Real-world outcomes of talimogene laherparepvec as salvage therapy for advanced melanoma Authors: Melissa M Yamada, Smitha Chandrasekhar, Ted A Gooley, Rita E Chen, Coley Doolittle-Amieva, George Ansstas, Shailender Bhatia, Evan T Hall, Paul T Nghiem, Song Y Park, David Y Chen Published: 2025 ABSTRACT AbstractBackgroundTalimogene laherparepvec (T-VEC) is an oncolytic herpes simplex virus therapy approved for treatment of unresectable and metastatic melanoma. However, real-world use often occurs in heavily pretreated patients, where evidence of effectiveness remains limited. This study evaluates treatment responses and clinical factors influencing T-VEC outcomes in patients with diverse treatment histories.MethodsWe analyzed patients with metastatic melanoma treated with T-VEC between 2015 and 2024. Objective responses (OR), complete response (CR) and partial response were assessed using univariate and multivariate Cox regression models. Durability of responses, progression-free survival (PFS), and overall survival (OS) were evaluated using Kaplan-Meier estimates.ResultsAmong 121 patients, 105 (87%) patients received ≥1 prior lines of systemic therapy; 48 (40%) had a current or prior history of distant metastatic disease, and 42 (35%) had both injectable and non-injectable disease at the T-VEC initiation. Median PFS was 12.2 months (95% CI 6.2 to 20.9), and median OS was 35.5 months (95% CI 25.8 to 63.9). Of 113 evaluable patients, 76 (67%, 95% CI 58% to 76%) achieved an OR, including 39 (35%, 95% CI 26% to 44%) CR. The probability OR by 6 months was 56% (95% CI 46% to 65%). Of the 39 patients achieving CR, 37 (95%) remained alive and progression-free at last follow-up (median 19.1 months). In multivariate analysis, the adjusted HR (aHR) for OR in patients with non-injectable distant metastases at T-VEC initiation relative to those without was 0.43 (95% CI 0.23 to 0.78; p=0.006). The aHR for OR among those immunosuppressed compared with those not immunosuppressed was 0.18 (95% CI 0.04 to 0.69; p=0.013), indicating a reduced likelihood of response for patients who were immunosuppressed. Unadjusted HRs for achieving an OR after 1, 2, and ≥3 prior therapies (vs none) were 1.20 (95% CI 0.57 to 2.52; p=0.627), 1.21 (95% CI 0.52 to 2.80; p=0.653), and 0.77 (95% CI 0.35 to 1.68; p=0.507), respectively.ConclusionsThis study demonstrates T-VEC’s potential efficacy in achieving meaningful disease control and response durability in patients with unresectable and/or metastatic melanoma, including those with diverse prior-treatment histories and comorbidities. WHAT IS ALREADY KNOWN ON THIS TOPIC Talimogene laherparepvec (T-VEC) is an approved oncolytic virus therapy for unresectable melanoma, but real-world evidence is limited, especially in heavily pretreated patients who are often ineligible for or refractory to immune checkpoint inhibitors. Talimogene laherparepvec (T-VEC) is an approved oncolytic virus therapy for unresectable melanoma, but real-world evidence is limited, especially in heavily pretreated patients who are often ineligible for or refractory to immune checkpoint inhibitors. WHAT THIS STUDY ADDS This multi-institutional study demonstrates that T-VEC can achieve meaningful response rates and durable disease control in a diverse, real-world cohort, while highlighting clinical factors—such as non-injectable disease and immunosuppression—that reduce response likelihood. This multi-institutional study demonstrates that T-VEC can achieve meaningful response rates and durable disease control in a diverse, real-world cohort, while highlighting clinical factors—such as non-injectable disease and immunosuppression—that reduce response likelihood. