Abstract
Purpose:
Investigating impression cytology (IC) results of various types of clinically suspected ocular surface lesions over a 14-year period in a referral center in Iran.
Methods:
IC findings obtained from patients with different types of ocular surface disorders between 2005 and 2018 were reviewed. Agreement between clinical suspicions and IC results was evaluated by calculating Cohen’s Kappa coefficient (CKC).
Results:
Clinical suspicions in 688 surveyed eyes were ocular surface squamous neoplasia (OSSN, 42.0%), limbal stem cell deficiency (LSCD, 36.3%), dry eye-related disorders (DERD, 11.5%), Acanthamoeba keratitis (AK, 7.2%), benign pigmented lesions (BPL, 1.9%), immune-related conjunctivitis (IRC, 0.7%), and malignant pigmented lesions (MPL, 0.4%). General agreement between clinical suspicions and IC results was 0.68 for all groups. This agreement was almost perfect in AK (CKC = 0.966) and BPLs (CKC = 0.843), and was substantial in MPLs (CKC = 0.749), OSSNs (CKC = 0.684), and LSCD (CKC = 0.612). CKC in IRC (0.567) and DERDs (0.443) was moderate. Histopathologic results were available in 22 eyes and were well-correlated with corresponding IC results (CKC = 0.86). Multiple post-treatment follow-up sessions of IC were performed in 51 eyes (11.4%) that had diagnosis of LSCD (31), OSSN (17), and MPL (3) at the first IC session.
Conclusion:
Our survey not only demonstrated an overall substantial agreement between IC results and primary clinical suspicions, but also showed an almost perfect correlation between IC results and existent histopathologic data. Therefore, IC as a non-invasive diagnostic modality can be of great importance in proper diagnosis of various ocular surface diseases especially when distinguishing malignant from benign lesions is required.
Introduction
Impression cytology (IC) is a simple, fairly, rapid, and minimally invasive method for collecting superficial layers of the corneal and conjunctival epithelium, and in addition to clinical signs and symptoms, it can provide a cytological proof for the diagnosis of a wide variety of ocular surface disorders.1,2 This outpatient-based procedure was reported to have a well correlation with histopathological diagnoses, and is also the gold standard for confirmation of the conjunctivalization of the cornea and limbal stem cell deficiency (LSCD). 3 By using IC, many ocular complications such as scarring, lid deformity, and LSCD induced by repeated surgical biopsies can be prevented;1,4 and therefore, IC as a reliable diagnostic tool, has been proved to substitute surgical biopsy for ocular surface lesions. 1 It also plays an essential role in diagnosis and management of patients with ocular surface lesions by discriminating malignant from benign lesions. 5 Nonetheless, this technique cannot replace histopathology in cases with surface keratosis in which surface keratin may interfere with proper sampling of the superficial cells. Hence, in cases with negative IC results, a tissue biopsy may be necessitated. 1
IC was established in the eye pathology laboratory of the Central Eye Bank of Iran in 2005; and since then, it was implemented for the diagnosis and follow-up of large numbers of patients with ocular surface disorders. This study was conducted to assess the agreement between the results of IC and the clinical suspicions in patients with various types of ocular surface lesions who were referred to the eye pathology laboratory of the Central Eye Bank of Iran over a 14-year period.
Materials and methods
Patients
In a retrospective case series, all IC and medical records obtained from patients with various types of ocular surface disorders who were referred to the eye pathology laboratory of the Central Eye Bank of Iran between 2005 and 2018 were reviewed. Full ethical approval was obtained from the Ethics Committee of the Ophthalmic Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran. The demographic data, clinical suspicions, and the results of the IC were recorded. Moreover, the clinical and IC diagnoses were categorized into seven groups as follows: (i) dry eye-related diseases (DERD); (ii) limbal stem cell deficiency (LSCD) in forms of chemical burns, thermal burns, mustard gas keratopathy, and aniridia; (iii) ocular surface squamous neoplasia (OSSN); (iv) benign pigmented lesions (BPL) such as nevi; (v) malignant pigmented lesions (MPL) such as malignant melanoma and primary acquired melanosis with atypia; (vi) Acanthamoeba keratitis (AK); and (vii) immune-related conjunctivitis (IRC) such as allergic keratoconjunctivitis.
