Mustard gas [bis-(2-chloroethyl) sulfide] is a lipophilic

Size: px

Start display at page:

Download "Mustard gas [bis-(2-chloroethyl) sulfide] is a lipophilic"

Barbra Clark
6 years ago
Views:

1 CLINICAL SCIENCE Confocal Microscopy in Chronic and Delayed Mustard Gas Keratopathy Mohammad-Reza Jafarinasab, MD,* Siamak Zarei-Ghanavati, MD, Mojgan Rezaei Kanavi, MD, Farid Karimian, MD,* Mohammad-Reza Soroush, MD, and Mohammad-Ali Javadi, MD* Purpose: To evaluate in vivo confocal microscopic features of the cornea in chronic and delayed mustard gas keratopathy (MGK). Design: Comparative cross-sectional study. Participants and Controls: Twenty-two eyes of 22 consecutive patients with MGK and 28 eyes of 28 normal unoperated subjects were enrolled. Methods: All subjects underwent corneal confocal scanning, and the findings were compared between the 2 groups. Main Outcome Measures: Qualitative (layer thickness measurement and cell count) and quantitative (nerve and deposit evaluation) findings of corneal confocal scanning. Results: All subjects with MGK and normal subjects were males, with no significant difference in mean age between the 2 groups. Mean central corneal thickness, mean epithelial thickness, mean cell density of basal epithelia, keratocyte density at 3 stromal layers, and endothelial cell density in the MGK group were significantly lower than those in the control group. Loss of keratocytes was predominantly observed in the anterior to middle stroma. In vivo confocal microscopy revealed lack of a subbasal nerve plexus, presence of intrastromal hyperreflective microdots, prominent thickened midstromal nerves, enlarged bizarre-looking keratocytes, amyloid degeneration, lipid keratopathy, posterior stromal folds, and endothelial cell pleomorphism and polymegathism. Conclusions: Corneal thinning, significant loss of keratocytes together with pleomorphic residual keratocytes, thickened midstromal nerve, stromal microdots, amyloid degeneration, and lipid keratopathy were remarkable findings observed in our cases. Although all the corneal layers were affected significantly by mustard Received for publication April 24, 2009; revision received October 13, 2009; accepted November 1, From the *Ophthalmic Research Center and Department of Ophthalmology, Labbafinejad Medical Center, Shaheed Beheshti University (MC), Tehran, Iran; Eye Research Center, and Department of Ophthalmology, Mashhad University of Medical Sciences, Mashhad, Iran; Iran Eye Bank, Tehran, Iran; and Janbazan Medical and Engineering Research Center, Tehran, Iran. The authors have no financial or proprietary interest in this study. Reprints: Siamak Zarei-Ghanavati, MD, Eye Research Center, Mashhad University of Medical Sciences, Khatam-al-Anbia Eye Hospital, Abootaleb Boulevard, Mashhad, Khorasan Razavi 96869, Iran ( zareis@mums.ac.ir). Copyright Ó 2010 by Lippincott Williams & Wilkins gas, the anterior to middle parts of the cornea were more involved than the posterior parts. Key Words: confocal, mustard gas, mustard gas keratopathy, in vivo microscopy (Cornea 2010;29: ) Mustard gas [bis-(2-chloroethyl) sulfide] is a lipophilic alkylating chemical warfare agent that was first used in Ypres, Belgium, during World War I by the German army 1,2 and recently by Iraqi forces against Iranian veterans during the Iraq Iran War ( ). 3 The eyes are the most sensitive organs to mustard gas. Ocular involvement, which is divided into acute, chronic, and delayed categories, has been reported to occur in 75% 90% of individuals exposed to mustard gas. Most patients with acute mustard gas related injuries recover completely after a few weeks. 4,5 Chronic and delayed mustard gas keratopathy (MGK) has been reported to occur in 0.5% 1% of a severely exposed group; it is usually progressive and difficult to treat. 6,7 The reported histopathologic features of corneal buttons from cases with MGK include destruction of epithelium and Bowman layer, stromal neovascularization, focal corneal thinning and ulceration, lipid and amyloid deposition, and scarring in the corneal stroma together with loss of keratocytes and endothelial cells However, in vivo studies are the only means for evaluating cellular and intercellular changes in mustard gas exposed victims who do not need invasive procedures such as keratoplasty. Confocal microscopy is a relatively new technique that enables real-time, in vivo, high-resolution imaging of the cornea and that, as a noninvasive method, evaluates microstructural changes in corneal layers. The purpose of this study was to evaluate confocal microscopic features of chronic and delayed MGK compared with those of normal unoperated controls. PATIENTS AND METHODS To conduct this study, we obtained full ethical approval from the Ophthalmic Research Center of Shaheed Beheshti Medical University and the Janbazan Medical and Engineering Research Ethics Committee. The study followed the tenets of the Declaration of Helsinki, and informed consent was obtained from all subjects. Patients with clinically diagnosed MGK and healthy volunteers as the control group were Cornea Volume 29, Number 8, August

