Graft-versus-host disease (GVHD) is a multi-organ complication from allogeneic bone marrow or hematopoietic stem cell transplantation used to treat malignant and benign hematologic diseases. The clinical hall-mark of ocular GVHD (oGVHD) is dry eye disease (DED) [1]. Autologous serum tears (AST) have been used to treat severe DED and are now integrated into oGVHD treatment [2]. Little is known about their effect on corneal nerves and corneal sensitivity in patients with oGVHD. We evaluated the effects of 12-week AST treatment on signs, symptoms, corneal sensitivity, and sub-basal nerve density (SND) in patients with moderate-to-severe DED resulting from oGVHD.
This single-center, prospective, observational study was approved by the Institutional Review Board of the Massachusetts Eye and Ear, conducted in accordance with the requirements of the Health Insurance Portability and Accountability Act, and adhered to the tenets of the Declaration of Helsinki. Patients aged 18–80 years old diagnosed with chronic GVHD who had moderate to severe DED were included. Moderate to severe DED was defined as having symptoms of DED for at least one month, Ocular Surface Disease Index (OSDI) survey score >22, and corneal fluorescein staining (CFS) ≥ 4 in at least one eye that failed to respond to previous treatments including lubrication, anti-inflammatory medications, and punctal occlusion (at least one punctum per eye). AST was prepared by collecting serum from 6 vials (approximately 42 ml total) of centrifuged blood from each patient. It was diluted to 20% with preservative free 0.9% injection quality sodium chloride and filtered with 0.22-μm filters. Five ml AST was then aliquoted into 15 ml bottles. Patients administered AST in each eye six times daily for 12 weeks. Patients were excluded if they had any change in anti-inflammatory or glaucoma regimen within one month, active ocular allergies, history of contact or scleral lenses or intraocular surgery within three months before enrollment or were unable to give informed consent.
The following clinical data were collected: CFS (NEI scale, graded 0–15), tear break up time (TBUT), and conjunctival lissamine green (LG) staining (NEI scale, graded 0–18). A Schirmer’s test with a topical anesthetic was performed for 5 min with eyes closed. Corneal sensitivity was measured with the Cochet–Bonnet esthesiometer. Quality of life and clinical symptoms were measured using the visual functioning questionnaire (NEI-VFQ-25; graded 0–100), the OSDI (graded 0–100), and symptom assessment in dry eye (SANDE) questionnaires I and II at each visit. SANDE I measures the frequency and severity of symptoms on a visual analog scale (0–100). SANDE II measures improvement (−50 to 0) or worsening (0–50) of frequency and severity of symptoms between visits [3]. Follow-up visits occurred at 6 weeks (±1 week) and 12 weeks (±2 weeks) after initiating AST.
Corneal SND was determined using in vivo confocal microscopy (IVCM) by measuring the corneal sub-basal nerve fiber lengths (SNFL) of 400 × 400 μm corneal sections. Five scans were obtained by a masked photographer and SNFL were traced using NeuronJ (http://www.imagescience.org/meijering/software/neuronj/), a semi-automated nerve analysis plug-in of ImageJ (National Institutes of Health, USA). SND was defined as total SNFL divided by a 0.16-mm2 area. Two observers measured SNFL independently. The mean value of both observers was calculated and the average SND values of both eyes was used.
R-3.6.1 Software (R Core Team, 2019, Vienna, Austria) was used for statistical analysis. Data for each variable were calculated by averaging values from both eyes. A linear mixed-effects model for repeated measures (MMRM) was used to assess the effect of AST in corneal signs and symptoms between baseline, 6 weeks and 12 weeks; and post hoc pairwise comparisons were done. Data is reported as mean ± standard deviation (SD). A p value of less than 0.05 was considered statistically significant.
Twenty patients with moderate-to-severe DED due to GVHD were enrolled, and 3 dropped out after baseline visit and were removed from analysis. Among the remaining 17 patients, 3 missed the 6-week visit and 3 patients missed the 12-week visit, while 11 completed both follow-up visits.
The average age was 65.8 years (range 33–78 years), with eleven (64.7%) females and six (35.3%) males. As shown in Table 1, patients reported improvement in DED symptoms with lower scores on OSDI (p = 0.007) and SANDE I in frequency (p = 0.04) and severity (p = 0.04) after treatment with AST. The corneal sensitivity (length of filament) increased at 6 week (4.57 ± 1.91 cm) and 12 week (5.16 ± 1.26 cm), compared to baseline (3.11 ± 2.38 cm, p = 0.001). There was also an increase in SND at 6 weeks (11.50 ± 6.12 mm/mm2) and 12 weeks (12.60 ± 6.26 mm/mm2), compared to baseline (8.82 ± 5.05 mm/mm2, p = 0.006). There were no statistically significant changes in clinical signs of DED including TBUT, Schirmer’s test, CFS, or LG staining.
