Ultra-high-frequency ultrasound in the detection of recurrent conjunctival melanoma: A case report

Abstract

BACKGROUND

Conjunctival melanoma is rare and prone to local recurrence, and optimal surveillance imaging of the ocular surface and periocular skin is not standardized. We report the use of ultra-high-frequency ultrasound (UHFUS) to map multiple clinically subtle recurrent nodules in a previously treated conjunctival melanoma (Co-M) patient undergoing immunotherapy. This case supports incorporating UHFUS into multidisciplinary follow-up, showing how it can non-invasively detect occult recurrence and guide staging and treatment planning.

CASE SUMMARY

A middle-aged patient with a history of conservatively treated left Co-M was referred for suspected local recurrence, reporting a new palpable subcutaneous lump near the lateral orbital rim and a pigmented lesion at the medial canthus of the same eye. Clinical examination confirmed a periocular nodule suggestive of relapse. Cervical lymph node ultrasound showed no suspicious lymphadenopathy. UHFUS (48 MHz) with Doppler of the periocular region revealed epidermal thickening at the medial canthus and three subepidermal nodules in the lower eyelid and lateral canthus, with irregular morphology and increased vascularity, highly suggestive of recurrent melanoma and satellite lesions. The patient was already receiving first-line anti-programmed cell death 1 immunotherapy; high-frequency ultrasound findings refined local staging, confirmed absence of nodal involvement, and supported multidisciplinary planning and imaging-based follow-up.

CONCLUSION

UHFUS can non-invasively detect and map recurrent Co-M, improving local staging and guiding multidisciplinary follow-up.

Key Words: High-frequency-ultrasound; High-resolution imaging; Melanoma; Ultra-high-frequency ultrasound; Conjunctival melanoma; Case report

Core Tip: This case report describes recurrent conjunctival melanoma in a previously treated eye, evaluated with ultra-high-frequency ultrasound (UHFUS) (48 MHz). UHFUS accurately depicted subtle periocular epidermal thickening and multiple subepidermal nodules not fully appreciated on clinical examination, while confirming the absence of suspicious cervical lymph nodes. Integrating UHFUS into multidisciplinary follow-up refined local staging in a patient on anti-programmed cell death 1 therapy and supported an imaging-based, organ-preserving management strategy.

INTRODUCTION

Conjunctival melanoma (Co-M) is a rare but highly aggressive ocular surface malignancy arising from melanocytes within the conjunctival epithelium[1-4]. Although it accounts for only a small fraction of all melanomas, Co-M is associated with high rates of local recurrence, regional lymph node metastasis, and distant spread, leading to substantial disease-specific mortality[5-7]. These features underscore the need for accurate local staging and long-term surveillance, particularly after conservative, eye-preserving treatments.

Current follow-up of Co-M relies mainly on slit-lamp biomicroscopy, anterior segment optical coherence tomography, and conventional B-scan ultrasonography[8,9]. However, these tools may be suboptimal for detecting small, subepidermal periocular nodules or satellite lesions after prior treatments and in the context of systemic immunotherapy[10-12]. Only limited reports have described the use of high-frequency ultrasound (HFUS) for periocular tumors, and its role in mapping recurrent Co-M remains poorly characterized[13].

Here, we report the case of a patient with previously treated left-eye Co-M who developed a suspected periocular recurrence while on anti-programmed cell death 1 (anti-PD-1) therapy and underwent 48-MHz HFUS evaluation. Our aim was to explore whether HFUS could identify clinically occult periocular recurrences, refine local staging, and inform multidisciplinary management. We hypothesized that HFUS would detect additional subepidermal nodules beyond clinical examination, thereby addressing a specific and increasingly relevant clinical problem: Surveillance of recurrent Co-M in the era of immunotherapy.

CASE PRESENTATION

Chief complaints

A patient with a history of conservatively treated Co-M of the left eye, previously managed at another institution, was referred to our high-frequency ultrasound unit at the University Hospital of Campania “Luigi Vanvitelli” for evaluation of a palpable subcutaneous nodule in the left periocular region, clinically suspicious for local recurrence.

History of present illness

The patient was receiving first-line systemic immunotherapy with an anti-PD-1 monoclonal antibody for the management of Co-M according to oncologic indications.

History of past illness

The most relevant past medical history was the previously treated left-eye Co-M, managed with eye-preserving, conservative treatment. Apart from ongoing anti-PD-1 immunotherapy, no other systemic therapies, depot injections, or herbal or over-the-counter medications were reported as influencing the clinical course. There was no history of drug allergies or previous adverse drug reactions relevant to the current presentation.

Personal and family history

No additional prior ocular tumors, hereditary cancer syndromes, or familial ocular diseases were reported as contributing to the current condition. Details of other systemic comorbidities, social habits, dietary history, and family history were not contributory to the present case.

