You are here

Article Text

Article menu

Download PDFPDF

Clinical science
Differential effects of primary disease and corneal vascularisation on corneal transplant rejection and survival
  1. Daniel Sibley1,
  2. Cathy L Hopkinson2,
  3. Stephen J Tuft1,
  4. Stephen B Kaye3,
  5. Daniel F P Larkin1
  6. On behalf of the National Health Service Blood and Transplant Ocular Tissue Advisory Group and contributing ophthalmologists (OTAG Study 26)
  1. 1 Moorfields Eye Hospital NHS Foundation Trust, London, UK
  2. 2 Statistics and Clinical Studies, National Health Service Blood and Transplant, Bristol, UK
  3. 3 Ophthalmology, Royal Liverpool University Hospital, Liverpool, UK
  1. Correspondence to Dr Daniel F P Larkin, Moorfields Eye Hospital NHS Foundation Trust, London EC1V 2PD, UK; Frank.Larkin{at}moorfields.nhs.uk

Abstract

Aims To investigate the relative risk of pretransplant corneal vascularisation on rate of rejection and graft failure within 5 years of surgery when categorised by indication for transplantation.

We analysed all adults recorded in the UK transplant registry who had a first cornea transplant for keratoconus (KC), pseudophakic bullous keratopathy (PBK) or previous infection (viral/bacterial/fungal/protozoan) between 1999 and 2017. We analysed the number of quadrants of the recipient cornea vascularised before transplant and type of vascularisation, the interval post-transplant to rejection, if any, and the outcome at 5 years post-transplant. Risk factors for rejection and transplant failure were modelled by multivariable risk-adjusted Cox regression.

Results Corneal vascularisation was recorded in 10%, 25% and 67% of patients with KC, PBK and infection, respectively. Individuals with PBK had an increased hazard of transplant rejection only when there were more than two quadrants of vascularisation (HR 1.5, p=0.004) when either superficial and/or deep vascularisation was present (HR 1.3 and 1.4, respectively, p=0.004). Individuals who had a transplant for previous infection had an increased hazard of rejection with four quadrants of vascularisation (HR 1.6, p=0.003). There was no risk-adjusted increase in transplant failure associated with vascularisation in any group. There was weak evidence of reduction in risk of rejection and/or failure associated with lamellar compared with penetrating transplantation in KC and PBK in vascularised recipient corneas.

Conclusion Vascularisation is a risk factor for corneal allograft rejection within 5 years. The indication for transplantation has a clinically significant effect on the magnitude of this risk.

  • cornea
  • eye (tissue) banking
  • neovascularisation

https://doi.org/10.1136/bjophthalmol-2019-314200

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Corneal transplantation is one of the most successful and commonly performed types of transplant.1 Although there is usually a low risk of failure there are a number of important risk factors associated with reduced graft survival. These include corneal vascularisation, active corneal inflammation or infection, previous failed graft, glaucoma and limbal stem cell failure. 2 3 The survival rate for penetrating keratoplasty (PK) is approximately 90% at 1 year and corneal allograft rejection episodes occur in approximately 1:6 patients who have had a PK.4 5 This low risk of allograft rejection compared with other transplanted tissues is a result of the relative immune privilege of corneal allografts, to which the avascularity of the cornea is an important contributory factor.6 The level of corneal vascularisation is commonly related to the level of pathology, with superficial pathology producing superficial vascularisation and deep pathology causing deep vascularisation.7 Corneal vascularisation facilitates alloimmunity—allowing alloreactive cells of the recipient immune system to gain early access to the graft.

    Of the reported risk factors for corneal transplant failure, preoperative vascularisation of the recipient cornea has the most statistically significant effect on actuarial survival.8–10 A meta-analysis of outcomes of 24 944 transplants reported an increased incidence of graft rejection and failure with an increasing number of quadrants of corneal vascularisation prior to transplantation.9 Despite the fact that vascularisation has been identified as having a significant negative influence on allograft survival, little is known of (1) its prevalence in large corneal transplant populations, (2) whether the effect of vascularisation varies according to the primary disease indication for transplantation, and (3) whether the effect of vascularisation differs according to whether the transplant is penetrating or lamellar. The prevalence of vascularisation is variable in corneal diseases, ranging from common after infection and pseudophakic bullous keratopathy (PBK), to uncommon in keratoconus (KC). With increased interest in interventions and pharmacological approaches that ameliorate the effects of corneal angiogenesis on transplant survival, there is a need for information on the prevalence of corneal vascularisation and its influence on transplant outcomes. The aims of this study were to examine the prevalence of pretransplant recipient corneal vascularisation and the relative risk of transplant rejection and failure according to the indication for transplantation and type of transplant undertaken.