Table 1.
Clinical Characteristics in Patients with oGVHD at baseline and after treatment with Autologous Serum Tears.
| N = 17 | Baseline mean ± SD | 6 weeks mean ± SD | 12 weeks mean ± SD | P-Value |
|---|---|---|---|---|
| OSDI | 57.10 ± 21.40 | 41.71 ± 25.33 | 36.21 ± 23.45 | 0.007 |
| SANDEI | ||||
| Frequency | 58.88 ± 28.74 | 32.35 ± 29.30 | 37.76 ± 30.75 | 0.04 |
| Severity | 58.64 ± 30.79 | 32.52 ± 27.71 | 38.29 ± 28.45 | 0.04 |
| SANDE II | ||||
| Frequency | −12.64 ± 13.76 | −2.11 ± 17,41 | 0.09 | |
| Severity | −14.35 ± 112.99 | −2.05 ± 16.74 | 0.06 | |
| TBUT (seconds) | 1.47 ± 2.02 | 1.67 ± 1.79 | 1.32 ± 1.98 | 0.87 |
| VFQ-25 | 70.57 ± 17.93 | 68.87 ± 24.41 | 61.86 ± 29.95 | 0.31 |
| Schirmer’s Test (mm) | 4.50 ± 5.5 | 3.66 ± 4.36 | 3.11 ± 3.87 | 0.5 |
| CFS | 7.79 ± 3.48 | 6.52 ± 3.50 | 6.47 ± 4.65 | 0.44 |
| Lissamine green staining | 4.88 ± 5.11 | 3.41 ± 3.78 | 2.94 ± 3.31 | 0.19 |
| Corneal Sensitivity (cm) | 3.11 ± 2.38 | 4.57 ± 1.91 | 5.16 ± 1.26 | 0.001 |
| SND (mm/mm2) | 8.82 ± 5.05 | 11.50 ± 6.12 | 12.60 ± 6.26 | 0.006 |
Results of mixed-effects model for repeated measures (MMRM) analysis are reported as mean ± standard deviation (SD). OSDI—ocular surface disease index, SANDE—symptom assessment in dry eye, TBUT—tear break-up time, VFQ-25—visual function questionnaire, CFS—corneal fluorescein staining, SND—sub-basal nerve density.
The post-hoc analysis revealed a significant difference in OSDI score between baseline and 6 weeks (p = 0.012), and between baseline and 12 weeks (p = 0.006) without a significant difference between weeks 6 and 12. The frequency and severity of symptoms assessed by SANDE I showed similar patterns, with significant difference between baseline and 6 weeks (p = 0.016 and p = 0.006, respectively), but not between weeks 6 and 12. The improvement in corneal sensitivity was significant from baseline to 6 week (p = 0.015) and from baseline to the 12 week follow up (p = 0.002), but not significant between 6 and 12 weeks. There was also a significant difference in SND at baseline compared to 6-weeks (p = 0.04) and from baseline to 12 weeks (p = 0.01) but no significant difference between 6 and 12 weeks.
In this prospective study, corneal sensitivity and SND increased in patients with moderate to severe DED associated with oGVHD after 12 weeks of AST treatment (as early as 6 weeks). Dry eye symptoms decreased with AST treatment, while clinical signs remained unchanged. This suggests that AST treatment may improve corneal nerve density and sensation with an amelioration of dry eye symptoms prior to improvement in clinical signs.
A similar study treating DED patients with 20% AST for three months found an improvement in OSDI and CFS (Oxford scale) but no change in SND [4]; however, less than half of the subjects had oGVHD and corneal sensitivity was not measured. Additionally, they measured the nerve density using a scanning slit corneal microscope that is different from the confocal scanning laser microscope used in our study. Measurements between these two devices are not interchangeable [5]. These differences can potentially explain the discrepancy in findings. In another study, patients with oGVHD were treated with 100% AST over 6 months and had improvement in visual acuity, CFS and OSDI; however this retrospective study did not measure SND or corneal sensitivity [2]. Patients with GVHD have impaired visual functioning and quality of life and the VFQ-25 has been reported to correlate with OSDI and CFS score in GVHD patients [6]. But in our study, despite an improvement in OSDI, we found no changes in VFQ-25 or CFS.