2 Jafarinasab et al Cornea Volume 29, Number 8, August 2010 enrolled in the study. A complete medical history was taken, and full ocular examination was performed in all subjects. Eyes with previous ocular surgery were excluded. All patients in the MGK group had official documents for exposure to mustard gas during the Iraq Iran War on file with the Veteran Foundation (Bonyad Janbazan) and adequate evidence of mustard gas related keratopathy on slit-lamp examinations, including conjunctival and limbal vascular tortuosity and telangiectasis, areas of limbal ischemia, peripheral corneal neovascularization, corneal scarring, stromal thinning, and intrastromal hyaline- or lipid-like deposits. 11,12 The control group was selected from age-matched male volunteers with no history of ocular or systemic disease, previous ocular trauma and/or operation, and contact lens wearing. Their corneal slitlamp biomicroscopic examination and topography were within normal limit. Confocal Microscopic Examination Confocal scanning was performed in all individuals by an ophthalmopathologist (M.R.K.). After instillation of a topical tetracaine eyedrop 0.5% (Sina Darou Laboratories Company, Tehran, Iran) to both eyes, a confocal scan (Confoscan 3.4; Nidek Technology, Padova, Italy) was performed on the cornea with the use of a viscogel (methyl cellulose 2.5%) as a coupling agent between the tip of the front lens (a objective lens) and the cornea. The automated and manual modes were used to capture images of all corneal layers in the central to midperipheral areas, and the scan was repeated to ensure obtainment of complete Z-scan graphs from the full thickness of the cornea. The images were recorded by a digital video camera. The coronal section of each image was 455 mm (horizontal) mm (vertical) with a lateral resolution of 1 mm and a depth of field of 10 mm. Each image was separated from the adjacent one by an average of 1.5 mm. First, we excluded patients with very severe MGK in whom the corneal opacity did not allow detailed confocal microscopy. Next, with the use of a Z-scan graph of the recorded images, the epithelial region including the basal epithelium, the subbasal nerve plexus, the anterior 10%, the middle and the posterior 10% of the stroma, the midstromal nerves, and the endothelium were determined. Then, the examiner (M.R.K.) selected 2 images with no motion artifacts from each layer. Density of the basal epithelial cells and endothelial cells was determined in the selected images by using the manual and automatic count systems, respectively, and maximum thickness of the midstromal nerve was measured by a distance calculating method. In addition to the endothelial cell density, the coefficient of variation (an index for polymegathism) and percentages of hexagonal cells were also determined. In the selected images from different stromal layers, the bright keratocyte nuclei were counted in a predefined area ( mm 2 ) with a manual counter to obtain the keratocyte density (cells/mm 2 ). With the use of Z-scan graphs, the central corneal thickness was measured as the distance between the endothelial peak and the last image of the epithelium. Epithelial thickness was determined as the distance between the beginning and the ending of the epithelial peak. Stromal thickness was calculated from a point just after the endothelial peak to a point just before the beginning of the epithelial peak. Depth of the involved stroma with bizarrelooking keratocytes and abnormal hyperreflectivity were also measured in the cases with MGK. Then, the averaged results for each parameter in the MGK group were compared with those of the normal unoperated group by using the statistical independent t test; P, 0.05 was considered significant. Furthermore, the appearance of the basal epithelial cells, subbasal nerve plexus, and stromal keratocytes and the presence of any abnormal intrastromal deposition were described in the MGK corneas. RESULTS Twenty-two eyes of 22 patients with MGK with a mean age of years and 28 normal unoperated eyes of 28 healthy volunteers with a mean age of years were entered in the study. There was no significant difference in age between the 2 groups, and all individuals were men. The patients were studied years after initial exposure to the mustard gas. Clinical presentations included conjunctival and limbal vascular abnormality and peripheral corneal neovascularization (in all cases), peripheral and central corneal deposits (20 cases, 90.9%), superficial corneal scarring (10 cases, 45.45%), and epithelial irregularity (8 cases, 36.36%). Quantitative Evaluation Central corneal thickness was significantly lower in the MGK group than that in the normal unoperated group ( vs mm; P, ). Mean epithelial thickness was and mm in the MGK and control groups, respectively (P, ). There was also a statistically significant difference in stromal thickness between the MGK and normal unoperated groups ( vs mm; P, ). Mean thickness of the involved stroma in the MGK group was mm (range: mm). Cell density values are demonstrated in Table 1. Mean basal epithelial cell density was observed to decline significantly in the MGK group ( cells/mm 2 ) compared with that in the normal unoperated group ( cells/mm 2 )(P, ). In the cases with MGK, keratocyte density was considerably lower than that in the normal unoperated cases at the 3 stromal layers [anterior stroma: vs cells/mm 2 (percentage of reduction: 95.3%; P, ); middle stroma: vs cells/mm 2 (percentage of reduction: 49.2%; P, ); and posterior stroma: vs cells/mm 2 (percentage of reduction: 11.0%; P = 0.004)]. The thickness of midstromal nerves in the MGK group ( mm) was statistically higher than that in controls ( mm) (P, ). Confocal images of basal cells, stromal keratocytes, midstromal nerves, and endothelial cells in the 2 groups are illustrated in Figure 1. Compared with values in the normal unoperated group, the endothelial cell density was statistically lower ( vs cells/mm 2 ; P = 0.024) and the q 2010 Lippincott Williams & Wilkins