    Methods

    Data are collected by NHS Blood and Transplant (NHSBT) on all corneal transplants performed in the UK from transplant record forms and follow-up forms at 1, 2 and 5 years post-transplant. The provision of this is a requirement of all surgeons undertaking corneal transplantation in the UK and data are stored on a database maintained by NHSBT.

    Patients, indication for transplant and outcomes

    All adults (>18 years) recorded on the UK Transplant Registry who received their first corneal transplant for KC, PBK or following infection (viral, bacterial, fungal and protozoan) between 1 April 1999 and 31 March 2017 were included. KC was chosen for analysis as a common disorder with a low risk of vascularisation. Transplant rejection was defined as any rejection episode (reversible or irreversible) of stromal or endothelial rejection in cases with PK, endothelial rejection in patients with endothelial grafts and stromal rejection in anterior lamellar grafts. Transplant failure was defined as a case identified on follow-up with either primary graft failure, irreversible rejection, acquired infection, recurrence of disease, endothelial decompensation, or dislocation of an endothelial keratoplasty (EK) graft. The interval from transplant to diagnosis of rejection or failure was calculated as rejection-free survival and graft survival, respectively.

    Pretransplant corneal vascularisation

    The number of preoperative quadrants (0–4) of vascularisation was recorded separately for superficial and deep vascularisation of the patient’s cornea. The numbers were too small to investigate these factors in combination, therefore the number of quadrants of vascularisation and the type of vascularisation (superficial, deep or both) were considered in separate models.

    Statistical analysis

    Kaplan-Meier estimates of rejection-free and transplant survival at 5 years were quantified for all transplants according to the number of preoperative quadrants of vascularisation (model 1) or type of vascularisation (model 2). Transplant rejection and survival up to 5 years were compared using the log-rank test. Multivariable Cox regression models for transplant survival and rejection were fitted using stepwise selection methods. Separate models were fit for each corneal diagnosis. Risk factors that were considered in the Cox regression models were: gender and age of the patient, graft type (PK, EK or deep anterior lamellar keratoplasty (DALK)), H-Y antigen-mismatched transplants (Yes: M→F, No: F→F, M→M or F→M),11 whether the transplant was human leucocyte antigen matched, presence of ocular surface disease and risk of glaucoma. For each risk factor and model, a value of p≤0.05 based on the likelihood ratio test (comparing models with and without the risk factor) was deemed sufficient evidence for the risk factor to be included in the model, only factors included are listed. In addition, the type of infection (viral, bacterial, fungal or protozoan) was included in the model for transplants undertaken following infection. HRs described the hazard of transplant rejection or failure for one or more quadrants of vascularisation compared with no quadrants or superficial, deep and/or both compared with no vascularisation, after adjustment for risk factors. To determine whether the presence of vascularisation increased the hazard of transplant rejection or failure for PK compared with lamellar transplants (EK and DALK), models including an interaction between transplant type and the number of quadrants or type of vascularisation were also analysed and are shown in the online supplementary tables 1a,b, 2a,b. Transplant survival was recorded as a percentage with 95% confidence limits (CI).

    Results

    Complete data were available from 13 435 transplants. Data on 2051 (13%) transplants were incomplete and these were excluded. There were 6174 transplants for KC, 4863 for PBK and 2398 after infection.

    Pretransplant vascularisation categorised by primary corneal disease

    Corneal vascularisation was recorded in 10%, 25% and 67% of patients with KC, PBK and infection, respectively. KC corneas with vascularisation in two, three or four quadrants were categorised together as ‘2 or more’ quadrants due to the limited number in the groups with three and four-quadrant vascularisation. Similarly, there were limited numbers of patients with deep only vascularisation for each indication and these were grouped for analysis as eyes having superficial and/or deep vascularisation. In KC and infection transplants, the majority had deep and superficial vascularisation (figure 1). In the majority of patients who had undergone surgery for PBK, the vascularisation was mostly superficial only (figure 1).