Physical examination

At referral, ophthalmologic examination revealed a pigmented lesion at the medial conjunctival canthus of the left eye associated with a clearly visible and palpable subcutaneous nodule along the lateral orbital rim, raising suspicion of local relapse (Figure 1).

Figure 1 Clinical photograph at the time of referral showing a pigmented lesion at the medial conjunctival canthus and a clearly visible subcutaneous nodule along the lateral orbital rim. These findings were suggestive of local recurrence of conjunctival melanoma.

Laboratory examinations

Routine laboratory tests, including hematologic and biochemical parameters, did not reveal abnormalities that impacted diagnostic or therapeutic decision-making.

Imaging examinations

As part of the restaging work-up, high-frequency ultrasound (24 MHz) of the cervical lymph node stations was performed. Ultrasound of the left laterocervical region demonstrated a single lymph node with preserved morphology, visible fatty hilum, regular cortex, and no signs of pathological enlargement or altered echogenicity, findings consistent with the absence of metastatic involvement (Figure 2). Subsequently, ultra-high-frequency ultrasound (UHFUS) (48 MHz) of the left periocular region was carried out, focusing on the pigmented lesion at the medial canthus and on the clinically palpable nodules.

Figure 2 High-frequency ultrasound (24 MHz) image of the left laterocervical region showing a single lymph node (outlined in blue dashed line), not suspicious for metastatic involvement. The node presents with preserved morphology, including a visible fatty hilum and a well-defined cortex, with no signs of pathological enlargement or altered echogenicity.

At the medial canthus/epicanthus region, UHFUS showed epidermal thickening extending into the hypodermis with disorganized echotexture and irregular margins, superficial to the lacrimal canal, with increased Doppler signal suggestive of neovascularization (Figure 3). In correspondence with the palpable nodule of the lower eyelid, a well-demarcated, subepidermal hypoechoic lesion measuring approximately 9 mm × 8 mm with scattered hyperechoic foci was visualized, consistent with a suspected conjunctival satellite metastasis; color Doppler demonstrated increased and accelerated perilesional vascularization (Figure 4). A second hypoechoic nodule with internal hyperechoic spots, located deeper in the subcutaneous tissue adjacent to the first satellite lesion, was identified, with a maximum diameter of about 12 mm × 10 mm (Figure 5). A third lesion at the lateral canthus appeared as a subepidermal nodule with irregular margins and slightly higher echogenicity than the other nodules, showing predominantly intralesional vascularization on Doppler evaluation (Figure 6). Taken together, UHFUS with color Doppler demonstrated multiple subepidermal and conjunctival nodules, including clinically occult lesions, with morphologic and vascular features highly suggestive of recurrent Co-M.

Figure 3 Multimodal assessment of suspected conjunctival melanoma recurrence. A: Real-time clinical photograph showing pigmentation and visible subcutaneous nodule in the periocular region; B: Ultra-high-frequency ultrasound image (48 MHz) revealing epidermal thickening with disorganized echotexture and irregular margins; C: Corresponding Doppler image demonstrating increased vascular signals consistent with neovascularization.

Figure 4 Multimodal evaluation of a subepidermal nodule in the lower eyelid, suspicious for conjunctival satellite metastasis. A: Real-time clinical photograph with arrow indicating the anatomical location of the palpable lesion along the left lower eyelid; B: Ultra-high-frequency ultrasound image (48 MHz) showing a well-demarcated, hypoechoic subepidermal nodule (9 mm × 8 mm) with scattered hyperechoic foci; C: Color Doppler image revealing increased and accelerated perilesional vascularization, consistent with active neoplastic tissue.

Figure 5 Identification and characterization of a second periocular lesion adjacent to the previously described satellite nodule. A: Real-time clinical photograph with arrow indicating the anatomical location of the deeper, non-palpable lesion in the left periocular region; B and C: Ultra-high-frequency ultrasound images (48 MHz) of the lesion in orthogonal planes, showing a well-defined, hypoechoic nodule with internal hyperechoic foci, measuring up to 12 mm × 10 mm. The lesion is located in deeper subcutaneous tissues compared to the satellite metastasis.

Figure 6 Detection of a third subepidermal lesion at the lateral canthus of the left eye. A: Clinical photograph with arrow indicating the anatomical location of the lesion at the lateral canthus; B: Ultra-high-frequency ultrasound image (48 MHz) showing a subepidermal nodule with irregular margins and slightly higher echogenicity compared to previously identified nodules; C: Color Doppler image demonstrating predominantly intralesional vascularization, suggestive of biologically active neoplastic tissue.

