Refractive surgery in systemic and autoimmune disease.

Patients with underlying systemic disease represent challenging treatment dilemma to the refractive surgeon. The refractive error in this patient population is accompanied by a systemic disease that may have an ocular or even a corneal component. The literature is rather sparse about the use of laser refractive surgery (LRS) and such procedure is not approved by the United States Food and Drug Administration (FDA) in this patient population. Patients with collagen vascular disease, diabetes mellitus (DM), allergic and atopic disease, or human immunodeficiency virus (HIV) are never ideal for LRS. Patients with uncontrolled systemic disease or ocular involvement of the disease should not undergo LRS. However, a patient with well-controlled and mild disease, no ocular involvement, and not on multidrug regimen may be a suitable candidate if they meet stringent criteria. There is a need for a large, multicenter, controlled trial to address the safety and efficacy of LRS in patients with systemic disease before such technology can be widely adopted by the refractive surgery community.

INTRODUCTION

Laser refractive surgery (LRS) is one of the most common surgeries performed worldwide.1 Almost one million patients undergo corneal refractive surgery each year in the United States.2 LRS has proven to be safe with excellent outcomes and high patient satisfaction.3 As the successful outcomes of LRS are highlighted in the popular culture, patients, including those with underlying systemic diseases, are inquiring about the possibility of undergoing such procedures.

This specific patient population represents a challenging treatment dilemma to the refractive surgeon as the refractive error is accompanied by a systemic disease that may have an ocular or even a corneal component. Further, the underlying disorder may cause an intraoperative or postoperative complication that is not encountered in the normal patient. Adoption of LRS in patients with autoimmune disease has been slow and the literature about the outcomes of LRS in this specific patient population is rather scarce.

The United States Food and Drug Administration (FDA) established a list of ocular and systemic contraindications to photorefractive keratectomy (PRK) in the mid 1990's. This list included patients with collagen vascular, autoimmune or immunodeficiency diseases, pregnant or nursing women, as well as patients who are taking one or both of the following medications: Isotretinoin (Accutane®) or amiodarone hydrochloride (Cordarone®). Other conditions that are labeled by the FDA as warnings included diabetes, history of herpes simplex or herpes zoster keratitis, significant dry eye that is unresponsive to treatment, and severe allergies. Furthermore, conditions that were labeled by the FDA as precautions included patients under the age of 18 years, those who are taking sumatriptan (Imitrex®), those with ocular disease, corneal abnormality, previous corneal or intraocular surgery, trauma in the ablation zone, or a history of glaucoma.4

Similar FDA guidelines were subsequently adopted for laser in situ keratomileusis (LASIK). In 2002, the American Academy of Ophthalmology (AAO) guidelines stated that uncontrolled autoimmune or other immune-mediated disease is a contraindication to LRS.5

It is important to note that the recommendations of the FDA and the AAO are not based on clinical studies and outcomes, but rather on the known ocular complications of these disorders as well as the documented outcomes of non-laser ocular surgery in such patients. The justification for such recommendations and guidelines were the possibility that an elective surgical procedure may worsen the ocular disease and result in suboptimal outcomes or significant corneal complications including perforations or corneal melt.

In this article, we examine the outcomes of LRS in collagen vascular disease, diabetes mellitus (DM), allergic and atopic disease, HIV, pregnancy, as well as history of dermatologic keloid formation. It is worthy to note that most patients requesting LRS are young or middle-aged. However, corneal involvement in collagen vascular diseases (CVD) or DM occurs late in the course of these diseases, when the dry eye progresses. The late outcomes in these patients are not predictable.

COLLAGEN VASCULAR DISEASES

CVD is a group of systemic autoimmune diseases including rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), Sjögren's syndrome (SS), and the seronegative spondyloarthropathies. RA is the most common among the CVD with an annual incidence of 40/100,000.6 Keratoconjunctivitis sicca (KCS) is the most common ocular complication, where around 25% of patients with RA have KCS.7 CVD may have ocular or even corneal involvement and may lead to serious complications such as corneal melt and perforation.

Because of the potential ocular involvement in CVD and based on the initial reports of higher complication rates of cataract surgery in patients with CVD, the FDA listed CVD as contraindication in LRS.