    Figure 1
    Figure 1

    Number of patients with keratoconus (KC), pseudophakic bullous keratopathy (PBK) and infection receiving transplants from 1999 to 2017 with superficial, deep or no vascularisation.

    Overall impact of vascularisation on rejection-free and transplant survival

    When we considered all three transplant indications together, differences in survival rates correlated with the number of quadrants and type of corneal vascularisation (p<0.0001 for both interval to first rejection and failure). The 5-year rejection-free survival rate was 83% (95% CI 82% to 84%) in avascular corneas compared with 70% (95% CI 64% to 74%) in corneas with four quadrants of vascularisation (figure 2A). The 5-year transplant survival rate was 78% (95% CI 77% to 79%) in avascular corneas compared with 50% (95% CI 45% to 55%) when there was vascularisation in four quadrants. Transplant survival decreased with increasing number of quadrants vascularised (figure 2B). The survival rate of 61% (95% CI 57% to 64%) for superficial vascularisation was lower for 5-year graft survival than either deep and superficial and/or deep (figure 3B). This might be due to other contributing risk factors that have not been accounted for. The survival rates for the type of vascularisation ranged from 76% to 78% for 5 years rejection free compared with 82% (95% CI 82% to 84%) for no vascularisation (figure 3A).

    Impact of vascularisation on survival in individual primary corneal disorders

    Results of the multivariable Cox regression models for rejection and failure for pretransplant vascularisation, according to the indication for transplant, are shown in tables 1 and 2, respectively. Transplant type and patient age were the most common risk-adjusted factors across indications. Following risk adjustment, for patients who had undergone transplantation for KC there was no evidence to suggest that preoperative corneal vascularisation was associated with transplant rejection (p=0.66) or failure (p=0.17) nor type of vascularisation with transplant rejection (p=0.96) or failure (p=0.21).

    Figure 2
    Figure 2

    Kaplan-Meier estimates of rejection-free (A) and transplant (B) survival according to the number of corneal quadrants vascularised prior to all transplants for keratoconus (KC), pseudophakic bullous keratopathy (PBK) or infection.

    Figure 3
    Figure 3

    Kaplan-Meier estimates of rejection-free (A) and transplant (B) survival according to the type of vascularisation prior to all transplants for keratoconus (KC), pseudophakic bullous keratopathy (PBK) or infection.

    View this table:
    Table 1

    Multivariable risk-adjusted Cox regression models for transplant rejection within 5 years

    View this table:
    Table 2

    Multivariable risk-adjusted Cox regression models for transplant failure at 5 years

    Patients who had undergone transplantation for PBK were at greater hazard of rejection when two to four quadrants were vascularised (p≤0.01 in 1 and 4), with approximately 1.5 times greater hazard of rejection (95% CI 1.1 to 1.9, 1.0–2.1, 1.1–2.0 for two, three and four quadrants, respectively) compared with corneas with no vascularisation. Similarly, we found an association between type of vascularisation and graft rejection (p=0.004). The hazard of rejection was 1.3 times greater (95% CI 1.1 to 1.6) for patients with superficial vascularisation and 1.4 times greater for patients with deep and/or superficial vascularisation (95% CI 1.1 to 1.8) compared with patients with no vascularisation. There was, however, no evidence that number of quadrants or type of vascularisation was associated with transplant failure in patients with PBK (p=0.57 and p=0.33, respectively), suggesting that other factors conferred a greater risk of failure in eyes with PBK.

    The hazard of rejection for corneas with infection was found to be greater only in patients with four quadrants of vascularisation (HR=1.6, 95% CI 1.2 to 2.1, p=0.003). In patients who had corneal transplantation for infection, the hazard of rejection was greater only in patients with vascularisation in four quadrants (HR=1.6, 95% CI 1.2 to 2.1, p=0.003). There was no evidence that increasing quadrants of vascularisation was associated with increased transplant failure for infection corneas (p=0.05), indicating evidence of a counterintuitive decreased risk for patients with vascularisation in only one quadrant. We found no evidence of an association between type of vascularisation and graft rejection (p=0.35) or graft failure (p=0.75) for patients with infection.

