Background

Talimogene laherparepvec (T-VEC) became the first oncolytic virus approved by the Food and Drug Administration for local injection of unresectable metastatic melanoma. As a genetically modified herpes simplex virus-1, T-VEC has a novel mechanism of action achieved via several genomic modifications to attenuate neurovirulence, increase specificity for cancer cells, enhance antigen presentation, and stimulate the immune response by insertion of the gene for human granulocyte–macrophage colony-stimulating factor (GM-CSF).1

The randomized open-label phase III trial comparing T-VEC to GM-CSF injections, known as Oncovex^GM-CSF Pivotal Trial in Melanoma (OPTiM), demonstrated a superior and significant durable response rate (complete response (CR) plus partial response (PR) lasting ≥6 months) in the T-VEC arm (19.3% vs 1.4%) as its primary endpoint.2 3 CR and PR were achieved in 16.9% and 14.6%, respectively, and median time to CR for those who achieved it was 8.6 months. Of note, assessment of treatment response in these studies has largely relied on clinical observation of changes in size and pigmentation with no interim pathological assessment in most cases.

While remnant pigmentation is an important component of response assessment, it is important to note that clinically suspicious pigmented macules or papules may represent tumorous melanosis, a histopathological term referring to the presence of melanin-laden macrophages in the absence of viable tumor cells. This phenomenon has been recognized by pathologists, but it has been rarely reported in the setting of cancer immunotherapy. Only two case reports to date have detailed its occurrence in the setting of T-VEC therapy.4 5 Additionally, a separate small study detailed the use of quantitative immunofluorescence as a novel diagnostic method in two patients post-T-VEC treatment to distinguish CD68+ macrophages with cytoplasmic SOX-10 costaining likely representing phagocytosed nuclear melanoma fragments from viable melanoma cells with nuclear SOX-10 expression.6 No significant data have been collected regarding the frequency of tumorous melanosis or its time of onset (ie, time of complete pathological response), including large-scale data reported in OPTiM (tumorous melanosis was, in fact, not mentioned). Several additional isolated case reports have been published detailing tumorous melanosis in association with other immunotherapies, including ipilimumab,7 pembrolizumab,8 9 and dabrafenib/trametinib.10 A single case series consisting of 10 patients with tumorous melanosis was also recently published, though none were treated with T-VEC.11

Here we report our experience with six cases of tumorous melanosis following successful T-VEC therapy, in which follow-up biopsies suggest that complete pathological response in a majority of cases occurs significantly sooner than clinically apparent response. To our knowledge, this is the largest case series reporting such findings. We provide a detailed timeline of the clinical courses for our patients, which could serve as a practical guideline for clinicians treating patients with intratumoral therapies.

Case summaries

Among 15 patients treated with T-VEC for melanoma in our institute between 2017 and 2021, we identified 6 patients for whom we performed serial biopsies including post-treatment specimens confirming tumorous melanosis.

The table 1, figure 1 and the online supplemental figure provide a summary of the clinical characteristics and responses with T-VEC therapy in our cohort. All patients included in this study provided informed consent for their clinical data to be used for medical research.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Serial clinical and pathological responses of representative patients. (A) Patient 1 had pigmented papules before treatment, which flattened 3 months after the first injection. Note that the number of lesions increased for the first 3 months of treatment, while the overall size and height of the lesions decreased. Histopathological findings of the patient’s earlier biopsy were those of an atypical, epithelioid proliferation of malignant cells with strong Melan-A red expression, which are overall compatible with malignant melanoma. Additional biopsy post-treatment however demonstrated prominent fibrosis and mixed inflammation with abundant pigmented histiocytes (melanophages). A Melan-A red stain was only focally positive, highlighting cytoplasmic contents of histiocytes. (B) Two apparent subcutaneous nodules of patient 2 coalesced into a single, larger nodule over the course of treatment. The size of the nodule increased over 5 months, which was interpreted as a treatment failure. Histopathological findings were remarkable for dermal fibrosis and numerous melanophages. A Melan-A red stain was only focally positive, highlighting cytoplasmic contents of histiocytes (C) Patient 3 had marked shrinkage one exophytic nodule after her first treatment dose. Remnant pigmented papules and subsequently macules persisted for more than 23 months of treatment, which on biopsy ultimately showed similar features to the prior two patients. There was fibrosis with mixed inflammation, focal cytoplasmic Melan-A red expression in melanophages, and no evidence of residual melanoma.

Among the six patients, five had cutaneous melanoma which was refractory to multiple lines of treatment including immune checkpoint inhibitor (ICI). One patient (case 6) elected for T-VEC injection before ICI for the unresectable but localized superficial tumors on her thigh. One patient (case 5) received concurrent pembrolizumab and T-VEC therapy to control distant metastasis.