Few small studies reported corneal ulceration and melting in CVD patients after cataract surgery,8 DSEK,9 PRK,10 and even trabeculectomy;11 but all of these cases had history or undiagnosed KCS or active systemic disease. However, multiple, larger studies looked at the safety of cataract surgery in CVD and none showed major visual complications.121314

Several retrospective studies have reported outcomes when LRS is performed in selected patients with controlled rheumatic diseases. Alio et al., reported the outcomes of LRS on 42 eyes with CVD (RA, SLE, dermatomyositis, scleroderma, ankylosing spondylitis, psoriatic arthritis, inflammatory bowel disease, arthritis, or Behçet's disease).15 The inclusion criteria included no active or history of ocular complications secondary to CVD, normal tear film functions, and no disease activity with or without concurrent medications for 6 months confirmed by a rheumatologist. Finally, patients needing multidrug regimens for disease control were excluded. Refractive outcomes at 6 months showed uncorrected visual acuity (UCVA) of 20/25 or better in more than 90% of this cohort. The spherical equivalent was −0.15 ± 0.38 diopters. Enhancement procedures were performed in 14.3%. Moderate degree of dry eye syndrome (DES) persisted in 9.5% of the eyes 6 months postoperatively. No corneal haze, melting, flap, or interface complications were observed in any of the study eyes.

Other retrospective studies showed similar results to Alio et al., using similar inclusion criteria.1617 Cobo-Soriano et al., studied LASIK outcome in 62 patients with connective tissue disorders, 91 patients with psoriasis, and 67 patients with intestinal inflammatory diseases. They reported good refractive outcome in these patients. No patients developed serious complications. Fifteen percent of patients developed superficial punctate keratopathy with subsequent resolution within 6 months.16 Another study reported good outcomes of LASIK in 49 eyes of patients with history of autoimmune diseases including SLE, RA, psoriatic arthritis, scleroderma, inflammatory bowel disease, Graves’ disease, Reiter's syndrome, and Hashimoto's disease. No eye developed corneal thinning, melting, persistent epithelial defect, persistent keratitis, scleral thinning, or scleritis.17 It is important to note that all patients included in both studies had a well-controlled autoimmune disease.

Also, Moshirfar et al., studied the effect of LASIK on patients with human leukocyte antigen (HLA)-B27 mutation and if LASIK increases the incidence of uveitis.18 They followed 46 eyes from patients with HLA-B27 mutation with no evidence of active uveitis for at least 1 year before the study. Twenty eyes received LASIK surgery after they were cleared for surgery by a rheumatologist, uveitis specialist, or both. Twenty-six eyes, representing the control population, did not undergo LRS. Fourteen of 20 eyes experienced a total of 20 episodes of postoperative uveitis in the LASIK group. During the same time period, 22 of the 26 control eyes experienced a total of 32 episodes. The incidence of uveitis between both groups was not statistically significant. The mean of when the uveitis episodes occurred was 13 months after surgery, and none occurred in the 1st month after surgery. This finding might indicate that surgery was not the precipitating factor in these patients’ uveitis flare ups. On the other hand, several other studies reported unfavorable outcome of LRS in patients with CVD. Li and Li reported corneal melt in 12 patients after LASIK; six of them had systemic diseases (one with thyroiditis, two with SLE, two with SS, and one with RA). The intervals between the corneal melt and LASIK ranged from 2 to 5 weeks. The medical status of these patients and whether the systemic disease was controlled or not was not mentioned.19 Lahners et al., reported two cases of peripheral keratitis 1 day after LASIK. One patient had history of RA and both showed evidence of previous inflammation.20 Liang et al., reported two patients with well-controlled SS that developed severe DES after LASIK.21 Diaz-Valle et al., reported a case of diffuse lamellar keratitis (DLK) associated with acute anterior uveitis 3 years after LASIK in a patient with a history of ankylosing spondylitis and recurrent acute anterior uveitis.22 Carp et al., reported a case of necrotizing keratitis 3 days after LASIK for myopia in patient with inflammatory bowel disease. The patient's systemic disease was severe enough to necessitate total colectomy for recurrent ulcerative colitis prior to the LRS.23 Liu et al., reported another case of ulcerative colitis that developed severe anterior uveitis with hypopyon 15 days after LASIK.24 Aman-Ullah et al., reported two cases of necrotizing keratitis after LRS in patients with controlled Crohn's disease. First case occurred 1 day after LASIK and the other case occurred 2 days following PRK.25 Finally, Javaloy et al., reported a case of recurrent DLK 4 months following LASIK and progressive deafness. The patient was later diagnosed with Cogan's syndrome.26