    Differential impact of vascularisation according to transplant type

    We found no evidence of an interaction between pretransplant vascularisation and the type of surgery (penetrating or lamellar), either for rejection or transplant failure in patients with KC, PBK and infection (online supplementary tables 1a,b and 2a,b). The effect of preoperative vascularisation and the type of corneal transplantation were independent and that the risks hazard of transplant rejection and failure associated with the number of quadrants vascularised before transplant was not influenced by whether the transplant was a PK or a lamellar transplant.

    Discussion

    This report of 13 435 first corneal transplants demonstrates an overall trend for an increased hazard of graft rejection with increasing quadrants of preoperative vascularisation in the host cornea. However, when data on corneal diagnoses are grouped together the Kaplan-Meier actuarial survival plots appear to overestimate the influence of pretransplant vascularisation and mask differences in risk according to primary corneal disease. Once differences in risk by indication were adjusted to control for confounding factors in the Cox regression models, the hazard of graft rejection was increased only in PBK and infection if there were more than two and four quadrants of vascularisation, respectively. The type of vascularisation was associated with an increased hazard of graft rejection in patients with PBK. On account of the well-recognised impact of recipient corneal vascularisation on PK survival,8 9 it is surprising that lamellar surgery did not confer a significant benefit compared with PK in respect of transplant rejection or failure within 5 years. This suggests the loss of angiogenic privilege caused by vascularisation is not mitigated by lamellar grafting. However, there are few if any published reports on the effect of a recipient corneal vascularisation on outcomes in DALK and EK procedures.

    Several antiangiogenic molecules have been identified in maintaining an avascular cornea including vascular endothelial growth factor receptor 2 (VEGFR-2) and endostatin. Both VEGFR-2 and endostatin are secreted by corneal epithelial and stromal cells and exogenous administration of a soluble VEGFR-2 inhibits lymphangiogenesis but not haemangiogenesis.12 Haemangiogenesis is associated with lymphangiogenesis13–15 and the latter is now understood to be critical in T cell-mediated rejection. Selective blockage of lymphangiogenesis in murine corneal transplantation has been reported to reduce rejection without affecting haemangiogenesis14 and the removal of ipsilateral lymph nodes prevented rejection.16 While imaging of lymphatics in vivo in corneal transplant patients is difficult, these experimental studies demonstrate that presence independently of lymph and blood vessels in the recipient bed indicate abrogation of angiogenic privilege implies loss of immune privilege and higher allograft rejection risk. Corneal vessels facilitate access of recipient immune cells to donor tissue and from donor cornea to recipient lymphoid tissue, allowing indirect allorecognition by the innate and adaptive immune systems.10

    The results from this study agree in part with previous reports of an increased risk of rejection in recipient corneas in which angiogenic privilege is lost.4 Our study did not find an increase in graft failure with increasing vascularisation when adjusted for risks including patient age and glaucoma. Vail et al reported17 that the presence of deep vessels increased the relative risk of graft failure by 1.53 but did not quantify vascularisation by number of vascularised quadrants or differentiate by disease indication for transplantation. In contrast, the Collaborative Corneal Transplantation Study,18 a prospective trial with a protocol for documentation of corneal vascularisation, did not find evidence for an increase in graft rejection or failure with vascularisation. There have been several subsequent large studies suggesting vascularisation as a risk factor for graft failure.19–21 A meta-analysis of 19 reports on patients undergoing PK for all indications found vascularisation increased the risk of graft failure with a pooled HR of 1.32 (95% CI 1.15 to 1.49, p=0.057).9 This is similar to our finding in KC but evidence was less strong (HR 1.5, p=0.08). The patients we report with PBK and infection had high rates of vascularisation and high rates of failure compared with KC. Although the higher rates of transplant failure in patients with PBK and infection are well known, we found that the high rate of failure was independent of vascularisation. One possibility is that the higher failure rate in these disorders may be accounted for by higher numbers of inflammatory cells in the recipient corneas at the time of transplant. High leucocyte counts in cornea at the time of transplant have been reported to correlate with higher risk of rejection22 and earlier graft failure.23

