5 spot spacing. Colour fundus photograph at 3 months postlaser shows localized, pigmented laser burns that have minimal laser burn expansion.

Retinopexy was performed using applications of a multi-spot arc pattern (Fig. 5). After mean follow-up of 4 months, 84% of eyes were successfully treated with a single session of laser using mean power of 358.3 mW.

Figure 5.

 Patient presented with superior retinal break. Colour fundus photograph of Pascal® 20-ms laser retinopexy burns taken at 1-hr postlaser. Grey–white intensity burns are visible that provide adequate laser surround, with no signs of retinal or choroidal haemorrhages present.

Discussion

The Pascal® laser photocoagulator was first introduced to Manchester Royal Eye Hospital in November 2006. In 313 consecutive cases of retinal vascular disorders, there were no ocular complications or adverse effects observed using the multi-spot PRP and macular Pascal® modes at pulse durations 10–30 ms. Seventy-two per cent of procedures had a successful clinical outcome, and 10- to 30-ms laser prevented visual loss in 85% of eyes. Significantly higher powers were required for PRP using Pascal® compared to conventional argon laser in PDR, I-BRVO, I-CRVO and NVG.

Photocoagulation exerts its therapeutic effect secondary to thermal effects at the retinal pigment epithelium (RPE) but can produce concurrent collateral damage to adjacent tissues (Mellerio 1966). Pascal® technology provides an exposure time of 10–20 ms for macular photocoagulation and 20–30 ms for PRP, resulting in less energy delivered to the eye per burn (Blumenkranz et al. 2006). Healing of retinal laser lesions has been recently reported using medium pulse laser parameters (Jain et al. 2008; Paulus et al. 2008). Furthermore, we have previously described effective uptake of 10- to 20-ms laser burns within the outer retina in clinical practice (Muqit et al. 2008).

Retinal photocoagulation can be painful, presumably because of photocoagulation of ciliary nerves in the suprachoroidal space (Bloom & Brucker 1997), and this may lead to significant sub-optimal laser application in patients (Royal College of Anaesthetists and the Royal College of Ophthalmologists 2001). Pascal® technology uses short pulse durations that result in minimal diffusion of heat to adjacent areas (Krauss et al. 1987). A reduction in pulse duration has been reported to be associated with reduced pain responses in diabetic retinopathy (Al-Hussainy et al. 2008). In our experience of 10- to 30-ms laser, all treatments were adequately performed using topical anaesthesia alone.

In a previous pilot study, we reported our initial experience using medium pulse Pascal® laser (Sanghvi et al. 2008). The current study has shown the clinical and visual outcomes using 10- to 30-ms laser to be comparable with conventional argon laser in PDR and CSMO. We did not experience any complications directly related to the higher power used at shorter pulse durations. Pulse energy or fluence is calculated as power × time/area. Thus, for an equivalent treatment burn, the shorter pulse duration will require a reduced total energy per burn to achieve a threshold ETDRS burn intensity compared with conventional 100-ms argon laser.

One of our main aims was to establish the effectiveness of 10- to 30-ms Pascal® laser. We observed successful outcomes in 72% of patients and prevention of visual loss in 85% of eyes, and these findings are comparable to the published standards using argon laser (The Diabetic Retinopathy Study Research Group 1976; Early Treatment Diabetic Retinopathy Study 1991; Branch Vein Occlusion Study Group 1986; The Central Vein Occlusion Study Group report 1995; Kaiser et al. 2000). No complications related to laser settings were noted in any patient.

Laser treatment parameters have evolved over the years based on clinical experience. More recently, the Writing Committee for the Diabetic Retinopathy Clinical Research Network (2007), DRCRN investigated the potential role of a mild macular grid technique for patients with CSMO in comparison with modified ETDRS laser using a greater number of burns at barely visible laser spot intensity. In PDR treatment, contemporary laser parameters (200- to 500-μm spot size, 50- to 200-ms pulses) in clinical practice using conventional green or yellow wavelength argon laser have recently been reported Diabetic Retinopathy Clinical Research Network et al., (2009). However, laser treatment parameters for exudative and neovascular complications secondary to retinal vein occlusions have remained unchanged over time and continue to be adapted from the ETDRS laser guidelines.

