Utility of ultrasound biomicroscopy in detecting invasive ocular surface squamous neoplasia

Abstract

Purpose:

To evaluate the role of preoperative ultrasound biomicroscopy (UBM) in diagnosing invasive ocular surface squamous neoplasia (OSSN).

Methods:

This retrospective single institution-based study included surgical cases of OSSN, in which UBM was performed preoperatively. The UBM characteristics of OSSN and signs of invasion were studied. The preoperative UBM imaging findings were correlated with the postoperative histopathological examination (HPE). The performance of UBM was studied in predicting invasive OSSN, considering HPE as the gold standard.

Results:

Fifteen patients were included in the study, with an average age of 61 ± 12.8 years and an 80% male preponderance (n = 12). A clear plane between the lesion and the sclera was considered as episcleral or ‘non-invasive,’ and disruption of this interface was interpreted as ‘invasive OSSN.’ Compared to postoperative HPE, UBM interpretation predicted invasive OSSN with an accuracy of 73.3%, a sensitivity of 71.4%, and 100% specificity.

Conclusions:

UBM is a useful adjunctive tool for assessing the invasion, lesion height, and depth of invasion in OSSN. On correlating with HPE, UBM showed a good performance in preoperative detection of invasive OSSN.

Ocular surface squamous neoplasia (OSSN) is a clinical diagnosis, and the adjunct use of various imaging modalities in OSSN has been described in the literature.[1,2,3,4,5,6,7,8] In addition to delineating lesion characteristics, imaging modalities like ultrasound biomicroscopy (UBM) and anterior segment optical coherence tomography (AS-OCT) can aid in detecting deeper invasion of OSSN.[5,6] The long-term safety and efficacy of topical pharmacotherapy for OSSN are well documented in the literature.[9,10,11] However, in patients with invasive OSSN (cases with stromal invasion), topical therapy is ineffective as the first-line management. Surgical intervention is the preferred management in invasive OSSN. AS-OCT may not be able to detect stromal invasion in OSSN with excess surface keratin and in large nodular tumors [Fig. 1]. UBM can better delineate lesion characteristics, including depth, and help in deciding line of management as well as preoperative planning [Fig. 1]. There are several studies that examine imaging in OSSN, but studies on invasive OSSN are sparse in the literature. In this study, we assessed the utility of UBM in detecting invasive OSSN and correlated the imaging findings with histopathological features.

Figure 1.

(a) Nodular variant of OSSN on the temporal limbus of the right eye; (b) AS-OCT through the tumor shows a hyper-reflective epithelium of the tumor with increased epithelial thickness and abrupt transition from the normal epithelium, but the posterior limit of the tumor cannot be visualized due to tumor thickness; (c) UBM scan showing the posterior tumor limit and episcleral position of the tumor in the same patient

Methods

The present study is retrospective in design and was conducted at a tertiary-care ophthalmology institute in North India. All cases of OSSN from January 2020 to July 2025 were identified from the electronic medical records (EMR) system. Patients who underwent surgical excision after preoperative UBM were included in the study. OSSN patients who did not undergo preoperative UBM and those who were treated only with pharmacotherapy were excluded from the study. The study adhered to the tenets of the Declaration of Helsinki, and institutional review board (IRB) approval was obtained for the study. Data were collected from the electronic medical records (EMR) system, including demographic details, preoperative and postoperative photographs, AS-OCT and UBM scans, surgical records, and histopathological reports. Tumor characteristics including size, site, feeder vessels, morphology, surface keratin, and pigmentation were noted. Systemic evaluation for lymph node involvement and metastasis was done. UBM was performed for all OSSN cases that were clinically suspicious for invasiveness (scleral fixity or reduced mobility of the lesion), for recurrent tumors, and for cases with suspected intraocular extension (scleral thinning, anterior chamber reaction, raised intraocular pressure, or anterior chamber mass lesion). Tumor thickness was measured, and scleral/corneal involvement was noted. Two clinicians (SD and OR) independently studied the UBM scans, and breach in the plane between the globe and the lesion (invasiveness) was determined by mutual concurrence. Cases with ambiguous findings or incomplete data, and cases where mutual agreement on invasiveness of lesion could not be reached were excluded from the study. For excision, patients were operated using the standard no-touch technique as described by Shields et al.[12] with 4 mm clear margins and double freeze-thaw cryotherapy of the excised margins and base. Patients with full thickness scleral involvement or intraocular spread underwent extended enucleation with ball implant. Histopathological examination (HPE) was performed for all cases. Patients with excised lesions having base positivity (invasive OSSN) on HPE were treated with adjuvant plaque brachytherapy.

Statistical analysis

Data collection was performed using MS Excel spreadsheet. Descriptive statistics were expressed as mean and standard deviations, and categorical variables were explained using frequencies/percentages.