IC technique
IC technique was fully described earlier. 5 Briefly, sampling for IC was performed under topical anesthesia with 0.5% tetracaine following slit-lamp examination of the lesion. Precut cellulose acetate filter strips (47 mm, pore size 0.47 µm, Schleicher & Schuell Microscience GMBH, Dassel, Germany) were applied to the site of interest using flat and round-tipped forceps during the recording of the direction of the oblique side. After a few-second gentle pressure of the strip onto the ocular surface, it was peeled off the lesion surface and fixed in cytology fixative. The sampling procedure was performed by two ocular pathologists (MRK and SBH) who were also ophthalmologists. The specimens were stained with a combination of Gill’s modified Papanicolaou stain and periodic acid-Schiff (PAS) and the stained specimens were mounted on glass slides using a DPX mountant. The slides were evaluated under light microscopy (BX41, Olympus, Japan) and each sample was reviewed by two ocular pathologists (MRK and SBH) for the final cytological diagnosis.
Clinical and cytopathological criteria for the diagnosis of DERD
Patients who had clinically two of three characteristics of DERD, including subjective symptoms in forms of ocular discomforts and visual disturbances, abnormal tear function tests in forms of <10 mm moisture of filter paper in 5 min on Schirmer’s test or tear break-up time of <10 seconds, and the presence of epithelial damage evidenced by vital staining tests, were suspected to DERD. 6 Diagnosis of DERD on IC samples was made based on the observation of epithelial squamous metaplasia in forms of goblet cells loss in conjunctival samples, reduced nuclear cytoplasmic ratio, nuclear pyknosis (Figure 1(a)) with cellular snakes, presence of variable degrees of inflammatory cells, and lack of cellular atypia.7–9
Representative impression cytology images of the study groups. (a) The image depicts the presence of cells with pyknotic nuclei and containing densely packed keratin filaments in a case with dry eye-related disease. (b) Note the presence of PAS-positive goblet cells among corneal epithelial cells in a case of limbal stem cell deficiency. (c) The image depicts the presence of epithelial cells with significant nuclear pleomorphism in a case with ocular surface squamous neoplasia. (d) Note the presence of nests of melanocytes containing cytoplasmic pigment and bland-looking nuclei in a case with benign melanocytic nevus. (e) The image represents the abundant presence of the atypical-looking fairly pigmented melanocytes in a case with conjunctival melanoma. (f) and (g) Note the presence of trophozoite (f, arrow) and cyst (g, arrow) in cases with Acanthamoeba keratitis. (h) The image depicts the presence of abundant inflammatory cells among the epithelial cells in a case with atopic keratoconjunctivitis.
Clinical and cytopathological criteria for the diagnosis of LSCD
Patients presenting with chronic inflammation, conjunctivalization, corneal neovascularization, epithelial haze, or persistent or recurrent epithelial defects, and had clinical history of thermal and chemical burns, mustard gas keratopathy, ocular cicatricial abnormalities or aniridia, were clinically suspected to LSCD. 10 Observation of PAS-positive goblet cells among the epithelial sheets on the corneal side was diagnostic of LSCD (Figure 1(b)) and corneal conjunctivalization.11,12
Clinical and cytopathological criteria for the diagnosis of OSSN
OSSN was clinically suspected when there was a conjunctival/corneal epithelial overgrowth of telangiectatic feeder vessels with or without surface keratinization. 13 The diagnosis of OSSN on IC was made based on observation of nuclear enlargement, irregular nuclear contour, coarsely clumped chromatin, nuclear pleomorphism, binucleation or multinucleation, and evident nucleoli (Figure 1(c)).4,14,15
Clinical and cytopathological criteria for the diagnosis of BPL
Patients with a long history of well-circumscribed pigmented lesions in interpalpebral conjunctival near the limbus with or without cysts, and darkly pigmented patients with bilateral perilimbal flat pigmentations were clinically suspected to BPL.16,17 The IC criteria for BPL such as nevi and primary acquired melanosis without atypia were the presence of nests or clusters of melanocytes containing cytoplasmic pigment and bland-looking nuclei, and the absence of mitosis (Figure 1(d)).5,18,19
Clinical and cytopathological criteria for the diagnosis of MPL
Patients with a short-time history of unilateral and flat conjunctival epithelial pigmented lesions that were not well-circumscribed, and those with pigmented conjunctival mass arising from previous nevus or pigmented lesion especially in caruncle, tarsal plate or forniceal conjunctiva, were clinically suspected to BPL. 20 The IC criteria for diagnosis of MPL were the presence of clusters of pleomorphic atypical cells, not resembling epithelial cells, of different sizes, with or without cytoplasmic pigment, irregular nuclear chromatin pattern, anisokaryosis characterized by large and irregular nuclei with prominent nucleoli, and mitoses. PAM with atypia was diagnosed on IC when the relative proportion of atypical melanocytes was low. However, malignant melanoma was cytologically diagnosed when were the atypical melanocytes abundant (Figure 1(e)).5,18,19
Clinical and cytopathological features for diagnosis of AK
Patients with history of contact lens wearing or eye contact with dirty water presenting with corneal epitheliopathy, radial keratoneuritis, or stromal infiltration were suspected to AK. 21 In IC samples, AK was diagnosed based on the observation of PAS-reactive Acanthamoeba cysts and/or hyperchromatic pear-shaped trophozoites (Figure 1(f) and (g)).7,21,22
Clinical and cytopathological criteria for the diagnosis of immune-related conjunctivitis
Patients with history of ocular allergy and presenting with bilateral irritating eyes, redness, and corneal pannus were suspected to IRD. 23 IRC was cytologically diagnosed based on the observation of variable degrees of inflammatory cells with or without squamous cell metaplasia (Figure 1(h)) and lack of cellular atypia.7,24,25
Histopathological examinations
To confirm the IC diagnoses, histopathological results of cases with biopsy specimens were reviewed, and their agreement with the corresponding IC results was investigated.