    There are two main limitations for this study. First, the categorisation of vessels as superficial or deep was dependent on surgeon-reported findings at pretransplant examination, rather than defined according to a descriptive protocol such as would be used in a clinical trial. Nor did vascularisation data entry enable description of density of vascularisation in any or multiple quadrants. As deep vessels are thought to confer higher risk15 24 this merits further investigation. The degree and extent of corneal vascularisation detected using biomicroscopy is often much less than is apparent using fluorescein and indocyanine green angiography.25 26 Second, details of any treatment for vascularisation or of immunosuppression including topical steroid before or after surgery were not available for analysis and the impact of such treatment could be examined only in a large randomised trial. These two limitations would be shared with any large multicentre or national transplant databases, but compensated for by large transplant numbers and statistical power. In addition to these limitations it must be recognised that presence of lymph vessels in corneas before transplant, unidentifiable at slit-lamp biomicroscope examination, may influence risk of rejection and subsequent failure to an equivalent or greater extent than blood vessels.14 27

    There are three main clinical implications from this study. First, it emphasises the need for surgeon and patient awareness of the relative influence of corneal vascularisation on the rejection of any proposed transplant. Interventions additional to topical corticosteroid should be considered in patients with PBK and infection in order to reduce risk of rejection.28 29 Second, it highlights the need for further research into effectiveness of interventions to reduce or eliminate the effect of established corneal vessels. Third, it indicates the necessity for equivalent allocation of patients in clinical trial groups with primary corneal disorders conferring variable risk. In these patients, anti-VEGF approaches have not been as effective and other pharmacological approaches,30 needle diathermy31 and photodynamic therapy32 have been investigated. A number of potential therapeutic targets have been identified in murine models with suture-induced corneal vascularisation, including VEGF-A,33 VEGFR-3,34 very late antigen-134 and VEGF-C.35

    In summary this report from a large corneal transplant data set confirms vascularisation as a risk factor for corneal allograft rejection at 5 years, but identifies differential influence on risk depending on the corneal disease indication for transplantation.