As can be seen, the literature addressing the outcome of LRS in CVD patients is rather scarce and inconclusive. Thus, there is a need for a large, multicentered, controlled trial to address this question. In the meantime, cautions should be exercised before performing LRS on CVD patients. If LRS is to be considered, the following guidelines need to be followed: Patients with moderate or severe disease as well as patients who need multidrug regimen to control the disease should be excluded; the systemic disease needs to be controlled with no evidence of activity or flare-up for a minimum of 6 months; no history of ocular involvement with normal eye exam including Schirmer test and tear-break-up time (TBUT); the patient has to be cleared for surgery by a rheumatologist, a uveitis specialist, or both; and finally at minimum the patient should have a documented negative workup for SS. Furthermore, the patient needs to understand that the long term outcome of LRS is unknown and that such a procedure is “off-label” in this patient population according to the FDA guidelines.

If a patient still wants to undergo the procedure, we favor LASIK over PRK. LASIK has a faster healing time in patients with potentially compromised corneal epithelial cells, has less risk of stromal haze and scar in patients with an underlying inflammatory disorder.

DIABETES MELLITUS

According to the American Diabetes Association, 25.8 million children and adults in the United States (8.3% of the population) have diabetes and 1.9 million new cases of diabetes were diagnosed in people aged 20 years and older in 2010.27

Ocular findings of DM include: Diabetic retinopathy, cranial nerve palsy, diabetic papillopathy, diabetic keratopathy (DK) and neuropathy, refractive instability, cataract, and increase risk of orbital infections. Diabetes is the leading cause of new cases of blindness among adults aged 20-74 years. In 2005-2008, 4.2 million (28.5%) of people with diabetes aged 40 years or older had diabetic retinopathy, and of these, almost 0.7 million had advanced diabetic retinopathy that could lead to severe vision loss.27 DK is common especially when other ocular or systemic complications are present.2829 Because of the changes in the epithelial basement membrane and the abnormality in the hemidesmosomes,303132 DK involves mainly the epithelium causing recurrent erosions and delayed epithelial healing.3334 Indeed, contact lenses might affect the cornea in diabetic more than nondiabetic patients as shown in one study where morphological endothelial changes were observed in diabetic patients compared with normal subjects.35 Corneal neuropathy usually occurs as part of the systemic diabetic neuropathy and can lead to poor epithelial adhesion and healing.36 Finally, abnormalities in the corneal endothelium of diabetic patients might lead to stromal edema.37

The FDA guidelines list DM as a relative contraindication to LRS. This is based on documented outcomes of non-laser ocular surgery in patients with diabetes and the potential risk of complications or negative outcome.38 A recent guideline by the AAO lists well-controlled diabetic disease as a relative contraindication and only uncontrolled disease as an absolute systemic contraindication.5

Poor wound healing has been identified in diabetic cornea. This may be due to delayed corneal wound healing. Indeed, DM patients have been reported to have persistent epithelial defect after pars plana vitrectomy.3940 They also have more inflammation after cataract surgery compared to normal subjects.41 Further, DM in general increases the risk of infection when the blood sugar is uncontrolled, but there is no clear evidence that DM increases the risk of ocular infection. However, the Endophthalmitis Vitrectomy Study showed that diabetic patients are more frequently associated with more virulent organisms and poor outcomes.4243

Few retrospective studies have studied the outcome of LRS in diabetic patients. Fraunfelder and Rich were the first to publish on this topic.38 They reported high complication rates (47%) in 30 eyes in diabetic patients who had LASIK compared to the control group (6.9%). The most frequent complications were punctate epithelial erosions (nine eyes) and persistent epithelial defects (nine eyes). Also, the refractive outcomes were worse in the diabetic eyes compared to the control group. There was no information about the glycemic control and if there were any systemic or ocular complications secondary to the DM. Similarly, few case reports showed worsening of diabetic retinopathy after LASIK in a patient who had uncontrolled blood sugar (hemoglobin A1c was 13.1%) and proliferative diabetic retinopathy before the surgery.444546 Furthermore, patients with type 1 diabetes may have higher risk of developing epithelial ingrowth after LASIK surgery.47

On the other hand, two more recent retrospective studies with larger sample size showed significantly less complication rates (6.5 and 11%) and very good refractive outcomes.1648 These later studies only included patients with good glycemic control and no evidence of systemic or corneal complications. In one of these studies, two patients had mild nonproliferative diabetic retinopathy and none of these patients showed worsening of their disease or other complications. Thus, it may seem, from the limited data that we have, that the risk of complications after LRS is related to poor glycemic control and LRS may be safe in patients with no systemic or ocular complications.