3 Cornea Volume 29, Number 8, August 2010 Confocal Microscopy in Chronic and Delayed MGK TABLE 1. Mean Cell Density and Percentage of Density Reduction in the Different Layers of Cornea in the Study Groups Cell Density MGK Group Normal Unoperated Group Percentage of Density Reduction Mean basal epithelial cell density (cells/mm 2 ) Mean keratocyte density in the anterior stroma (cells/mm 2 ) Mean keratocyte density in the middle stroma (cells/mm 2 ) Mean keratocyte density in the posterior stroma (cells/mm 2 ) Mean endothelial cell density (cells/mm 2 ) P endothelial cell area ( vs mm 2, P, ) and coefficient of variation were higher in the MGK group (32.66% vs %; P = 0.027). The percentage of hexagonal cells in cases with MGK was statistically lower than that in controls (55.66% vs %; P = 0.043). Qualitative Evaluation On confocal scan of the cases with MGK, the regular mosaic appearance of the basal epithelia was not observed and the cells looked pleomorphic, with some having ill-defined borders and some having nonspecific hyperreflective borders (Figs. 1A1, 2). The subbasal nerve plexus was apparent in only 2 cases with MGK. In all cases with MGK, bizarre and markedly enlarged keratocytes and hyperreflective microdots of 1 2 mm in diameter were noted in the stroma (Fig. 2A). Focal areas of nonspecific interstitial hyperreflectivity and disarrangement, representing stromal haziness and scarring, were found predominantly in the anterior stroma. Also included were posterior stromal folds in 14 cases (63.66%) with MGK (Fig. 1D2). Lipid keratopathy as a consequence of previous stromal vascularization was observed as bright needle-like crystalline deposits (Fig. 2B) in the stroma in 10 cases (45.45%). Amyloid deposits appeared as hyperreflective cotton candylike deposits (Fig. 2C) in 18 cases (81.81%), 3 of which needed keratoplasty, and the amyloid deposits were confirmed consequently by histopathology as red apple-green birefringent deposits under polarized light on Congo red staining. DISCUSSION Mustard gas toxicity in the cornea is thought to begin from accumulation of this lipophilic substance in the lipid layer of the tear film. The gas acts as an alkylating agent that induces structural changes, impairs normal cellular homeostasis, and eventually causes cell death. 1,2,7 Although deep corneal penetration of this chemical agent seems to be difficult, confocal microscopy in the current study revealed involvement of all the corneal layers with involvement shown predominantly in the anterior and middle layers in patients with chronic and delayed MGK. However, readers should be aware that this cross-sectional study explains the current corneal findings that have occurred an average of 24 years after mustard gas exposure. Therefore, the findings represent just a snapshot of the long and complicated process and cannot prove any cause effect relationships. Javadi et al, 10 on histopathologic examination of 22 corneal buttons with chronic and delayed MGK, reported corneal thinning and ulceration, loss of keratocytes and endothelial cells, focal inflammation, stromal vascularization, and degenerative sequels of long-standing inflammation. In the current series, we observed similar features on confocal microscopy, although focal areas of inflammatory cell infiltrates were not apparent. It seems that inflammatory manifestation of MGK is an infrequent condition, as reported previously. 9 Foci of amyloid degeneration and lipid keratopathy, as consequences of a long-standing degenerative process and stromal neovascularization, were present in 81.81% and 45.45% of cases with MGK, respectively. Stromal neovascularization is a part of a healing process; therefore, it seems that degenerative processes dominate in chronic and delayed MGK. Previous histopathologic studies have shown epithelial injuries induced by mustard gas. 10,13 In our study, the mosaic pattern of basal epithelial cells was not apparent on confocal scan and most of the cells looked pleomorphic with damaged or irregular high-contrast boundaries. These morphologic changes in addition to decreased basal cell density and epithelial thickness shown in the current study may be related to epithelial stem cell deficiency documented in MGK. 7 Lack of the subbasal nerve plexus was another confocal microscopic feature in most of the mustard gas exposed corneas, which could be secondary to destruction of the Bowman layer and occurrence of subepithelial fibrosis. However, the effects of dry eye on subbasal nerve density cannot be excluded. Most of the cases with MGK were reported to suffer from dry eye, 9,12 and a series of studies on the effects of dry eye on subbasal nerve density yielded different results Using confocal microscopy in dry eyes, Erdelyi et al 14 reported significant reduction in subbasal nerve density. We are not sure whether significant loss of the subbasal nerve plexus in cases with MGK is a direct MGK-related neuropathy or is related to mustard gas induced dry eye. However, loss of this nerve plexus may act as a predisposing factor for a severe ocular surface disorder. Interestingly, confocal microscopy of our cases with MGK showed increased thickness of midstromal nerves. Severe loss of keratocytes has been reported on histopathologic examination of a series of corneal buttons with MGK after corneal transplantation. 10 Keratocyte density decreased progressively from the anterior to the posterior stroma in normal cornea. Moller-Pedersen et al 17 demonstrated a 30% decrease in cell density over the entire anterior posterior stromal thickness and a gradient pattern of keratocyte distribution. This gradient was reversed in our cases with MGK, with severe and even total loss of keratocytes in the anterior stroma. To the best of our q 2010 Lippincott Williams & Wilkins 891