MULTIDISCIPLINARY EXPERT CONSULTATION

The cervical lymph node ultrasound and subsequent 48 MHz UHFUS examinations of the periocular region were requested as part of the restaging work-up under ongoing immunotherapy, with the aim of better defining local and regional disease extent and supporting multidisciplinary therapeutic planning.

FINAL DIAGNOSIS

Recurrent Co-M of the left eye with multiple periocular subepidermal satellite nodules demonstrated by UHFUS, without ultrasonographic evidence of cervical lymph node metastasis in a patient on anti-PD-1 immunotherapy.

TREATMENT

At the time of referral, the patient was already undergoing first-line systemic therapy with an anti-PD-1 monoclonal antibody for Co-M. This case report focuses on the imaging-based assessment and mapping of recurrent lesions; detailed description of subsequent local therapeutic interventions (such as surgery or radiotherapy), if any, is beyond the scope of the present report.

OUTCOME AND FOLLOW-UP

Cervical lymph node ultrasound confirmed the absence of sonographic signs of nodal metastasis, while UHFUS (48 MHz) of the periocular region demonstrated multiple subepidermal nodules with morphologic and vascular features highly suggestive of recurrent Co-M, including clinically occult satellite lesions. These findings refined local and regional staging and supported multidisciplinary planning of further management while the patient was receiving ongoing anti-PD-1 immunotherapy. The present report focuses on the diagnostic contribution of UHFUS; detailed long-term clinical outcomes and extended follow-up imaging are not reported.

DISCUSSION

Co-M is a rare but aggressive ocular surface malignancy with substantial risks of local recurrence, lymphatic spread, and melanoma-related mortality. Reported local recurrence rates remain high, ranging from approximately 20% in more recent large series to over 50%-60% in older cohorts, despite advances in surgery and adjuvant therapies[14]. These data support the need for meticulous long-term surveillance, particularly in patients treated with eye-preserving strategies and in those receiving systemic checkpoint inhibitors, in whom patterns of local and in-transit recurrence may be complex[15].

In current practice, follow-up of Co-M relies mainly on slit-lamp biomicroscopy, anterior segment optical coherence tomography, ultrasound biomicroscopy (UBM), and cross-sectional imaging when indicated. High-frequency ultrasound (HFUS, approximately 20-50 MHz) and UBM have been used to measure conjunctival tumor thickness, assess scleral or intraocular invasion, and guide surgical planning[16]. However, their role in systematically mapping small, subepidermal periocular satellites or recurrent nodules-especially in previously treated, complex anatomical areas such as the canthi and eyelid skin-has been only sparsely addressed[17,18]. In parallel, dermatologic literature has shown that HFUS and UHFUS (≥ 40 MHz) can non-invasively characterize cutaneous malignancies, defining lesion depth, margins, and vascularity to complement clinical examination and dermoscopy[19-23].

Our case adds to this body of work by illustrating how 48 MHz UHFUS with Color Doppler can be applied to a patient with previously treated Co-M on anti-PD-1 therapy, in whom both a pigmented medial canthal lesion and a lateral periocular nodule raised suspicion of recurrence. Unlike most prior ophthalmic reports where HFUS/UBM was primarily used preoperatively to assess conjunctival tumors and intraocular extension[16], here UHFUS was used in a surveillance setting to systematically scan the entire periocular region. This approach revealed not only the clinically evident lesions but also additional subepidermal nodules with malignant-appearing morphology and vascular patterns, effectively “mapping” a multifocal periocular recurrence that would likely have been underestimated by clinical examination alone. The negative 24 MHz cervical lymph node ultrasound, consistent with the absence of regional metastases, further refined staging in a non-invasive manner.

From a temporal perspective, the sequence of events-primary Co-M treated conservatively, a disease-free interval, then development of new periocular nodules in the same anatomical region during follow-up-supports a strong temporal association between the current lesions and tumor recurrence. This is biologically plausible given the known propensity of Co-M for local recurrence and periocular spread. Morphologically, the hypoechoic subepidermal nodules with irregular margins, internal hyperechoic foci, and increased perilesional or intralesional vascularity on color Doppler are compatible with malignant lesions, in line with HFUS descriptors reported for cutaneous melanoma and other skin cancers[24]. Alternative diagnoses such as benign cysts, inflammatory nodules, or vascular malformations are less likely, given the imaging pattern and clinical context. Although formal causality scales such as the Naranjo nomogram are designed for drug-related adverse events rather than oncologic recurrence, applying a similar logic (temporal relationship, biological plausibility, typical imaging pattern, and limited alternative explanations) would place the likelihood that these nodules represent recurrent melanoma in the “probable” range. Table 1 HFUS/UHFUS (with Doppler) features that may help differentiate conjunctival/periocular melanoma from benign melanocytic lesions.