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings support T-VEC as a salvage option for select patients with advanced melanoma and underscore the need for biomarkers and combination strategies to enhance systemic effects, guiding future research and personalized treatment approaches. These findings support T-VEC as a salvage option for select patients with advanced melanoma and underscore the need for biomarkers and combination strategies to enhance systemic effects, guiding future research and personalized treatment approaches. Background Melanoma is an aggressive skin cancer, yet significant advancements in our understanding of melanoma pathogenesis have translated to highly effective therapeutic agents, including small molecule inhibitors targeting the mitogen-activated protein kinase (MAPK) pathway and immune checkpoint inhibitors (ICIs).1 2 Notably, checkpoint blockade therapies have led to durable responses, with a 43% overall survival (OS) at 10 years in patients treated with combined cytotoxic T-lymphocyte-associated protein 4 and programmed cell death protein 1 (PD-1) blockade.3 4 Despite these breakthroughs, over 50% of patients with melanoma will not achieve durable responses, including 10–20% of patients with melanoma who are unable to receive these systemic therapies due to medical conditions contraindicating their use, leaving these patients in need of effective therapeutic options.3 4 Talimogene laherparepvec (T-VEC) is an oncolytic immunotherapy derived from herpes simplex virus type 1 (HSV-1) that has been modified to enhance tumor-selective viral replication, oncolysis, and immune activation; additionally, transgenic granulocyte-macrophage colony-stimulating factor promotes dendritic cell recruitment and maturation. Viral oncolysis of tumor cells may be a source for cross-priming tumor-specific T-cell responses and tumor elimination.357 One major contributor to ICI resistance may be an immunologically “cold” tumor microenvironment (TME), which excludes immune effectors from the tumor. Intratumoral therapy offers an opportunity to directly modulate the TME and overcome systemic checkpoint blockade failure, presenting a viable treatment option for this challenging subset of patients with melanoma.8 9 While T-VEC showed promise in the pivotal OPTiM trial in patients with limited prior exposure to systemic therapy, particularly PD-1 inhibitors,10 in practice, T-VEC is often used for patients who have failed ICI treatment or are ineligible to receive it. Real-world data on its effectiveness remains scarce, particularly regarding key response characteristics such as timeline to response, durability, and factors influencing outcomes, especially in patients with multiple prior lines of systemic therapy. This study evaluates the real-world efficacy of T-VEC in a multi-institutional cohort of patients with heavily pretreated melanoma while identifying key clinical characteristics influencing response. Methods Study design and patient selection This retrospective cohort study was conducted across two research centers, the University of Washington (UW) and Washington University in St. Louis (WashU). The data were analyzed according to the UW and WashU institutional review board-approved study protocol STUDY00018904 and 202309151, respectively. Eligible patients were aged ≥18 years and received T-VEC therapy for histologically confirmed melanoma and sufficient clinical data. The study cohort consisted of 121 patients (n=52 UW; n=69 WashU) who received T-VEC for malignant melanoma since January 2015. Data cut-off date was January 2024. Data collected Comprehensive clinical characteristic data were collected for each patient in the cohort to evaluate their impact on treatment outcomes. This included demographics, initial stage of disease, history of distant metastasis throughout the disease course, extent of disease at the time of T-VEC therapy, tumor characteristics, gene mutations, relevant medical history of immunosuppression (including HIV/AIDS, organ transplant, autoimmune disease, hematologic malignancy, or other causes), herpes virus (HSV) infection, prior treatment history, T-VEC administration regimen, evidence of clinical inflammation, and treatment responses. Staging was determined per American Joint Committee on Cancer eighth edition.11 For assessing molecular drivers, v-Raf murine sarcoma viral oncogene homolog B (BRAF) and neuroblastoma RAS viral oncogene homolog (NRAS) mutations were assumed to be mutually exclusive.12 Prior history of recurrent HSV or positive serology prior to T-VEC initiation was assigned chronic HSV status. Clinical inflammation was defined as erythema or redness at the injection site after initiation of T-VEC, identified either through