Melanoma is a malignant tumor of melanocytes, which are neural-crest–derived cells responsible for the production of melanin. Melanoma is the most lethal form of skin cancer, accounting for >75% of skin cancer–related deaths1. Approximately 232,000 new cases of melanoma are diagnosed globally per year, with >70% of these cases diagnosed in Australia, Europe, and North America2. In Asian countries, the incidence of melanoma is relatively low. In 2015, estimates of new cases of melanoma and melanoma-related deaths in China were ~8000 and 3200, respectively3.
In China, nearly 40% of patients with melanoma have advanced disease at the time of diagnosis (stage III, 25.1%; stage IV, 12.8%)4. Five-year survival rates for stages I, II, III, and IV melanoma are 94%, 44%, 38%, and 4.6%, respectively; median survival is 5.00, 4.25, 2.83, and 1.42 years, respectively4. The proportions of patients with mucosal (22.6%) and acral melanoma (41.8%) are higher in China than in Western countries5. For example, mucosal and acral melanomas account for <5% of melanoma cases in the United States6. Prognosis is poor for patients with mucosal melanoma; reported median overall survival (OS) in patients with advanced mucosal melanoma hospitalized in Beijing is 11 months7.
Outside China, steady progress has been made in the development of targeted therapies and immunotherapies for locally advanced and metastatic melanoma. Anticytotoxic T-lymphocyte–associated antigen-4 (CTLA-4) monoclonal antibodies (eg, ipilimumab), the combination of BRAF plus mitogen-activated protein kinase (MEK) inhibitors (eg, dabrafenib and trametinib), and programmed death ligand 1 (PD-L1) inhibitors (eg, nivolumab and pembrolizumab) have been approved for the treatment of locally advanced or metastatic melanoma in >50 countries, including the United States and across Europe. In China, treatment options for unresectable locally advanced or metastatic melanoma have been limited to chemotherapeutic agents, such as dacarbazine.
Available treatment options for Chinese patients, such as cytotoxic chemotherapies, have limited activity even in the first-line setting. According to the Chinese Guidelines on the Diagnosis and Treatment of Melanoma (2015 edition)4, the efficacy of dacarbazine-based regimens is marginal, with an objective response rate (ORR) of 7%, progression-free survival (PFS) of 1.6 months, and OS of 0–6 months. In a multicenter, randomized, double-blind phase 2 trial, the ORR of Chinese patients with advanced mucosal melanoma who were receiving first-line therapy was 3.7% in the dacarbazine arm8.
There is no standard treatment regimen for second-line therapy in China. In a phase 2 trial conducted to evaluate recombinant human endostatin (rhES; Endostar; Shandong Simcere Medgenn Bio-Pharmaceutical Co. Ltd, Nanjing, China) plus dacarbazine in patients with metastatic melanoma, this combination was associated with significant improvement in median PFS (4.5 vs. 1.5 mo; P=0.013) and OS (12.0 vs. 8.0 mo; P=0.005) compared with dacarbazine plus placebo8. The most commonly used second-line regimen in patients with advanced mucosal melanoma who were hospitalized in Beijing was paclitaxel plus carboplatin plus bevacizumab, which was associated with a median PFS of only 2.5 months7. Data reflecting the real-world effectiveness of first-line and second-line therapies in patients with melanoma in China are limited. To address this issue, we conducted a study to describe real-world clinical outcomes in patients with advanced and metastatic melanoma in China.
Study setting and data sources
This was a retrospective, observational cohort study using patient electronic medical records (EMRs) obtained in an academic setting of the Beijing Cancer Hospital (BCH), which treats the highest volume of patients with melanoma locally and nationwide. A melanoma patient pool was identified from EMRs, and eligible patients who initiated treatment between January 1, 2014, and December 31, 2015 (index date) were enrolled according to study inclusion/exclusion criteria. The cohort end date of December 31, 2017, was used to censor survival follow-up.