We do not perform LRS on any patient with uncontrolled DM or with diabetic ocular involvement. The DM has to be tightly controlled with hemoglobin A1c < 7.5 for a minimum of 6 months. Further, there should not be any evidence of systemic or ocular complications. We perform a thorough slit lamp exam to exclude tear film abnormality or insufficiency, subtle corneal epithelial defects, basement membrane irregularity, corneal esthesiometry to rule out neuropathy, and a careful dilated fundus exam to look for any evidence of diabetic retinopathy or papillopathy. We prefer LASIK to PRK because of the poor healing in diabetics. Since femtosecond laser has less risk of epithelial defects/sloughing compared to microkeratome, it may be safer to use femtosecond laser in diabetic patients.49

ALLERGIC AND ATOPIC DISEASES

Uncontrolled ocular allergy is considered a contraindication to LRS. DLK and DES have been reported after LASIK in patients with allergy.50 DES is common in atopic patients and DES is the most common complication post LASIK surgery.51 Moss et al., reported 25% increased risk of DES in patients with a history of allergy.52 Dry eye may result infrequently in impaired wound healing and decreased optical quality of the cornea.53 Boorstein et al., showed greater risk of DLK in untreated atopic patients compared to nonatopic patients or atopic patients taking an oral systemic nonsedating histamine receptor-1 antagonist.54 Myrowitz reported a case of DLK in a patient with systemic mastocytosis 5 days after uneventful LASIK surgery. The patient was successfully treated with topical corticosteroids.55

There are no mast cells in normal conjunctival epithelium, but in cases of chronic allergic conjunctivitis, mast cells have been isolated from the conjunctiva. There are few case reports of corneal complications after LRS in patients with limbal vernal keratoconjunctivitis (VKC) as well as contact dermatitis of the eyelids.5657 Also, there is a higher risk of corneal haze and myopic regression after PRK in patients with untreated allergic conjunctivitis compared to treated allergic conjunctivitis or to normal patients.58 On the other hand, Asano-Kato et al., showed no difference in adverse outcomes to LASIK in patients with a history of allergic conjunctivitis (84 eyes of 49 patients) compared to normal patients (140 eyes of 77 patients). Furthermore, there was no significant difference in UCVA between the two groups.59

We do not perform LRS on patients if there is uncontrolled or active allergic conjunctivitis, if there is evidence of DES, or in patients with chronic conjunctivitis and T-cell mediated reaction like VKC or atopic keratoconjunctivitis (AKC). Because of increased risk of DES in allergic patients, aggressive lubrication and punctal plugs are recommended. Patients with mild atopy should use systemic antihistamine medications preoperatively in an effort to decrease the incidence of DLK. However, antihistamine medications might worsen the DES and the patients need to be monitored closely.

HUMAN IMMUNODEFICIENCY VIRUS

Immunodeficiency is listed as a contraindication for LRS. However, a patient with HIV does not mean that their immunity is low. Acquired immune deficiency syndrome (AIDS) is defined as a CD4 lymphocyte count below 200 cells/μl (or a CD4+ T-cell percentage of total lymphocytes of less than 15%) and/or the presence of AIDS-defining opportunistic infection.60 The risk of HIV transmission with LRS is unknown. Hagen et al., studied the likelihood of virus transmission by using the excimer laser on a virus-infected tissue culture plate, while an uninfected tissue culture plate was in an inverted position over the infected plate. Six hundred pulses were applied and the experiment was repeated 20 times, all of the uninfected plates remained uninfected.61