4 Jafarinasab et al Cornea Volume 29, Number 8, August 2010 FIGURE 1. Confocal scan in healthy cornea (A1, B1, C1, D1, E1) and MGK cornea (A2, B2, C2, D2, E2). Note the basal cell layer in a healthy cornea (A1) with a regular mosaic pattern compared with pleomorphic basal cells with illdefined borders and some with high reflective boundaries (arrow, A2). Note the quiescent keratocytes in healthy corneal stroma; the cell density decreases progressively from the anterior (B1) to middle (C1) to posterior stroma (D1). Complete loss of keratocytes in the most anterior layer of stroma of a patient with MGK (B2); unlike a healthy cornea, keratocyte density increases progressively from the anterior to the posterior stroma (B2, C2, D2). Note the pleomorphic keratocytes (C2, curved arrow) and presence of posterior stromal folds (arrowheads, D2). The thickness of the midstromal nerve in MGK (E2) was higher than that in the normal unoperated cornea (E1) q 2010 Lippincott Williams & Wilkins

5 Cornea Volume 29, Number 8, August 2010 Confocal Microscopy in Chronic and Delayed MGK FIGURE 2. A, Stromal hyperreflective microdots. B, Lipid keratopathy as bright needle-like crystalline deposits in the stroma. C, Amyloid deposits as cotton candy-like hyperreflective structures. knowledge, such a severe reduction in keratocyte density has not been reported in the literature, although milder reduction was observed in persons with a history of photorefractive keratectomy or laser in situ keratomileusis surgery. 18 Although a possible relationship between decreased keratocyte density and corneal nerve changes might be considered, this hypothesis will require further study. In addition to keratocyte loss, we observed bizarre and abnormally enlarged pleomorphic keratocytes in all cases with MGK, a feature that was not detected in the normal unoperated cases. This could be explained by the teratogenic and mutagenic activities of the mustard gas, 19,20 although to the best of our knowledge, the association between exposure to this chemical agent and ocular surface malignancy has not been reported. Considering the ability of normal keratocytes to divide and migrate after injury, 21 we suppose that the genetically altered keratocytes may lose their ability to repopulate the injured stroma in MGK. Furthermore, this cytotoxic agent may trigger the apoptotic pathway 22 and lead to chronic progressive keratocyte death. Presence of necrotic tissue within the stroma in addition to lipid and amyloid deposition was another histopathologic feature in a mustard gas injured cornea. 