retrospective photographic review by dermatologists or clear documentation of inflammation in the medical record. Outcome measurement We used investigator-assessed objective response (OR) per modified WHO tumor response criteria.13 14 A minimum of three cycles of T-VEC was required for an accurate assessment of response. Overall response assessment evaluated injected lesion and non-injected lesion responses, the latter of which represents the ability of T-VEC to provoke response in inaccessible metastases. Pseudoprogression was defined as an initial increase in the number or size of lesions, per WHO criteria, followed by OR with a best overall response of complete response (CR) or partial response (PR) within the first 3 months of treatment. Best overall response duration following treatment initiation was defined by time to future systemic treatment, progression, or death. If none of these occurred, best overall response duration was determined to be indefinitely continuing. Reasons for termination of T-VEC were defined as treatment best overall response of CR, PR, or stable disease (SD) with surgical removal of remaining lesions, progressive disease (PD), adverse reactions and toxicity, patient desired break from treatment, or death. Adverse reactions and toxicity included HSV reactivation, neurological symptoms of confusion and altered mental status, and immune reactions defined by the National Cancer Institute Common Terminology Criteria for Adverse Events V.5.0. Statistical analysis The probability of overall and progression-free survival (PFS) was estimated using the Kaplan-Meier method, with time measured following T-VEC termination. Additionally, the probabilities of CR and OR (CR or PR) as a function of time following T-VEC termination were summarized using cumulative incidence estimates, where death without response was regarded as a competing risk. Response data were unavailable for eight patients. Furthermore, the probability of progression following the attainment of response as a function of time was summarized using cumulative incidence estimates. Cox regression models were fit to evaluate the association between individual patient parameters and the attainment of response following T-VEC, treating response as a time-to-event outcome. Cause-specific HRs were used to estimate the “risk” of achieving response; HR values above one are consistent with increased response in the group that composes the numerator reflective of the HR. Most parameters were encoded as dichotomous variables, with exceptions including prior treatment lines (coded as 0–3), injection depth, tumor subtype (coded by individual characteristics). Largest tumor size and the number of injections were treated as continuous linear variables. Results Clinical characteristics of cohort Of 121 patients with melanoma, 52 females (43%) and 69 males (57%) were treated with T-VEC with a median age of 66 years (range, 20–93) at T-VEC initiation. On average, patients began T-VEC treatment 4.1 years after their initial diagnosis of melanoma. Of the 121 patients with evaluable response data, 105 (87%) had received one or more lines of systemic therapy. This was further categorized into 42 patients (35%) who had received one line of systemic therapy, 19 patients (16%) who had undergone two lines, and 44 patients (36%) who had received three or more lines of therapy. Eight patients (7%) received adjuvant ICI therapy, and two (2%) received adjuvant chemotherapy including isolated limb perfusion. Systemic therapies administered for metastatic disease included ICI monotherapy (n=54; 45%), dual checkpoint inhibitors (n=22; 18%), targeted therapy (n=15; 12%), combination of ICI and targeted therapy (n=2; 2%), conventional chemotherapy (n=2; 2%), interferon (n=4; 3%), and other treatments, including clinical trial agents (n=7; 6%). Additionally, one patient participated in a T-VEC clinical trial prior to receiving T-VEC as part of this cohort. 98 patients (81%) had prior exposure to immune checkpoint inhibitors. Concurrent treatment was administered to 62 patients (51%). Among these, ICI monotherapy was the most common (n=35; 29%), followed by dual checkpoint blockade (n=9; 7%), targeted therapy (n=10; 8%), and combination ICI with targeted therapy (n=4; 3%). Four patients (3%) received miscellaneous or investigational agents, including clinical trial drugs or kinase inhibitors not categorized above (online supplemental table 1). 