Understandably, performing LRS on HIV or AIDS patients is controversial. There is severe lack of reporting on the outcome of LRS in this patient population. Hovanesian et al., reported a case of bilateral bacterial keratitis after LASIK in an HIV patient.62 A web-based survey was distributed among refractive surgeons to evaluate current practices of refractive surgeons in performing refractive surgery in persons with HIV or AIDS. Among the surgeons who responded to the survey, 50% considered persons with HIV acceptable candidates for LRS, while 12.5% considered persons with AIDS acceptable candidates. Of the surgeons who perform LRS in persons with HIV or AIDS, 72.7% do not take additional precautions when operating on these patients.63

If LRS is to be performed in a person with HIV, we recommend taking additional precautions during the procedure. These precautions should include performing one eye at a time, the surgeon and all the staff in the laser room should wear a filter mask, the laser plume should be evacuated immediately after the surgery, and the case should be done at the end of the day. HIV patients with evidence of DES or patients with AIDS should not have LRS. Further, we strongly encourage further investigation into this topic to better characterize the risks in these patients.

PREGNANCY

According to the FDA, LRS is contraindicated in pregnant or nursing women. Hormones produced during pregnancy and lactation can cause change in the refraction. The corneal curvature has been found to increase during the second and third trimester with resolution postpartum or after the cessation of breastfeeding.64 Sex hormone receptors have been isolated from human corneas.65 In an experimental study, the corneal biomechanical stability decreased when the cornea was exposed to estrogen. These data suggest that changes in estrogen levels might lead to corneal ectasia after LASIK.66 Furthermore, there are few case reports that showed increased risk of ectasia after pregnancy.676869 In our opinion, elective refractive procedures should be postponed until few months after delivery and cessation of breastfeeding and documented refractive stability.

DERMATOLOGIC KELOID

Keloid is a dermatological problem that occurs after trauma or injury leading to proliferation of dense fibrous tissue that extends beyond the boundaries of the original injury. Dermatologic keloids are more common in African Americans.70 Tanzer et al., showed good refractive outcome of PRK in six eyes in African American patients with history of keloid formation. No serious complication was reported and no difference in postoperative UCVA or best corrected visual acuity (BSCVA) was found between patients with dermatologic keloid and non-keloid formation.71 Artola et al., reported a retrospective study of five patients (ten eyes) with history of keloid formation.72 All patients were myopic and underwent LASIK. After 1 year, the mean UCVA was 20/20. None of the eyes required enhancement for residual error or regression up to 1 year postoperatively. No complication or corneal haze was reported. Also, Cobo-Soriano et al., showed no difference in the refractive outcome or complications following LASIK between patients with dermatologic keloids and those without.16 Finally, Lee et al., reported excellent outcome of Epi-LASIK in three patients with history of keloid formation.73

Keratinocytes and fibroblasts play a major role in keloid formation through increased expression of growth factors.74 As a result of corneal epithelial injury that occurs during LRS, cytokines and growth factors will be released in the tear film.75 Theoretically, this might increase the risk of subepithelial fibrosis and corneal haze.76 However, recent studies found that wound tension and stretching play a role in formation of keloid,74 and since the corneal tissue is not under surface tension this might prevent corneal haze in patients with dermatologic keloid.

CONCLUSIONS

As evident from our findings, there is paucity in the literature addressing the outcome of LRS in patients with systemic disease with often conflicting or confusing results. Further, the recommendations of the FDA and the AAO are not based on clinical studies and outcomes, but rather on the known ocular complications of these disorders as well as the documented outcomes of non-laser ocular surgery in such patients.

A patient with an underlying systemic disorder is never ideal for LRS. Within the last year, two young patients with underlying CVD underwent LASIK at an outside facility. One presented to The Wilmer Eye institute with hypopyon-associated uveitis and the second with peripheral corneal infiltrates. At our institute, we approach such patients with extreme care and caution. Patients with uncontrolled systemic disease or ocular involvement of the disease should not undergo LRS. However, a patient with well-controlled and mild disease, no ocular involvement, and not on multidrug regimen may be a suitable candidate if they meet our stringent criteria. We pay careful attention during the preoperative evaluation of each patient in order to avoid preventable complications and maximize patient outcomes. A detailed and documented discussion, including FDA and AAO guidelines, and informed consent takes place. Also, this group of patients is at risk to develop cataract at early age either because of the disease itself or as a side effect of the medications and this also has to be discussed with the patients as well. Each patient is addressed on a case-by-case basis. There is a need for a large, multicenter, controlled trial to address the safety and efficacy of LRS in patients with systemic disease before such technology can be widely adopted by the refractive surgery community.