10,13 In our study, the presence of intrastromal microdots might be related to foci of stromal necrosis. These changes in the interstitial matrix of the cornea might be a potential barrier for repopulation of stromal keratocytes. In our series, the endothelial layer seemed to be less affected by mustard gas. Although endothelial changes in the MGK group were statistically significant, there was still adequate endothelial cell density to perform lamellar keratoplasty or other intraocular surgery in such cases. However, readers should be aware of selection bias in this study because we included patients with less severe MGK in whom the corneal clarity allowed confocal microscopy. In conclusion, this study demonstrated in vivo microstructural abnormalities in different corneal layers in chronic and delayed MGK. Although abnormal features were observed in all corneal layers, the anterior to middle areas were affected more severely than the posterior parts. REFERENCES 1. Blodi FC. Mustard gas keratopathy. Int Ophthalmol Clin. 1971;11: Duke-Elder S, MacFaul PA. Injuries. Part 2. Non-mechanical injuries. In: Duke-Elder S, ed. System of Ophthalmology. Vol. 14. London, United Kingdom: Henry Kimpton; 1972: United Nations Security Council. Report of the Mission Dispatched by the Secretary General to Investigate Allegations of the Use of Chemical Weapons in the Conflict Between the Islamic Republic of Iran and Iraq. New York, NY: United Nations; April 25, S/19823 and S/ 19823/Addendum Mandel M, Gibson WS. Clinical manifestations and treatment of gas poisoning. J Am Med Assoc. 1917;69: Gates M, Moore S. Mustard gas and other sulphur mustards. In: Division 9, National Defense Research Committee, eds. Chemical Warfare Agents, and Related Chemical Problems. Washington, DC: Office of Scientific Research and Development; 1946: Safarinejad MR, Moosavi SA, Montazeri B. Ocular injuries caused by mustard gas: diagnosis, treatment, and medical defense. Mil Med. 2001; 166: Pfister R. Ocular effects of mustard agents and lewisite. In: Pechura CM, Rall DP, eds. Veterans at Risk: the Health Effect of Mustard Gas and Lewisite. Washington, DC: Academic Press; 1993: Kadar T, Turetz J, Fishbine E, et al. Characterization of acute and delayed ocular lesions induced by sulfur mustard in rabbits. Curr Eye Res. 2001; 22: Javadi MA, Yazdani S, Sajjad Hi, et al. Chronic and delayed-onset mustard gas keratitis. Ophthalmology. 2005;112: Javadi MA, Yazdani S, Kanavi MR, et al. Long-term outcomes of penetrating keratoplasty in chronic and delayed mustard gas keratitis. Cornea. 2007;26: Solberg Y, Alcalay M, Belkin M. Ocular injury by mustard gas. Surv Ophthalmol. 1997;41: Etezad-Razavi M, Mahmoudi M, Hefazi M, et al. Delayed ocular complications of mustard gas poisoning and the relationship with respiratory and cutaneous complications. Clin Experiment Ophthalmol. 2006;34: Pleyer U, Sherif Z, Baatz H, et al. Delayed mustard gas keratopathy: clinical findings and confocal microscopy. Am J Ophthalmol. 1999;128: Erdelyi B, Kraak R, Zhivov A, et al. In vivo confocal laser scanning microscopy of the cornea in dry eye. Graefes Arch Clin Exp Ophthalmol. 2007;245: q 2010 Lippincott Williams & Wilkins 893