27 patients (22%) received T-VEC monotherapy without concurrent systemic treatment, and treatment data were unavailable for 32 patients (26%). 48 patients (40%) had a current or prior history of distant metastatic disease during their disease course. At the initiation of T-VEC therapy, 42 patients (35%) presented with both injectable and non-injectable disease, while the remaining patients had only injectable lesions. 16 patients (13%) were immunosuppressed. Seven of these patients (6%) had autoimmune disorders, four patients (3%) had a history of organ transplant, three patients (2%) had a hematologic malignancy, and two patients (2%) were under chronic immunosuppression for other underlying conditions. Detailed descriptions of clinical and demographic characteristics are provided in table 1. Table 1Baseline clinical and demographic characteristics of patientsCharacteristicNo. (%)SexFemale 52 (43), male 69 (57)Age at diagnosis (years)Median 66 (range 20–93)Age at T-VEC initiation (years)Median 68 (range 27–96)Time from diagnosis to T-VEC initiation (years)Median 2.5 (range 0.1–24.4)Stage at diagnosisI: 17 (15); II 21 (18); III: 66 (54); IV 13 (11)Tumor subtypeCutaneous 100 (83); acral 17 (14); mucosal 4 (3)Largest tumor size at the time of initial T-VEC injection (cm)Median 1.8 (range 0.2–17.6)Injection site location**Head and neck 29 (25); trunk 21 (19); extremities 82 (70)Injection depth** Subcutaneous49 (41) Lymph node9 (7) Superficial†45 (37) Multiple18 (15)Distant metastases or non-injectable regional lesions at T-VEC initiation No history or presence of metastases, no non-injectable lesions73 (60) History or presence of metastases, no non-injectable lesions6 (5) History or presence of metastases, non-injectable lesions42 (35)Concurrent systemic treatment No59 (49) Yes62 (51)Prior lines of systemic therapy 016 (13) 142 (35) 219 (16) 3+44 (36)Prior ICI No23 (19) Yes98 (81)Immunosuppression No104 (86) Yes16 (13) Autoimmune disorder7 (6) Organ transplant4 (3) Hematologic malignancy3 (2) Chronic immunosuppression2 (2)Chronic HSV at T-VEC initiationNo 54 (45); yes 9 (7)MutationBRAF 38 (31); NRAS 22 (18); KIT 8 (7)*Not mutually exclusive.†Tumors that are confined to dermal lesions with pigmentation.BRAF, V-raf murine sarcoma viral oncogene homolog B; HSV, herpes virus; ICI, immune checkpoint inhibitor; KIT, KIT proto-oncogene, receptor tyrosine kinase; NRAS, neuroblastoma RAS viral oncogene homolog; T-VEC, talimogene laherparepvec. Baseline clinical and demographic characteristics of patients Not mutually exclusive. Tumors that are confined to dermal lesions with pigmentation. BRAF, V-raf murine sarcoma viral oncogene homolog B; HSV, herpes virus; ICI, immune checkpoint inhibitor; KIT, KIT proto-oncogene, receptor tyrosine kinase; NRAS, neuroblastoma RAS viral oncogene homolog; T-VEC, talimogene laherparepvec. Response characteristics, timeline, and durability Median follow-up of the cohort was 29.2 months (95% CI 26.0 to 36.7 months). Median follow-up among survivors was 24.4 months (range, 1.6–90.9 months). Median OS was 35.5 months (95% CI 25.8 to 63.9 months; figure 1A), and median PFS was 12.2 months (95% CI 6.2 to 20.9 months; figure 1B). Kaplan-Meier estimates following T-VEC therapy. (A) OS following T-VEC therapy. (B) PFS following T-VEC therapy. The number of patients at risk is displayed below each curve at corresponding time points. OS, overall survival; PFS, progression-free survival; T-VEC, talimogene laherparepvec. Among 113 patients with evaluable response data, 39 patients (35%) achieved CR, 37 patients (33%) PR, 10 patients (9%) SD, and 27 patients (24%) PD (table 2), with similar outcomes observed between UW (31.4% CR) and WashU (37.1% CR). Among patients who presented with both injectable and non-injectable disease at the start of T-VEC, no apparent effect was observed in non-injectable lesions. Four patients who achieved CR in injected lesions maintained SD in distant sites without progression (table 3, light green). Among