    References

    1. ↵
      2. Gain P ,
      3. Jullienne R ,
      4. He Z , et al
      . Global survey of corneal transplantation and eye banking. JAMA Ophthalmol 2016;134:167–73.doi:10.1001/jamaophthalmol.2015.4776
      OpenUrl
    2. ↵
      2. Williams KA ,
      3. Roder D ,
      4. Esterman A , et al
      . Factors predictive of corneal graft survival. Report from the Australian corneal graft registry. Ophthalmology 1992;99:403–14.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Brown SI ,
      3. Bloomfield SE ,
      4. Pearce DB
      . A follow-up report on transplantation of the alkali-burned cornea. Am J Ophthalmol 1974;77:538–42.doi:10.1016/0002-9394(74)90468-1
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      2. George AJT ,
      3. Larkin DFP
      . Corneal transplantation: the forgotten graft. Am J Transplant 2004;4:678–85.doi:10.1111/j.1600-6143.2004.00417.x
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Amouzegar A ,
      3. Chauhan SK ,
      4. Dana R
      . Alloimmunity and tolerance in corneal transplantation. J.i. 2016;196:3983–91.doi:10.4049/jimmunol.1600251
      OpenUrl
    6. ↵
      2. Niederkorn JY ,
      3. Larkin DFP
      . Immune privilege of corneal allografts. Ocul Immunol Inflamm 2010;18:162–71.doi:10.3109/09273948.2010.486100
      OpenUrlCrossRefPubMed
    7. ↵
      2. Faraj LA ,
      3. Said DG ,
      4. Al-Aqaba M , et al
      . Clinical evaluation and characterisation of corneal vascularisation. Br J Ophthalmol 2016;100:315–22.doi:10.1136/bjophthalmol-2015-306686
      OpenUrlAbstract/FREE Full Text
    8. ↵
      2. Williams KA ,
      3. Lowe M ,
      4. Bartlett C , et al
      . Risk factors for human corneal graft failure within the Australian corneal graft registry. Transplantation 2008;86:1720–4.doi:10.1097/TP.0b013e3181903b0a
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      2. Bachmann B ,
      3. Taylor RS ,
      4. Cursiefen C
      . Corneal neovascularization as a risk factor for graft failure and rejection after keratoplasty. Ophthalmology 2010;117:1300–5.doi:10.1016/j.ophtha.2010.01.039
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. Cursiefen C ,
      3. Cao J ,
      4. Chen L , et al
      . Inhibition of Hemangiogenesis and Lymphangiogenesis after Normal-Risk Corneal Transplantation by Neutralizing VEGF Promotes Graft Survival. Invest. Ophthalmol. Vis. Sci. 2004;45:2666–73.doi:10.1167/iovs.03-1380
      OpenUrlAbstract/FREE Full Text
    11. ↵
      2. Hopkinson CL ,
      3. Romano V ,
      4. Kaye RA , et al
      . The influence of donor and recipient gender incompatibility on corneal transplant rejection and failure. Am J Transplant 2017;17:210–7.doi:10.1111/ajt.13926
      OpenUrl
    12. ↵
      2. Albuquerque RJC ,
      3. Hayashi T ,
      4. Cho WG , et al
      . Alternatively spliced vascular endothelial growth factor receptor-2 is an essential endogenous inhibitor of lymphatic vessel growth. Nat Med 2009;15:1023–30.doi:10.1038/nm.2018
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      2. Smolin G ,
      3. Hyndiuk RA
      . Lymphatic drainage from vascularized rabbit cornea. Am J Ophthalmol 1971;72:147–51.doi:10.1016/0002-9394(71)91605-9
      OpenUrlPubMedWeb of Science
    14. ↵
      2. Dietrich T ,
      3. Bock F ,
      4. Yuen D , et al
      . Cutting edge: lymphatic vessels, not blood vessels, primarily mediate immune Rejections after transplantation. J.i. 2010;184:535–9.doi:10.4049/jimmunol.0903180
      OpenUrl
    15. ↵
      2. K. Chauhan S ,
      3. Dohlman TH ,
      4. Dana R
      . Corneal lymphatics: role in ocular inflammation as inducer and Responder of adaptive immunity. J Clin Cell Immunol 2014;05.doi:10.4172/2155-9899.1000256
    16. ↵
      2. Yamagami S ,
      3. Dana MR
      . The critical role of lymph nodes in corneal alloimmunization and graft rejection. Invest Ophthalmol Vis Sci 2001;42:1293–8.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      2. Vail A ,
      3. Gore SM ,
      4. Bradley BA , et al
      . Conclusions of the corneal transplant follow up study. collaborating surgeons. Br J Ophthalmol 1997;8:631–6.
      OpenUrl
    18. ↵
      1. The Collaborative Corneal Transplantation Studies Research Group
      . The Collaborative corneal transplantation studies (CCTS). effectiveness of histocompatibility matching in high-risk corneal transplantation. Arch Ophthalmol 1992;110:1392–403.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      2. Tan DT ,
      3. Janardhanan P ,
      4. Zhou H , et al
      . Penetrating keratoplasty in Asian eyes: the Singapore corneal transplant study. Ophthalmology 2008;115:975–82.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      2. Price MO ,
      3. Thompson RW Jr ,
      4. Price FW Jr
      . Risk factors for various causes of failure in initial corneal grafts. Arch Ophthal 2003;121:1087–92.doi:10.1001/archopht.121.8.1087
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      2. Sellami D ,
      3. Abid S ,