Statistical analysis
The data were presented as mean, standard deviation, median, range, frequency, and percentage. The Cohen’s Kappa coefficient (CKC) was calculated to analyze: (a) the agreement between the rate of clinical suspicion in each disease category and the rate of IC positive result for the corresponding category; (b) the agreement between the existent histopathological diagnoses and their corresponding IC results; and (c) the overall agreement between the clinical suspicions and the positive IC results. The CKC results were interpreted as slight agreement (0.01–0.20), fair agreement (0.21–0.40), moderate agreement (0.41–0.60), substantial agreement (0.61–0.80), and almost perfect or perfect agreement (0.81–1.00). 26 All statistical analyses were performed by applying SPSS (IBM Corp., Released 2013, IBM SPSS Statistics for Windows, Version 22.0 Armonk, NY, IBM Corp.). Also, the data for multiple (more than once) post-treatment follow-up sessions of IC were investigated.
Results
Between 2005 and 2018, IC records of 688 eyes from 577 Caucasian patients (68% male) with the mean age of 47 years old (age range between 7 and 90 years old) were retrospectively reviewed and their agreement with corresponding clinical diagnoses was analyzed. Bilateral involvement was observed in 133 patients and 51 eyes (11.4% of the 448 IC positive eyes) out of 36 patients who had already been listed for the first session of IC and had positive IC results, underwent multiple (more than once) post-treatment follow-up sessions of IC for investigating residual or recurrent lesions. The overall agreement between the clinical suspicions and the IC results was 0.68. Slight agreement or fair agreement was observed in none of the disease categories. Table 1 demonstrates the rates of agreement between the clinical suspicions and the IC results.
Clinical suspicions, IC results, and the Cohen’s Kappa agreement between the clinical suspicions and the corresponding IC results are shown. The eyes with multiple post-treatment follow-up sessions of IC were also specified.
The histopathologic results were available for 22 eyes including OSSN (13 eyes), benign melanocytic nevi (four eyes), conjunctival melanoma in the context of PAM with atypia (three eyes), and pterygium (two eyes). The corresponding IC results were all well supported by the histopathological features with the CKC of 0.86.
LSCD
About 250 cases were clinically suspected to LSCD including 206 chemical/thermal injuries, 38 mustard gas injuries, two cases of aniridia, and one case of combined prior ocular surface surgery and topical chemotherapy. Out of 206 LSCD cases, 117 (57%) had IC results indicative of LSCD and the Cohen’s Kappa agreement between the clinical diagnosis and the IC results in this group was 0.612. Multiple post-treatment follow-up sessions of IC were performed in 31 (26.5%) of the 117 eyes that were diagnosed as LSCD at the first IC session, in which persistent LSCD was disclosed in 26 cases (83.8%).
OSSN
Clinical suspicion to OSSN was observed in 289 cases, out of which, 191 (66%) cases were cytologically proved to be OSSN with the Cohen’s Kappa agreement of 0.684. The IC results among the remained 98 cases were squamous metaplasia in 80, LSCD in 11, non-specific inflammation in 5, and benign melanocytic lesion in 2. In other words, the IC results ruled out the presence of atypia in 34% of clinically suspected OSSNs. In cases with no atypia on IC, the corresponding ophthalmologists were advised to consider repeating IC within 3 months if the patients would not respond to medical treatments and were still clinically suspected to OSSN. The histopathological features were available in 15 eyes and diagnostic for OSSN in 13 eyes that showed positive IC results for OSSN, and for pterygium in two eyes that had IC results indicative of squamous cell metaplasia without cellular atypia. Excluding the cases with the histopathological diagnosis of pterygium, none of the atypia-negative IC cases were referred for repeat IC. In 17 (8.9%) out of the 191 eyes diagnosed as OSSN at the first IC session, multiple post-treatment follow-up sessions of IC were performed, disclosing residual and/or recurrent OSSN in 11 cases (64.7%).