those with PR in injected lesions, two had SD in non-injectable tumors, while 17 experienced PD in non-injectable tumors (table 3, dark green). Table 2Summary of best overall response among 113 evaluable patientsAll patientsBest overall responseComplete response39 (35%)Partial response37 (33%)Stable disease10 (9%)Progressive disease27 (24%) Summary of best overall response among 113 evaluable patients Table 3Response patterns among patients with both injectable and non-injectable disease at T-VEC initiationBest response, injected sitesCRPRSDPDUnknownBest non-injected responseCR00000PR00000SD42330PD0274121Unknown00007CR, complete response; PD, progressive disease; PR, partial response; SD, stable disease; T-VEC, talimogene laherparepvec. Response patterns among patients with both injectable and non-injectable disease at T-VEC initiation CR, complete response; PD, progressive disease; PR, partial response; SD, stable disease; T-VEC, talimogene laherparepvec. 30 of the 39 CRs occurred within 6 months, with CRs ranging from 21 to 722 days. At 6 months, the estimated probability of CR was 28% (95% CI 20% to 37%) and OR was 56% (95% CI 46% to 65%). By 1 year, the estimated probability of CR was 32% (95% CI 23% to 41%) and OR was 64% (95% CI 54% to 72%). At 2 years, the probability of CR was 38% (95% CI 28% to 47%) and OR 71% (95% CI 61% to 76%) (figure 2). Among responders, 22 patients had an initial increase in the number or size of lesions within the first 3 months of T-VEC treatment, which was deemed to be pseudoprogression. Estimated probability of achieving response following T-VEC therapy. (A) Probability of achieving a CR as a function of time. (B) Probability of achieving an objective response (CR or PR). The number of patients at risk is displayed below each curve at corresponding time points. CR, complete response; PR, partial response; T-VEC, talimogene laherparepvec. Of the 39 patients who achieved CR as their best response, 37 remain alive without progression at last contact, while 2 died from other causes. The median follow-up among survivors is 19.1 months from achievement of CR (range, 0.7–80.3 months) (figure 3). Among 37 patients who achieved a PR as best response, 10 are alive without progression at last contact, with a median follow-up after PR of 12.1 months (range, 0.9–71.1 months) (figure 3). The median time to progression among those who had PR was 3.2 months. Probability of progression following objective response to T-VEC therapy. (A) Probability of progression following the attainment of CR. (B) Probability of progression following the attainment of a PR. The number of patients at risk is displayed below each curve at corresponding time points. CR, complete response; PR, partial response; T-VEC, talimogene laherparepvec. Clinical characteristics associated with response Univariate associations with ORs, including CR and PR, are summarized in online supplemental table 2. Among these factors, the HR for OR among those who had a history of distant metastases and non-injectable lesions present during initiation of treatment relative to those with neither was 0.39 (95% CI 0.21 to 0.71; p=0.002). Immunosuppressed patients had an 85% lower hazard of OR relative to immunocompetent patients (HR=0.15, 95% CI 0.04 to 0.63; p=0.009). The HRs for achieving an OR after 1, 2, and 3 or more prior therapies, relative to no prior therapy, were 1.20 (95% CI 0.57 to 2.52; p=0.627), 1.21 (95% CI 0.52 to 2.80; p=0.653), and 0.77 (95% CI 0.35 to 1.68; p=0.507), respectively. HRs for presence versus absence of mutations showed an HR of 0.98 (95% CI 0.59 to 1.61; p=0.925) for BRAF, 1.14 (95% CI 0.60 to 2.18; p=0.694) for NRAS, and 0.71 (95% CI 0.28 to 1.80; p=0.466) for KIT proto-oncogene, receptor tyrosine kinase (KIT) mutations. A multivariable Cox regression model for OR (table 4) identified similar associations as the univariate models. Patients with a prior or current history of distant metastatic disease but no non-injectable lesions during T-VEC therapy had an adjusted HR (aHR) of 0.74 (95% CI 0.47 to 1.18; p=0.203) compared with those without a history of distant metastases. Those with both distant metastatic disease and non-injectable lesions had an aHR of 