All patients had T-VEC therapy for 4–23 months until the remaining lesion was surgically removed (case 2), or pathological CR was achieved. Disease burden varied from two to innumerable dermal or subcutaneous nodules. The size of tumors ranged between 1.0 mm to 1.6 cm. Clinical responses, such as decreasing size of the papular lesions, were typically observed within 2–3 months from the beginning of T-VEC injection. However, complete resolution of pigmentation was delayed by many months in the majority of tumors.

We performed biopsies of five cases when elevated tumors decreased to macular lesions without change in size or pigmentation for several months. In one case (case 2), increased size of the injected tumor prompted surgical excision, which demonstrated no viable tumor cells in pathology. Of note, while the anatomically increased size of the tumor was reported, the maximum standardized uptake value (SUV_max) of the lesion decreased from the previous scan at the time of the excision. The transient increase of tumor size was thus deemed to represent pseudopregression.

Pathological evaluation of those pigmented macular lesions demonstrated fibrosis, inflammation, and varying degrees of pigment-laden melanophages (ie, tumorous melanosis) and no residual viable melanoma in all cases. They were carried out between 4 months and 23 months from the initial treatment, which preceded clinical resolution of disease.

No patient developed regrowth or recurrent melanoma of the injected lesions as of the time of this report. Two patients developed new in-transit metastatic tumors 5 and 2 months after the last dose of T-VEC (case 3 and 5, respectively). All patients experienced grade 1 flu-like symptoms after injections, which tended to decrease in subsequent cycles. No grade 3 or higher adverse events were reported. Patients’ underlying conditions, including Crohn’s disease, immune-related adverse events from pervious treatment, and chronic kidney disease, did not aggravate with T-VEC injection. One patient (case 4), who also was treated with multiple lines of systemic treatment, developed multiple myeloma 34 months after the last injection of T-VEC.

Comment

The treatment of metastatic melanoma has evolved considerably over recent years to include a variety of agents. Among them, T-VEC therapy has demonstrated promising results, especially in treating cutaneous or superficial metastatic melanoma. However, practical data to guide clinicians, such as the variable timing and patterns of responses and toxicity, has rarely been reported.

Typically, clinical assessment of treatment response is made based on tumor size and remnant pigmentation. Despite the understandability of this approach, novel treatments have shown atypical response patterns such as pseudoprogression—a transient increase in tumor size due to inflammation that precedes diminution—especially in melanoma.12

For superficial tumors being assessed by size and color, the evaluation could be challenging due to tumorous melanosis, the presence of melanophage-rich cells without the remaining viable tumor. Indeed, tumorous melanosis has been recognized as a mimicker of metastatic disease in the clinical setting and has been described in patients who received immunotherapy.8 10 11 Rare reports of this phenomenon have been documented in patient groups undergoing T-VEC therapy.13 14 To date, moreover, little has been described regarding time course and extent of tumorous melanosis.

The clinical impression remains the primary deciding factor for therapeutic considerations. A substantial decrease in lesion size is interpreted as a promising, but still incomplete, response. Routine biopsies of lesions to evaluate residual disease is not common practice. In our cohort of six patients, we recognized that visible lesions post injection are not necessarily indicative of residual melanoma. In all six patients, scout biopsies of clinically apparent residual tumor masses after T-VEC therapy simply demonstrated prominent tumorous melanosis with no morphological evidence of viable melanoma, confirmed by immunohistochemical stains in all cases. Histologically, the excisions contained prominent fibrosis with variably dense lymphohistiocytic inflammation and pigment incontinence. These features are evidence of a substantial host response.

In managing patients with advanced disease who require non-surgical treatments, it is thus crucial to recognise that residual tumor mass can be seen post treatment, consisting of only a dense and reactive inflammatory response. It should be noted that biopsies were only performed in this series of cases due to the ambiguous or worrisome clinical setting; thus, we could not accurately assess the incidence of tumorous melanosis in our entire T-VEC cohort.

Our findings emphasize the important contribution of periodic histopathological assessment. The findings also highlight that, in some cases, biological response to therapy may precede and occasionally contradict the assessment of clinical response. The responding tumor in one of our patients showed decreased fluorodeoxyglucose (FDG) uptake and tumorous melanosis in the setting of increasing size of the lesion. As such, assessing metabolic activities of the tumor by FDG-positron emission tomography (PET) could assist the response assessment in select cases. Thus, consideration for biopsy and/or FDG-PET to assess treatment response in select cases with stable tumor size and/or residual pigmentation may aid in preventing unnecessary extended therapy.

Our study is limited by the retrospective, observational nature of data, as well as the small size of our cohort. Potential future studies will be needed to further evaluate the kinetics and pathophysiology of tumorous melanosis, specifically addressing time from initiation of therapy and the course of response progression in a controlled prospective cohort, which will include assessment of true incidence of tumorous melanosis. Lastly, elucidating the precise roles of the immune cells in tumorous melanosis will be an important next step to further our understanding of the tumor microenvironment and to identify potentially useful biomarkers.