6 Jafarinasab et al Cornea Volume 29, Number 8, August Hosal BM, Ornek N, Zilelioglu G, et al. Morphology of corneal nerves and corneal sensation in dry eye: a preliminary study. Eye. 2005;19: Zhang M, Chen J, Luo L, et al. Altered corneal nerves in aqueous tear deficiency viewed by in vivo confocal microscopy. Cornea. 2005;24: Moller-Pedersen T, Ledet T, Ehlers N. The keratocyte density of human donor corneas. Curr Eye Res. 1994;13: Erie JC, McLaren JW, Hodge DO, et al. Long-term corneal keratoctye deficits after photorefractive keratectomy and laser in situ keratomileusis. Trans Am Ophthalmol Soc. 2005;103: Takeshima Y, Inai K, Bennett WP, et al. p53 mutations in lung cancers from Japanese mustard gas workers. Carcinogenesis. 1994;15: Dacre JC, Goldman M. Toxicology and pharmacology of the chemical warfare agent sulfur mustard. Pharmacol Rev. 1996;48: Wilson SE, Mohan RR, Ambrósio R Jr, et al. The corneal wound healing response: cytokine-mediated interaction of the epithelium, stroma, and inflammatory cells. Prog Retin Eye Res. 2001;20: Helena MC, Baerveldt F, Kim WJ, et al. Keratocyte apoptosis after corneal surgery. Invest Ophthalmol Vis Sci. 1998;39: q 2010 Lippincott Williams & Wilkins

Mustard Gas Induced Ocular Surface Disorders

Challenging Case Mustard Gas Induced Ocular Surface Disorders Section Editor: Alireza Baradaran-Rafii, MD Ophthalmic Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran Sulfur