0.43 (95% CI 0.23 to 0.78; p=0.006). The aHR for OR achievement for immunosuppressed patients relative to non-immunosuppressed patients was 0.18 (95% CI 0.04 to 0.69; p=0.013). Table 4Clinical factors associated with objective response: multivariate Cox regression analysisCharacteristicValueNumber OR/number of patients (%)HR (95% CI), p valueGlobal p value (if applicable)SexFemale34/50 (68)1--Male42/63 (67)0.74 (0.47 to 1.18), 0.203--Age at T-VEC initiation--n/a1.05* (0.90 to 1.23), 0.519--History or current presence of distant metastases, presence of non-injectable lesions at the time of initial T-VEC injectionNo history or presence of metastasis, no non-injectable lesions58/71 (82)10.024History or presence of metastasis, no non-injectable lesions5/6 (83)0.81 (0.31 to 2.11), 0.670--History or presence of metastasis, non-injectable lesions13/36 (36)0.43 (0.23 to 0.78), 0.006--ImmunosuppressionNo74/100 (74)1--Yes2/13150.18 (0.04 to 0.73), 0.016--*Modeled as continuous linear variable, HR reflects relative increase in hazard associated with a 1 year increase in age.T-VEC, talimogene laherparepvec. Clinical factors associated with objective response: multivariate Cox regression analysis Modeled as continuous linear variable, HR reflects relative increase in hazard associated with a 1 year increase in age. T-VEC, talimogene laherparepvec. Univariate Cox regression models for CR are summarized in online supplemental table 3. Notably, the unadjusted HR of CR achievement among those with non-injectable metastatic lesions relative to those without was 0.21 (95% CI 0.08 to 0.60; p=0.003). Immunosuppressed patients had an 81% lower hazard of CR compared with patients without overt immune compromise (HR=0.19, 95% CI 0.03 to 1.42; p=0.106). Larger tumors at T-VEC initiation were associated with a 38% decreased hazard of CR per 1 cm increase in size (HR=0.62, 95% CI 0.43 to 0.88; p=0.007). The HR for CR among patients with subcutaneous tumors, relative to those with superficial dermal lesions, was 0.42 (95% CI 0.20 to 0.89; p=0.023). Clinical inflammation following T-VEC injection, compared with no inflammation, was associated with an HR of 2.35 (95% CI 1.18 to 4.71; p=0.016). The HRs for achieving CR after 1, 2, and 3 or more prior therapies, relative to no prior therapy, were 1.67 (95% CI 0.56 to 4.97; p=0.357), 1.85 (95% CI 0.56 to 6.14; p=0.317), and 1.00 (95% CI 0.31 to 3.19; p>0.999), respectively. HRs for CR based on mutation status (presence vs absence) showed HRs of 0.80 (95% CI 0.39 to 1.65; p=0.547) for BRAF, 1.61 (95% CI 0.68 to 3.82; p=0.283) for NRAS, and 0.92 (95% CI 0.27 to 3.09; p=0.890) for KIT mutations. A multivariable Cox regression model for CR is summarized in online supplemental table 4. Patients with a prior history of distant metastatic disease but no non-injectable lesions during T-VEC therapy had a 78% lower adjusted hazard of CR achievement (aHR=0.22, 95% CI 0.03 to 1.88; p=0.168) compared with those with neither. Patients with both a prior history of distant metastatic disease and non-injectable lesions at initiation of T-VEC had an aHR of 0.21 (95% CI 0.07 to 0.61; p=0.004) for achieving CR relative to those with neither. Patients who had clinical inflammation following T-VEC injection had a 120% higher adjusted hazard of CR compared with those without (aHR=2.20, 95% CI 1.09 to 4.44, p=0.028). In patients who achieved a CR to T-VEC therapy, prior ICI treatment responses varied. 12 patients exhibited primary resistance to prior ICI therapy. Eight patients developed acquired resistance to ICI, with initial response to the therapy but subsequent disease progression. For patients who received ICI as adjuvant therapy and initially achieved disease-free status, five experienced recurrence after more than 6 months, while three had recurrence in less than 6 months. One patient had ICI therapy discontinued for reasons unrelated to disease progression; however, the disease showed a positive response following this interruption. Four patients had incomplete ICI response data (online supplemental table 5). Discussion In our study, T-VEC demonstrated comparable response rates and durable efficacy to prior studies, across a heterogeneous cohort of patients with