Medical charts in BCH contain detailed information about patient disease characteristics, therapies, laboratory data, computed tomography, magnetic resonance imaging, and clinical outcomes. This hospital performs regular imaging in patients with melanoma as part of routine care, which was a valuable resource for confirmation of clinical responses in the current study. Generally, clinical responses are assessed according to Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1), and patients are followed up after discharge from the hospital. Trained researchers retrospectively reviewed all available EMRs (eg, medical charts, clinical notes, imaging scans, and follow-up records) to extract relevant information into a structured case report form for each patient who met study inclusion and exclusion criteria. The study protocol (EP05026.036.01) and a waiver of patient consent were approved by the medical ethics committee of BCH (No. 52 Fu-cheng Road, Haidian District, Beijing 1000142, P.R. China) on November 20, 2017. Reporting of data for this study adheres to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) and STROCSS (Strengthening the Reporting of Cohort studies in Surgery) guidelines9,10.
Adult patients with unresectable, locally advanced (stage IIIB or IIIC) or metastatic (stage IV) melanoma were enrolled in this retrospective study based on the following selection criteria.
- Eighteen years or older at the time of initiation of first-line or second-line therapy.
- Histologically confirmed diagnosis of stage IIIB, IIIC, or IV melanoma.
- At least one measurable lesion as defined by RECIST v1.1 on imaging studies (computed tomography or magnetic resonance imaging).
- Initiated either systemic first-line or second-line anticancer therapy in the selected hospital between January 1, 2014, and December 31, 2015 (adjuvant therapy was not considered first-line).
- Documented response status data [ie, complete response (CR), partial response (PR), progressive disease (PD), and stable disease (SD)] available in the EMRs according to RECIST v1.1.
- Diagnosis of uveal or ocular melanoma.
- Immunotherapy between January 1, 2014, and December 31, 2015 (adjuvant treatments such as interferon were allowed).
- Enrollment as a clinical trial participant for any melanoma therapies between January 1, 2014, and December 31, 2015.
- Diagnosis of human immunodeficiency virus or active hepatitis B or C infection at the time of treatment under evaluation.
Primary objectives were to describe the demographic and disease characteristics of patients with unresectable locally advanced and metastatic melanoma at the time of initiation of first-line or second-line anticancer therapies and to describe the systematic treatment pattern and estimate the real-world ORR per RECIST v1.1 by first-line or second-line therapy in these patients. Exploratory objectives included duration of response (DOR), disease control rate (DCR), time to progression (TTP), PFS per RECIST v1.1, and OS by first-line or second-line therapy.
Patients received first-line or second-line anticancer therapy during routine care. This information was extracted from EMRs. Outcomes of interest included real-world ORR, DCR, DOR, time to response, PFS, TTP, and OS. Real-world ORR was defined as the proportion of patients in the analysis population who had CR (disappearance of all target lesions and any pathologic lymph nodes) or PR (a reduction in the sum of diameters of the target lesions of ≥30% from baseline).
Assessment of preliminary response was conducted by the investigator handling the chart review and based on radiography at baseline (within 4 wk before treatment) and radiography during treatment and was confirmed by subsequent radiographic assessment at least 4 weeks from the first date of documentation of response. The principal investigator then reviewed this decision; in the case of a disagreement, the principal investigator rereviewed the original records/images before making a final judgement. DCR was defined as the proportion of patients in the analysis population who had CR, PR (decrease in the sum of diameters of target lesions of ≥30% from the baseline sum of diameters), SD (neither PR nor PD). Patients with missing data were considered to have PD. DOR was defined as the time from first documented evidence of CR or PR until PD (increase in the sum of diameters of target lesions of ≥20% from the smallest sum on study, including baseline) or death for patients who demonstrated CR or PR. Patients who died without documented PD were counted as event (PD) at the death date. Response duration for patients whose disease had not progressed or who had not died by the time of analysis was censored. PFS was defined as the time from the first day of treatment to the first documented PD or death from any cause, whichever occurred first. Those who did not have an event during follow-up were censored. TTP was defined as the time from the first day of treatment to the first documented PD (death not included per RECIST v1.1). Those who did not have an event during follow-up were censored. OS was defined as the time from the first day of treatment to death from any cause. Patients whose death was not document by the time of the final analysis were censored.