      4. Bouaouaja G , et al
      . Epidemiology and risk factors for corneal graft rejection. Transplant Proc 2007;39:2609–11.doi:10.1016/j.transproceed.2007.08.020
      OpenUrlCrossRefPubMed
    22. ↵
      2. Hau S ,
      3. Clarke B ,
      4. Thaung C , et al
      . Longitudinal changes in corneal leucocyte density in vivo following transplantation. Br J Ophthalmol 2019;103:1035–41.doi:10.1136/bjophthalmol-2018-312386
      OpenUrlAbstract/FREE Full Text
    23. ↵
      2. Williams KA ,
      3. White MA ,
      4. Ash JK , et al
      . Leukocytes in the graft bed associated with corneal graft failure. Analysis by immunohistology and actuarial graft survival. Ophthalmology 1989;96:38–44.
      OpenUrlPubMedWeb of Science
    24. ↵
      2. Maguire MG ,
      3. Stark WJ ,
      4. Gottsch JD , et al
      . Risk factors for corneal graft failure and rejection in the Collaborative corneal transplantation studies. Collaborative corneal transplantation studies Research Group. Ophthalmology 1994;101:1536–47.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      2. Anijeet DR ,
      3. Zheng Y ,
      4. Tey A , et al
      . Imaging and evaluation of corneal vascularization using fluorescein and indocyanine green angiography. Invest. Ophthalmol. Vis. Sci. 2012;53:650–8.doi:10.1167/iovs.11-8014
      OpenUrlAbstract/FREE Full Text
    26. ↵
      2. Romano V ,
      3. Steger B ,
      4. Zheng Y , et al
      . Angiographic and in vivo confocal microscopic characterization of human corneal blood and presumed lymphatic neovascularization. Cornea 2015;34:1459–65.doi:10.1097/ICO.0000000000000609
      OpenUrl
    27. ↵
      2. Diamond MA ,
      3. Chan SWS ,
      4. Zhou X , et al
      . Lymphatic vessels identified in failed corneal transplants with neovascularisation. Br J Ophthalmol 2019;103:421–7.doi:10.1136/bjophthalmol-2018-312630 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30348644
      OpenUrlAbstract/FREE Full Text
    28. ↵
      2. Chatel M-A ,
      3. Larkin DFP
      . Sirolimus and mycophenolate as combination prophylaxis in corneal transplant recipients at high rejection risk. Am J Ophthalmol 2010;150:179–84.doi:10.1016/j.ajo.2010.03.010
      OpenUrlPubMed
    29. ↵
      2. Reinhard T ,
      3. Böhringer D ,
      4. Enczmann J , et al
      . Hla class I and II matching improves prognosis in penetrating Normal-Risk keratoplasty. Dev Ophthalmol 2003;36:42–9.
      OpenUrlPubMed
    30. ↵
      2. Cursiefen C ,
      3. Bock F ,
      4. Horn FK , et al
      . GS-101 antisense oligonucleotide eye drops inhibit corneal neovascularization: interim results of a randomized phase II trial. Ophthalmology 2009;116:1630–7.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      2. Spiteri N ,
      3. Romano V ,
      4. Zheng Y , et al
      . Corneal angiography for guiding and evaluating fine-needle diathermy treatment of corneal neovascularization. Ophthalmology 2015;122:1079–84.doi:10.1016/j.ophtha.2015.02.012
      OpenUrlCrossRefPubMed
    32. ↵
      2. Hou Y ,
      3. Le VNH ,
      4. Clahsen T , et al
      . Photodynamic therapy leads to time-dependent regression of pathologic corneal (lymph) angiogenesis and promotes high-risk corneal allograft survival. Invest. Ophthalmol. Vis. Sci. 2017;58:5862–9.doi:10.1167/iovs.17-22904
      OpenUrl
    33. ↵
      2. Bachmann BO et al
      . Promotion of graft survival by vascular endothelial growth factor a neutralization after high-risk corneal transplantation. Arch Ophthal 2008;126:71–7.doi:10.1001/archopht.126.1.71
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      2. Zhang H ,
      3. Grimaldo S ,
      4. Yuen D , et al
      . Combined blockade of VEGFR-3 and VLA-1 markedly promotes high-risk corneal transplant survival. Invest. Ophthalmol. Vis. Sci. 2011;52:6529–35.doi:10.1167/iovs.11-7454
      OpenUrlAbstract/FREE Full Text
    35. ↵
      2. Dohlman TH ,
      3. Omoto M ,
      4. Hua J , et al
      . VEGF-trap aflibercept significantly improves long-term graft survival in high-risk corneal transplantation. Transplantation 2015;99:678–86.doi:10.1097/TP.0000000000000512
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors Study concept and design: DS, CLH, DFPL. Acquisition, analysis or interpretation of data: DS, CLH, DFPL. Drafting of the manuscript: DS, CLH, DFPL. Critical revision of the manuscript for important intellectual content: SBT, SBK, DFPL. Statistical analysis: CLH. Study supervision: DFPL.

    • Funding This research was supported by the National Institute for Health Research (NIHR) Moorfields Biomedical Research Centre and NIHR Moorfields Clinical Research Facility, based at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology.

    • Disclaimer The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

    • Competing interests None declared.

    • Patient consent for publication Not required.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data availability statement Data are available upon request.