Results

Fifteen patients met the inclusion criteria in the study period. The average age of the study population was 61 ± 12.87 years, and the male: female ratio was 4:1 (12 males and 3 females). The average duration of presentation was 4 ± 2.8 months after noticing the lesion. Three patients (20%) had earlier undergone excision elsewhere and presented with recurrence or residual lesion. Six patients (40%) had received topical chemotherapy but had not responded to treatment. Five patients (33.3%) had right eye affection, and in 10 cases (66.6%), the left eye was affected. As per the 8^th edition of the American Joint Committee on Cancer (AJCC) Staging Manual,[13] two cases (13.3%) were staged as T1N0M0, eight cases (53.3%) as T2N0M0, and five (33.3%) as T3N0M0. One patient (6.6%) was a known case of xeroderma pigmentosa with prior history of OSSN in the contralateral eye, which was intraepithelial and had resolved completely with topical medications. Two patients (13.3%) had scleral thinning along with the tumor. UBM was performed for all the patients (100%), and the average tumor thickness measured was 2.66 ± 1.24 mm. OSSN was observed as a mass lesion with a hyper-reflective outer surface and a hyporeflective stroma [Fig. 2]. Intraepithelial lesion was diagnosed when a clear demarcation was seen between the posterior lesion surface and the underlying sclera/cornea [Fig. 2]. Disruption of this plane was diagnosed as invasive OSSN [Fig. 2]. On UBM, invasive OSSN showed variable reflectivity in the penetrated tissues. Patients with full-thickness scleral involvement showed variable echogenicity in the affected tissues [Fig. 2]. Four patients (26.6%) had full thickness scleral involvement on UBM, and one patient (6.6%) had intraocular spread (ciliary body involvement). These five patients (33.3%) underwent extended enucleation. In addition to these five cases, five other cases (33.3%) were diagnosed as having scleral invasiveness on UBM, while the remaining five cases (33.3%) were diagnosed as intraepithelial tumors. Ten patients (66.6%) underwent excision with the no-touch technique with wide margins. Nine of these patients (90%) were noted to have deeper invasion on histopathological examination. On correlating with postoperative HPE, preoperative UBM evaluation achieved an accuracy of 73.3% in detecting invasive OSSN. The sensitivity and specificity of UBM for invasive OSSN detection were 71.4% and 100%, respectively. Metastatic workup and viral serology were negative in all the patients. Clinical details with UBM and HPE findings of the cohort are elucidated in Table 1. Performance of UBM is summarized in Table 2.

Figure 2.

(a) UBM scan in a case of OSSN showing a hyperechoic tumor surface (white arrow), hypoechoic tumor stroma (red star), and the demarcation line of separation from the underlying cornea (yellow arrow); (b) the red dotted line highlights the demarcation line between the posterior tumor surface and the cornea in the same scan as a; (c) scleral invasion (red arrow) picked up on UBM – note the hyperechogenecity near the red arrow and the absence of demarcation between the posterior tumor surface and the corneal surface; (d) intraocular invasion of OSSN on UBM showing full thickness scleral involvement (red arrow) and the increased thickness of the iris (yellow arrow). No differentiation can be made between the tumor and the underlying structure of the sclera

Table 1.

Clinical, UBM, and histopathological details of the study population (N=15)

Parameter		N (%)
Case distribution as per AJCC classification
T1     T2     T3		2 (13.3%) 8 (53.3%) 5 (33.3%)
Clinical characteristics
Morphology     Nodular mass     Papillary     Leukoplakic     Gelatinous     Associated scleral thinning     Intraocular spread     Presence of feeder vessels     Pigmented tumor     Clinical signs of invasion         T1         T2         T3		6 (40%) 3 (20%) 3 (20%) 3 (20%) 2 (13.3%) 1 (6.6%) 15 (100%) 1 (6.6%) 9 (60%) 1 (6.6%) 4 (26.6%) 4 (26.6%)
Systemic comorbidities
Xeroderma pigmentosa     Hepatitis B infection		1 (6.6%) 1 (6.6%)
Systemic evaluation
Lymphadenopathy     Distant metastases		0 0
Invasion on UBM		10 (66.6%)
T1     T2     T3		1 (6.6%) 4 (26.6%) 5 (33.3%)
Average tumor size on UBM		2.66+1.24 mm
Management
Extended enucleation     Excision with no-touch technique     Adjuvant plaque radiotherapy		5 (33.3%) 10 (66.6%) 10 (66.6%)
Invasive OSSN on histopathology		14 (93.3%)
T1     T2     T3		1 (6.6%) 8 (53.3%) 5 (33.3%)

Table 2.