Outcomes were stratified according to first-line and second-line regimens. Evaluations of CR, PR, PD, and SD were consistent with RECIST v1.111. Best overall response was assigned.
ORR point estimates and 95% confidence intervals (CIs) were provided using the binomial exact method. For the TTP, DOR, PFS, and OS end points, Kaplan-Meier curves and median estimates of time from the Kaplan-Meier curves were provided as appropriate. SAS 9.4 software (SAS Institute, Cary, NC) was used for data. Descriptive analysis was applied to demographic and disease characteristics and to treatment pattern, yielding summary statistics. The estimated sample sizes to allow estimation of ORR with acceptable precision are presented in Table1.
In total, 248 eligible adult patients with unresectable locally advanced (stage IIIB or IIIC) or metastatic (stage IV) melanoma were identified in the EMRs of BCH from January 2014 to December 2015. Twenty-eight provinces in China were represented in the study population. Approximately 70% of patients were diagnosed with acral or mucosal histology; the remaining patients were categorized as nonchronically sun-damaged or chronically sun-damaged subtypes. Approximately 95% of patients had stage IV melanoma; the proportions of patients with stage M1a, M1b, and M1c were 18.9%, 20.4%, and 54.9%, respectively. Approximately 5% of patients had brain metastases and ~40% had liver metastases. Approximately 80% of patients had elevated lactate dehydrogenase levels before receiving anticancer treatment (86.9% for first-line; 75.0% for second-line).
Baseline patient characteristics are summarized in Table2. Of the 248 patients included in the study, 221 received first-line therapy, 116 received second-line therapy (they received first-line therapy in other hospitals), and 89 received both first- and second-line therapy at BCH. In patients receiving first-line therapy, Eastern Cooperative Oncology Group (ECOG) performance status was 0, 1, or >1, in 29.9%, 51.1%, and 5.0%, respectively. The median tumor size at baseline in first-line patients was 42.0 mm. Before enrollment at BCH, 19.5% of first-line patients received chemotherapy and 59.3% underwent surgery. In patients who received second-line therapy, ECOG performance status was 0, 1, or >1, in 31.9%, 38.8%, and 5.2%, respectively. Median tumor size at baseline in second-line patients was 49.5 mm. Before enrollment at BCH, 91.4% of second-line patients had received chemotherapy and 59.5% of patients had undergone surgery.
The most commonly used first-line regimens were combination therapies: dacarbazine plus cisplatin plus rhES (36.7%) and paclitaxel plus carboplatin plus bevacizumab (22.2%). The most commonly used second-line regimens were paclitaxel albumin plus carboplatin plus bevacizumab (22.4%), paclitaxel plus carboplatin plus rhES (15.5%), and paclitaxel albumin plus cisplatin plus rhES (12.1%). Other regimens are shown in Table3.
Treatment duration for first-line therapies ranged from 0.1 to 25.4 months; treatment duration was 3 months or more in ~40% of patients and was 6 months or more in 9.5% of patients. Treatment duration for second-line therapies ranged from 0.5 to 15.3 months; fewer patients undergoing second-line therapies had a treatment duration of 3 months or more (27.6%) than in those receiving first-line therapy (40.7%), although the proportions of patients with a treatment duration of 6 months or more was the same for first-line and second-line therapy, at 9.5%.