Patients with DED have decreased corneal sensitivity that correlates with decreased SND [7], and corneal sensitivity improves with corneal nerve regeneration [8]. Indeed in our study, corneal sensitivity and SND improve significantly as early as 6 weeks after AST treatment and remain increased at 12 weeks, compared to baseline. These changes, however, appear to plateau between 6 and 12 weeks. Similarly, we observed significant improvement in symptoms (OSDI and SANDE I) between baseline and 6 weeks, but no additional significant improvement between 6 and 12 weeks. This trend suggests that the onset of AST on corneal innervation and symptomatology may be rather fast (within weeks), and its effects maintain afterwards. Our study can serve as a pilot for larger scale prospective study with longer follow-up to answer questions including if AST regimen should be maintained chronically and if the frequency of AST use can be tapered after the initial intense (6 times/day) regimen.
There are limitations to our study. Our study has a small number of cases, and several patients were lost to follow-up due to long distance travel issues, GVHD-related systemic comorbidities, and the COVID-19 pandemic. We followed patients for 12 weeks and long-term effects were not assessed. Observers measuring SND were not masked which may impose bias. There was no control group and due to the observational nature of the study, the improvements cannot be fully attributed to the treatment with AST. Our cohort was treated at a tertiary eye care center and findings may not be generalizable to other populations.
In summary, AST is a therapeutic option for treating moderate to severe DED associated with oGVHD. As early as 6 weeks, it improves corneal nerve density, corneal sensitivity, and DED symptoms. These findings warrant further long-term and a larger scale investigation to determine the benefits of AST as a treatment option for this serious ocular surface condition.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Footnotes
Declaration of competing interest
The authors do not have any conflicts of interest to disclose.
References
- [1].Ogawa Y, Dana R, Kim S, et al. Multicenter prospective validation study for international chronic ocular graft-versus-host disease consensus diagnostic criteria. Ocul Surf Sep 18 2022;26:200–8. 10.1016/j.jtos.2022.09.002. [DOI] [PubMed] [Google Scholar]
- [2].Tahmaz V, Gehlsen U, Sauerbier L, et al. Treatment of severe chronic ocular graft-versus-host disease using 100% autologous serum eye drops from a sealed manufacturing system: a retrospective cohort study. Br J Ophthalmol Mar 2017;101 (3):322–6. 10.1136/bjophthalmol-2015-307666. [DOI] [PubMed] [Google Scholar]
- [3].Schaumberg DA, Gulati A, Mathers WD, et al. Development and validation of a short global dry eye symptom index. Ocul Surf Jan 2007;5(1):50–7. 10.1016/s1542-0124(12)70053-8. [DOI] [PubMed] [Google Scholar]
- [4].Mahelkova G, Jirsova K, Seidler Stangova P, et al. Using corneal confocal microscopy to track changes in the corneal layers of dry eye patients after autologous serum treatment. Clin Exp Optom May 2017;100(3):243–9. 10.1111/cxo.12455. [DOI] [PubMed] [Google Scholar]
- [5].Schrems WA, Hoesl LM, Dastjerdi M, Dana R, Pavan-Langston D, Hamrah P. In vivo confocal microscopy comparison of anterior human corneal structures by laser scanning and white light systems in normal and diseased corneas. Investig Ophthalmol Vis Sci 2009;50(13). 3710–3710. [Google Scholar]
- [6].Saboo US, Amparo F, Abud TB, Schaumberg DA, Dana R. Vision-related quality of life in patients with ocular graft-versus-host disease. Ophthalmology Aug 2015;122 (8):1669–74. 10.1016/j.ophtha.2015.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Benitez-Del-Castillo JM, Acosta MC, Wassfi MA, et al. Relation between corneal innervation with confocal microscopy and corneal sensitivity with noncontact esthesiometry in patients with dry eye. Invest Ophthalmol Vis Sci Jan 2007;48(1): 173–81. 10.1167/iovs.06-0127. [DOI] [PubMed] [Google Scholar]
- [8].Muller RT, Abedi F, Cruzat A, et al. Degeneration and regeneration of subbasal corneal nerves after infectious keratitis: a longitudinal in vivo confocal microscopy study. Ophthalmology Nov 2015;122(11):2200–9. 10.1016/j.ophtha.2015.06.047. [DOI] [PMC free article] [PubMed] [Google Scholar]