There is evidence that resolution of retinal neovascularization is significantly related to the cumulative total number of burns and that successful photocoagulation requires considerably more treatment than that suggested by earlier studies. (Cordeiro et al.1997; Rogell 1983). Furthermore, a greater area of retinal ablation at the initial treatment session is associated with regression of neovascularization (Bailey et al. 1999). The Pascal® laser multiple-spot pattern allows the delivery of as many as 25 spots in approximately 500 ms for PRP. We found that an average of 1374 burns may be rapidly and effectively applied to the retina for a primary and full PRP, using multi-spot arrays with a spot spacing of 1–1.5 burns. A higher number of burns were required for single-session PRP in PDR and for full PRP in I-CRVO and NVG to achieve satisfactory retinal coverage.

Currently, clinicians remain hesitant to perform single-session primary PRP on patients with PDR, especially with good vision. Macular oedema in 43% and visual loss of two or more lines in 8% has been reported following PRP (McDonald & Schatz 1985). The DRCRN reported that PRP performed in a single sitting was not associated with any adverse effects on vision or macular thickness (Diabetic Retinopathy Clinical Research Network 2009). Patients with I-CRVO and NVG commonly have poor vision, and single-session PRP is undertaken more frequently. Our study has shown that the use 10- to 30-ms Pascal® laser appears to be a safe system for undertaking single-session treatments in retinal vascular disorders.

Macular burns produce loss of photoreceptors (Ishiko et al. 1998), and laser scars can expand postoperatively (Schatz et al. 1991). This occurs because the coagulated zone extends beyond the boundaries of the initial laser spot because of heat diffusion from the RPE target site. In contrast, at shorter pulses (10–30 ms), while the desired therapeutic effect of the laser spot is maintained, retinal damage is increasingly confined to the outer retina. The lesion sizes vary much less with laser power and remain comparable to the spot diameter. Pulse durations of approximately 10–20 ms represent an optimal compromise between speed, higher spatial localization, reduced collateral damage and sufficient width of the therapeutic window (Jain et al. 2008). When using Pascal® multi-spot arrays, the controlled spot spacing may further help to avoid overlapping of burns and collateral damage.

Macular photocoagulation achieved successful resolution of macular oedema in 63–89% of eyes treated. The multi-spot arrays were not associated with inadvertent damage to overlying macular vessels, and there were no signs of subretinal exudation, intraretinal bleeding or subretinal haemorrhage postlaser. We observed the Pascal® multi-spot arrays to allow rapid and safe application of laser burns.

The main limitations to this study were the retrospective design and lack of randomization. The Pascal® system was first introduced to the United Kingdom in 2006 at Manchester, and currently, this study is the only large clinical series that has investigated Pascal® in clinical practice. We feel that our results have an important relevance for ophthalmologists performing laser using 10- to 30-ms parameters. The large number of clinical disorders investigated may confound the interpretation of visual outcomes; however, the thresholds for 10- to 30-ms laser photocoagulation in macular oedema, retinal vein occlusions, NVG and retinopexy have yet to be published in the literature. Assessment of pain responses postlaser is currently being investigated in a prospective clinical trial at Manchester. The main strength of our study is the reporting of safe benchmarks for 10- to 30-ms laser photocoagulation using multi-spot Pascal®. Our findings will allow clinicians to evaluate the therapeutic window and thresholds for laser using these contemporary parameters.

Laser photocoagulation remains the first-line treatment for the complications of retinal vascular disease. In this large case series, we found the clinical efficacy using 10- to 30-ms pulse duration Pascal® laser to be comparable with published conventional standard protocols used for retinal disorders. Higher power, multi-spot arrays of 10- to 30-ms pulse duration laser may be safely used in clinical practice with effective prevention of visual loss.

Acknowledgements

Financial Disclosure: George Marcellino is an employee of Optimedica Corporation. Paulo Stanga has received financial support from Optimedica Corporation. Funding support received from Optimedica Corporation. Work has been presented in May 2009 at the Association of Research and Vision in Ophthalmology Meeting, Fort Lauderdale, USA. Work has been presented in May 2009 at the Royal College of Ophthalmologists Annual Congress, UK. Support for the NIHR Manchester Biomedical Research Centre is acknowledged.

Ancillary