melanoma with diverse systemic treatment histories. Factors such as the presence of non-injectable lesions, history of distant metastatic disease, and immunosuppression were associated with a reduced likelihood of achieving an OR. In contrast, variables such as increasing number of prior treatment lines (including ICIs), concurrent systemic therapy, and tumor mutation status did not show definitive evidence of an association with response. While most responses occurred within the first 6 months, approximately 10% were observed beyond this time point. Treatment was generally well tolerated, with only one confirmed severe adverse event due to disseminated T-VEC infection.15 Clinical factors and responses In our real-world cohort, regression of non-injected lesions was rare, and the presence of non-injectable tumors reduced response likelihood, consistent with prior findings that early metastatic melanoma (stage IIIB–IVM1a) and lower tumor burden predict response.16 17 The regression of non-injected tumors is thought to be an immune-mediated process in which local therapies—such as radiation or intralesional injections—induce tumor cell death, promote the release of tumor-associated antigens, and activate systemic antitumor immune responses.18 In prior studies evaluating responses in non-irradiated lesions following radiation therapy, features of a favorable immune response include elevated numbers of cytotoxic T lymphocytes, increased antigen presentation by dendritic cells, reduced immunosuppressive populations such as myeloid-derived suppressor cells, and heightened natural killer cell activity.19 These immune biomarkers collectively indicate an immunologically “hot” TME capable of propagating systemic antitumor immunity. T-VEC treatment recapitulates some of these features.20 Despite its potential to trigger immune responses locally, tumor response in non-injected lesions was rare in our cohort, and whether they represent recall of pre-existing immune responses or de novo antitumor responses could be a subject for future study. Notably, immunosuppressed patients—who were excluded from prior studies—demonstrated a lower likelihood of response in our study. While existing studies have shown that immunosuppressed patients with melanoma often exhibit diminished responses to immune checkpoint inhibitors, previous investigations of T-VEC have not systematically analyzed this subgroup. Our findings highlight the need for further research to guide treatment strategies in immunocompromised patients.21 22 Tumor mutations and responses Molecular classification of melanoma has been largely guided by driver mutations in the MAPK pathway, with hotspot activating mutations in BRAF and NRAS accounting for approximately 75% of cases.23 Sequencing meta-analyses have identified associations between driver mutations and tumor mutational burden, as well as secondary drivers that may affect immunotherapy responses.4 24 In our cohort, BRAF mutations were assessed in 113 patients and detected in 38 cases, though most cases were examined using an immunohistochemistry assay for BRAFV600E (VE1) which assesses only for this single mutation. In a minority of cases, next-generation sequencing revealed NRASQ61K/R mutations in 22 patients, which was the most frequently mutated gene after BRAF in our cohort. In this study, mutation of either BRAF or NRAS did not show apparent association with outcomes, though this should be interpreted with caution since genomic studies were performed on a minority of patients in this study.2527 More data are required to obtain a more robust conclusion with respect to these mutations. Real-world cohort response characteristics Our study demonstrates high efficacy of T-VEC, with an objective response rate (ORR) of approximately 70% with durable responses, despite a heavily pretreated cohort.28 29 Among patients who achieved a CR, 37 of 39 (95%) remain alive without progression, with a median follow-up time at last contact of 19.1 months from CR (range, 0.7–80.3 months). Previous studies have similarly demonstrated significant durability of responses in T-VEC-treated patients who reached response. Notably, in the OPTiM trial, among T-VEC-treated patients with a