DERD
DERD was suspected clinically in 79 cases, and out of them, 76 (96%) cases were cytologically proved with a Cohen’s Kappa agreement of 0.443. The IC results disagreed with the clinical diagnosis of DERD in two cases of unremarkable epithelium on cytology and one case of cytologically-proven OSSN.
Pigmented lesions: BPL and MPL
Out of 16 pigmented lesions, 13 lesions were clinically suspected to BPL in forms of conjunctival nevi, and three were clinically suspected to MPL in form of PAM or melanocytic nevi with atypia. IC results were consistent with the clinical diagnosis in 11 (84%) BPL and 3 (100%) MPL with the Cohen’s Kappa agreement of 0.843 and 0.749, respectively. The IC results were not in line with the clinical diagnosis of BPL in one case of pigmented OSSN and one case of melanocytic nevi with atypia on cytology. In other words, IC led us to figure out the presence of atypia in two cases of clinically diagnosed BPL. The histopathological features were available and diagnostic in four eyes for benign melanocytic nevi, in three eyes for conjunctival melanoma in the context of PAM with atypia, which had positive IC results for BPL and MPL, respectively. Multiple post-treatment follow-up sessions of IC were performed in three eyes (100%) that were diagnosed as MPL at the first IC session, in which residual and/or recurrent MPL were disclosed in all three cases (100%).
AK and IRC
In our series, AK and IRC were clinically suspected in 49 and 5 cases, respectively. IC results in the AK group demonstrated the presence of Acanthamoeba cysts and trophozoites in 46 (94%) with the Cohen’s Kappa agreement of 0.966. In the IRC group, the IC results supported the primary clinical diagnoses in 4 (80%) cases with the clinical diagnosis of atopic (three eyes) and vernal (one eye) keratoconjunctivitis, showing a Cohen’s Kappa agreement of 0.567. Also, IC results in one case of the IRC demonstrated DERD.
Discussion
Our study demonstrated an overall substantial level of agreement between the IC results and the primary clinical diagnoses in various types of ocular surface disorders. This agreement, in the OSSN and LSCD groups that constituted the most populated groups surveyed in this study, was substantial; however, it was substantial to almost perfect in the least populated groups such as MPL, BPL, and AK whose clinical pictures can be quite distinctive.
The IC results in the OSSN, MPL, and BPL groups, by distinguishing neoplastic from non-neoplastic lesions, were of great help to be subsequently subjected to non-invasive medical treatment modalities, based on the preferences of the surgeons, rather than invasive surgical interventions. Excisional biopsy as the most traditional and precise method for the treatment of neoplastic ocular surface lesions, may damage adjacent normal tissue and induce LSCD, scar formation and visual disorders. 27 Along with the recent implementation of topical interferon alfa-2b for the management of OSSN 28 and mitomycin C for the management of MPL in our university-based ophthalmic centers (unpublished data); it was not possible to have post-surgical specimens for histopathology in most of our surveyed cases. However, in our survey, the IC results showed 100% correlation with the present histopathology reports as the gold standard diagnostic method. Multiple reports indicated a high correlation between the IC results and histopathology reports in OSSN, MPL, and BPL.1,5,29,30 The high agreement (CKC = 0.86) between the IC results and histopathology in our survey, in addition to the high experience of our ocular pathologists in IC sampling and interpretations, could be resulted from the presence of intraepithelial components in the OSSN and melanocytic lesions.
The IC results in our survey could exclude the presence of atypia in 34% of clinically suspected OSSNs and prevent making unnecessary surgical damages. However, there is a possibility that some of the IC-negative results for OSSN may be resulted from the presence of IC drawbacks such as lack of full sensitivity of this method (up to 92% sensitivity)1,4,31 and its limitations in representation of epithelial deeper layers that is necessary for the detection of low-grade OSSNs with deeply seated atypical cells. 32 To reduce the possibility of false-negative results in our study; several repeated applications of cellulose acetate filter paper had been performed for the lesion of interest. Moreover, the ophthalmologists were advised to refer the atypia-negative IC patients if the lesions would not respond to medical therapy and the OSSN was still suspected.