Table 1 High-frequency ultrasound/ultra-high-frequency ultrasound (with Doppler) features that may help differentiate conjunctival/periocular melanoma from benign melanocytic lesions.

Ultrasound feature (HFUS/UHFUS ± Doppler)	Melanoma/malignant melanocytic lesion (suggestive)	Benign melanocytic lesion (more typical)
Overall shape/symmetry	Asymmetric, irregular or multilobulated	Symmetric, round/oval
Margins	Irregular, ill-defined, infiltrative appearance	Smooth, well-defined
Echogenicity	Markedly hypoechoic relative to surrounding tissue	Mildly hypoechoic to isoechoic
Internal echotexture	Heterogeneous; disorganised echoes; possible internal hyperechoic foci	Homogeneous or finely granular; usually no marked heterogeneity
Intralesional cystic spaces (especially conjunctival nevi)	Usually absent	Often present (small anechoic cysts), particularly in conjunctival nevi
Epithelial/epidermal changes	Thickening and/or architectural disruption; possible ulceration	Epidermis/epithelium usually preserved; no disruption
Depth/thickness	Greater thickness; extension into hypodermis/subcutaneous tissue; loss of normal planes	Limited depth; confined to superficial dermis/subepithelial tissue; preserved planes
Vascularity (color/power Doppler)	Increased intra- and/or perilesional flow; tortuous vessels; higher flow signals	Absent or minimal flow; if present usually faint/peripheral
Perilesional reaction	Perilesional edema or architectural distortion may be present	No significant perilesional reaction
Multiplicity/satellites	Possible satellite or in-transit nodules; multifocality raises suspicion	Typically solitary; satellites uncommon
Change over time (if serial scans available)	Rapid growth and/or increasing vascularity	Stability over time

UHFUS: Ultra-high-frequency ultrasound; HFUS: High-frequency ultrasound.

CONCLUSION

This case also raises the question of how best to integrate UHFUS into the multidisciplinary management of Co-M in the era of immunotherapy. Checkpoint inhibitors such as anti-PD-1 agents are increasingly used in advanced or recurrent ocular melanomas and can achieve durable responses, but patterns such as pseudoprogression and mixed responses are recognized in cutaneous melanoma[25]. In our patient, the detection of multiple periocular nodules during ongoing anti-PD-1 therapy emphasizes that local disease may progress despite systemic treatment and that precise local staging remains crucial for planning potential surgery, radiotherapy, or other focal interventions. UHFUS could be particularly valuable for: (1) Baseline mapping before starting systemic therapy; (2) Early detection of subtle new lesions; and (3) Monitoring response of periocular nodules when organ-preserving strategies are pursued.

Several limitations of this case report must be acknowledged. First, it is a single-patient observation from a tertiary referral setting, which limits generalizability. Second, histopathologic confirmation for all UHFUS-detected nodules is not available, preventing a direct one-to-one correlation between imaging and pathology; this is a common constraint in imaging-based follow-up, especially when aggressive surgery is avoided. Third, UHFUS is highly operator-dependent and not yet widely available in ophthalmology or dermatology units; standardized acquisition protocols and lesion descriptors for conjunctival and periocular melanoma are still lacking. Fourth, long-term clinical and imaging follow-up data are limited in this report, so the impact of UHFUS-guided staging on ultimate outcomes (local control, eye preservation, survival) cannot be fully assessed. These methodological limitations mean that our findings should be interpreted as hypothesis-generating rather than definitive.

Despite these constraints, the case has several strengths that support its validity and clinical relevance: (1) A clear temporal sequence from primary Co-M to periocular recurrence; (2) Concordant clinical and sonographic findings across multiple, spatially related nodules; (3) Integration of regional nodal assessment with cervical ultrasound; and (4) Use of a very high-frequency (48 MHz) probe with Doppler, enabling simultaneous evaluation of micro-morphology and vascularity. Together, these aspects argue that UHFUS provided accurate, clinically meaningful information that complemented standard examination and influenced multidisciplinary planning.In conclusion, this case illustrates that UHFUS can non-invasively detect and map periocular recurrences of Co-M, including clinically occult satellite nodules, and refine local and regional staging in a patient on anti-PD-1 therapy. Although long-term oncologic outcome cannot be fully assessed from this single observation, our findings support the evidence-based recommendation that clinicians managing Co-M consider adding high- or UHFUS to standard slit-lamp and cross-sectional imaging when evaluating suspected local recurrence. In daily practice, UHFUS may be particularly useful in previously treated, anatomically complex periocular areas where surgery is organ-preserving and imaging must guide multidisciplinary planning. Future research should include prospective studies with larger cohorts, standardized UHFUS protocols, and histopathologic correlation to clarify the impact of UHFUS-guided staging on treatment selection, eye preservation, and survival.