CR, 72% (95% CI 57% to 83%) were free from recurrence 3 years after achieving a CR.10 A recent retrospective study found that among 112 patients with metastatic melanoma treated with T-VEC after failure of immunotherapy, 37% had a CR with a median disease-free survival of 32 months (95% CI 17 to not evaluable).29 Overall, our findings in combination with results from prior studies suggest that for patients who achieve a CR, responses to T-VEC are highly durable. Responses to T-VEC in the context of prior or concurrent systemic therapies vary across studies, partly due to the heterogeneity in patient populations, prior treatments, and timing of trials. In the final analysis of the OPTiM pivotal trial, patients treated with T-VEC as a second-line or later therapy had a lower likelihood of response.10 However, the OPTiM trial was conducted from 2009 to 2011, when ICI was not widely available. In a clinical trial designed to assess T-VEC responses in patients with prior anti-PD-1 exposure, subgroup analyses of patients with disease progression to prior ICI therapy showed similarly low response rates, 0–7%.30 A recent trial evaluating the combination of T-VEC and pembrolizumab versus pembrolizumab monotherapy in ICI-naïve patients demonstrated a median PFS of 14.3 months (95% CI 10.3 to 22.1) with the combination therapy, compared with 8.5 months (95% CI 5.7 to 13.5) with pembrolizumab alone.31 Our study revealed no definitive evidence of an association between concurrent systemic therapy, including ICI therapy such as pembrolizumab, and achieving an OR. The heterogeneity of response in these studies suggests that pretreatment conditions and patient selection may influence T-VEC efficacy observed in this study, continued efficacy regardless of prior therapies sets T-VEC apart as an orthogonal therapy to ICI and other systemic treatments, and reflects a gap in knowledge of where precisely T-VEC intervenes in the cancer-immunity cycle.9 In addition, this study highlights the extended response timeline observed with T-VEC compared with other systemic therapies, which aligns with OPTiM data where median time-to-response was 9 months. Unlike ICIs, which often elicit responses within a few months, T-VEC’s median time-to-response is numerically longer.32 This apparent delayed response time may be considered when evaluating T-VEC treatment responses. That said, most responses in our cohort occurred within 6 months, providing a reasonable time frame for clinicians to assess and re-evaluate responses. At present, checkpoint blockade salvage therapies are limited. Lifileucel, an autologous cellular therapy approved for metastatic melanoma that has failed PD-1 inhibitor therapy, has an ORR of 31.4% (5.9% CR) with a median response duration of 36.5 months. While real-world data is not yet available, our data suggests that T-VEC may be a reasonable alternative for patients who are ineligible to receive lifileucel or as a bridge to cellular therapy.33 Conclusion Our findings suggest that T-VEC may provide a viable therapeutic option for patients with advanced melanoma, including those who have experienced failure with other treatments, such as ICIs or targeted therapies. The presence of distant metastatic disease or non-injectable tumors was associated with a lower likelihood of achieving a response, highlighting the potential utility of T-VEC in managing specific subsets of patients with melanoma. While ICIs are generally considered first-line therapy for metastatic melanoma due to their systemic activity and high overall response rates, a subset of patients—particularly those with lower disease burden and predominantly injectable lesions—may be excellent candidates for T-VEC, regardless of prior treatment history. Future research could focus on augmenting oncolytic virus’s effect in non-injected lesions, potentially enhancing its systemic efficacy when combined with other therapies. In addition, further studies to identify biomarkers predictive of response are a high priority to improve patient selection. Overall, our findings support the need for continued exploration of oncolytic virus therapies as an important component of the melanoma treatment landscape. Supplementary material