Only 17 out of 191 IC-diagnosed OSSN cases were referred by the corresponding ophthalmologists to our ocular pathology laboratory for post-treatment follow-up ICs, in which residual and/or recurrent OSSNs were clinically suspected. It seems that in the rest of the cases, the effect of topical treatments was only evaluated clinically and none of the cases were suspicious to residual or recurrent OSSN on follow-up examinations.
A moderate level of agreement was observed between the IC results and the clinical diagnosis of LSCD in the current study. The diagnosis of LSCD based on IC in our series was crucial because by having the certain diagnosis of LSCD in our patients, there could be a high potential for preventing the progression of LSCD and ocular surface damage via application of appropriate therapeutic modalities to restore functional limbal stem cells. 33 The moderate rate of IC-negative results for LSCD can be explained by the misinterpretation of a fibrovascular pannus clinically as a LSCD. The clinical diagnosis of LSCD is mainly made based on past medical history and the presence of clinical signs of corneal epitheliopathy and corneal superficial vascularizations. However, the clinical signs in cases with old fibrovascular pannus could be misinterpreted as a LSCD.34,35
Although IC results in our survey could support the diagnosis of DERD in 96% of the clinically suspected DERD, the statistical analysis did not show a substantial to perfect correlation between the IC results and the corresponding clinical suspicions in this group (CKC = 0.443). Interestingly, in one case with the clinical suspicion to DERD, the IC result indicated the presence of OSSN and was of great help in changing the therapeutic strategy in that case.
Our survey demonstrated an almost perfect agreement between the IC results and the clinical diagnosis of AK (CKC = 0.996). A high yield of Acanthamoeba cysts and/or trophozoites (94.6%) by IC was previously reported in cases with clinical diagnosis of AK. 21 Given that microbiologic cultures for the diagnosis of AK can be time-consuming and may yield up to only 68% positive results, 36 and considering that other high-tech diagnostic methods such as PCR 37 and confocal microscopy 38 may not be available in most of the ophthalmic centers; the IC technique can be considered as a good option in those ophthalmic centers that have a pathology or ocular pathology department.
In multiple studies, IC has been utilized not only for the diagnosis of atopic and vernal keratoconjunctivitis but also for investigating the possible effects of various topical medications on conjunctival immune cells.23,39 In our survey, the clinical diagnoses of IRC were moderately supported by the IC results, excluding one case in which the cytology yielded the diagnosis of DERD with no evidence of inflammation.
In the current study, multiple post-treatment follow-up sessions of IC were performed in 11.4% of those eyes showing positive IC results either for neoplastic lesions such as OSSN and MPL or for LSCD as a nonneoplastic entity. Performing IC in the former, along with previous studies,1,29,30,40,41 was of great help in the detection of residual or recurrent neoplastic lesions, as the basis for further therapeutic strategies. In the LSCD group, post-treatment IC sessions were performed to detect persistence of the LSCD that was necessary for further limbal or corneal graft surgeries.
One of the shortages of the IC in OSSN and MPL cases, is the disability of this technique in distinguishing in situ from invasive OSSN and MPLs. 15 In these cases, IC in an outpatient clinic setting, by detecting the presence of atypical cells, could be of great help for making better informed decisions regarding the treatment modality. In addition to advantages of IC such as preservation of limbal stem cells, 4 providing a filter paper-sized flat mount of the superficial epithelia with well-preserved morphology compared to the small size of conjunctival smears with deformed epithelial morphology or to conjunctival incisional biopsies with the risk of missing lesions, makes IC an ideal sampling technique for the diagnosis of ocular surface disorders.27,42
In conclusion, in our survey, the clinical suspicions of ocular surface disorders, excepting the DERD and IRC groups, were well supported by the IC results and the IC results were very well-correlated with the existent histopathological data. Despite various drawbacks of IC, it can be crucial for the primary and follow-up managements of the patients with various types of ocular surface disorders, especially in distinguishing malignant from benign lesions. However, it should be remembered that the clinically suspected OSSN lesions should be carefully followed even when the IC does not show atypia.
Footnotes
Author contributions
MRK contributed to the conception or design of work, performing impression cytology, interpreting impression cytology slides, drafting of the article and critical review of the article. SMM and MA contributed to the data interpretation and drafting of the article. SBH contributed to performing impression cytology and interpreting impression cytology slides. MRK contributed to the data collection and data analysis.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethics approval
Full ethical approval was obtained from the Ethics Committee of the Ophthalmic Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