Comparing the performance of UBM with histopathological examination

Parameter		Number/Percentage
Total cases in the study		15
Invasive OSSN on UBM evaluation		10 (66.6%)
Noninvasive OSSN on UBM evaluation		5 (33.3%)
Invasive OSSN on histopathological evaluation		14 (93.3%)
Noninvasive OSSN on histopathological evaluation		1 (6.7%)
Accuracy of UBM in detecting invasive OSSN		73.3%
Sensitivity of UBM in detecting invasive OSSN		71.4%
Specificity of UBM in detecting invasive OSSN		100%

Discussion

Management of OSSN has evolved in the recent past, with the support of auxiliary imaging and the advent of topical pharmacotherapy. The use of topical drugs has become popular as a first-line treatment for OSSN; however, surgical excision with wide margins remains the gold standard treatment. In suspected invasive OSSN, confirmation of corneal stromal or scleral invasion and depth of invasion is important. Cases with superficial or partial invasion should be planned with sclerectomy/keratectomy or adjuvant plaque brachytherapy. Cases that turn up with ‘base positivity’ on HPE also warrant adjuvant treatment such as plaque brachytherapy. When the invasion is full-thickness or intraocular, extended enucleation should be planned. Clinical suspicion of invasive OSSN may arise when the tumor is fixed to underlying sclera and has limited mobility. In advanced cases with intraocular invasion, clinical signs like anterior chamber reaction, high intraocular pressure, or anterior chamber mass lesion may be observed.[14] AS-OCT is a routinely used noninvasive imaging technique in the clinics which provides high-resolution images of OSSN, and characteristic findings of the tumor are appreciated.[15] AS-OCT helps in differentiation of intraepithelial tumor from invasive tumor by showing a plane of separation from the sclera in intraepithelial cases.[5] AS-OCT carries the benefit of providing high-resolution images; however, as it is an optical method, it has its own limitations including back shadowing in cases of excess surface keratin and thick tumors. Delineating the posterior limit of the tumor on AS-OCT is not possible in such cases. UBM is a noninvasive technique like AS-OCT, but it trades resolution for depth. UBM relies on high-frequency ultrasonic waves and unlike AS-OCT is not affected by the presence of surface keratin or height of tumor. As it provides visualization to a greater depth than AS-OCT, it can be used to image the posterior limit in larger tumors.

We performed UBM for all cases of OSSN that were clinically suspected to be invasive, for nodular tumors, and for tumors that recurred after surgery. UBM could help in diagnosing invasive OSSN with an accuracy of 73.3% when compared to HPE. Further, the thickness of residual bed and scleral thickness could also be measured on UBM, for planning the dosimetry of adjuvant plaque brachytherapy. UBM can also be used to delineate the extent of invasiveness and plan sclerectomy during excision. Cases which could not be picked as invasive OSSN on UBM were microinvasive tumors or tumors with focal invasiveness on HPE [Fig. 3]. Four cases were diagnosed as noninvasive on UBM but turned out microinvasive or with focal base positivity on HPE. Of these four cases, intraoperative episcleral adhesiveness was noted in one case, while the rest were freely dissected during surgery. While UBM could not detect focal invasiveness, it was utilized to plan adjuvant plaque brachytherapy in these cases by measuring the scleral thickness for dosimetry calculation. Meel et al.[6] noted that UBM was helpful in diagnosing OSSN with intraocular extension (IO) and thick nodular tumors >5 mm height were more likely to have intraocular extension. The average tumor height noted for OSSN with IO extension was 5.06 mm in their study. In our study, only one patient had intraocular extension, with a tumor height of 4.13 mm. A previous Indian study[16] noted an average tumor height of 2.93 mm on UBM, which was near to the average tumor height noted by us (2.66 mm). This study also noted a greater tumor height in cases with IO extension (4.3 mm).[16]

Figure 3.

(a) Left eye leukoplakic OSSN in a 62-year-old male patient; (b) on UBM scan, the entire lesion was noted to be episcleral and diagnosed as noninvasive; (c) HPE picture (10x) showing the tumor cells breaching the basement membrane, invading the stroma, and reaching the till the base of the tumor with a focal breach in the posterior limit (red arrow) – a histopathological diagnosis of invasive OSSN with ‘base positive’ was made, and the patient underwent plaque brachytherapy

Our study showed good performance of UBM as a clinical aide for invasive OSSN but has certain limitations. UBM is a subjective technique and has a learning curve, therefore prone to misinterpretation. The sample size was relatively small, and the study was not blinded. One-third of our cases (n = 5, 33.3%) had either full-thickness scleral involvement (n = 4) or intraocular spread (n = 1) and may have introduced a performance bias. Finally, UBM being an ultrasonographic technique cannot diagnose focal or micro invasiveness, which can only be diagnosed on histopathology.

Conclusion

UBM is a practical tool with good performance for aiding the clinician in diagnosing invasive OSSN, for prognostication and operative planning. Although UBM is a less commonly employed imaging technique, effective use as an auxiliary tool provides useful information in suspected cases of invasive OSSN.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.