As of December 31, 2017, median follow-up was 14.7 months. All patients were followed-up until end of study. Of the 248 patients included in the study, 195 (78.6%) had died. In the overall study population, median OS was 10.5 months (95% CI, 9.4–12.1 mo), and the 12-month OS rate was 43.9%. Clinical outcomes in patients receiving first-line therapies are summarized in Table4. Among the 221 patients who received first-line therapy, 2 confirmed CRs and 12 confirmed PRs were reported, confirmed by follow-up imaging, resulting in an ORR of 6.3% (95% CI, 3.5%–10.4%). The median DOR of the 14 confirmed cases of CR or PR was 9.1 months (range, 1.7–28.4 mo), and DCR was 68.3% (95% CI, 61.8%–74.4%). Median PFS was 3.5 months (95% CI, 2.9–4.2 mo) and the 12-month PFS rate was 10.6%. Median TTP was 3.5 months (95% CI, 2.9–4.2 mo). As of December 31, 2017, a total of 171 (77.4%) deaths had occurred; median OS was 10.5 months (95% CI, 9.2–12.1 mo) and the 12-month OS rate was 43.5%.
Clinical outcomes in patients receiving second-line therapies are summarized in Table4. There were 4 confirmed cases of PR, substantiated by follow-up imaging, among the 116 patients who received second-line therapy, resulting in an ORR of 3.4% (95% CI, 0.9%–8.6%). The median DOR of the 4 confirmed cases of PR was 7.5 months (range, 4.6–24.2 mo), and the DCR was 57.8% (95% CI, 48.2%–66.9%). Median PFS was 2.3 months (95% CI, 2.0–3.0 mo), and the 12-month PFS rate was 5.2%. Median TTP was 2.3 months (95% CI, 2.0–3.0 mo). As of December 31, 2017, a total of 101 (87.1%) deaths had occurred; median OS was 7.5 months (95% CI, 6.5–8.7 mo) and the 12-month OS rate was 30.5%. Compared with patients receiving first-line therapy, those receiving second-line therapy had shorter median OS (7.5 vs. 10.5 mo) and median PFS (2.3 vs. 3.5 mo) (Table4), as was apparent from the Kaplan-Meier estimates of OS (Fig.1) and PFS (Fig.2) for first-line and second-line therapy.
Subgroup analyses of clinical outcomes (ORR, PFS, and OS) by treatment line and by different histologic groups (acral, mucosal, nonchronically sun-damaged, and chronically sun-damaged) are summarized in Table5. For first-line patients, ORR in the acral and mucosal subgroups was 6.0% and 6.3%, respectively, and was 15.4% in the chronically sun-damaged subgroup. Median PFS in the acral and mucosal subgroups was 3.3 months and 4.1 months, respectively, and was 5.7 months in the chronically sun-damaged subgroup. Median OS in the acral and mucosal subgroups was 10.5 months and 10.3 months, respectively, and 13.1 months in the chronically sun-damaged subgroup. No statistical testing was conducted to analyze these differences. Similar trends were observed in second-line patients (Table5); estimates of clinical outcomes between the various subgroups were similar, although the ORR in the chronically sun-damaged histologic subgroup seemed to be higher than those reported for other subgroups (7.1% vs. 2.4%–4.2%).
To our knowledge, this is one of the few real-world studies to report the most commonly used anticancer regimens and associated clinical outcomes in patients with advanced melanoma in China. As previously noted, BCH treats patients with melanoma from all over China, including rural and urban areas. Most patients received combination regimens for first-line and second-line therapies; the most common regimens were dacarbazine plus cisplatin plus rhES for first-line therapy and paclitaxel albumin plus carboplatin plus bevacizumab for second-line therapy. These therapies seem to be associated with suboptimal clinical outcomes in advanced melanoma, with low rates of tumor response (6.3% and 3.4%) and limited survival rates (12-mo PFS, 10.6% and 5.2%; 12-mo OS 43.5% and 30.5%) in the real-world setting.
Although commonly considered therapies for advanced melanoma, dacarbazine-based regimens were reported to have limited clinical efficacy. In a randomized, double-blind clinical trial of first-line dacarbazine therapy in Chinese patients, the ORR was 3.7%, median PFS was 1.5 months, and median OS was 8 months8. Better outcomes were observed for first-line therapy in the current study, which may be at least in part attributable to the use of combination therapies by most patients; however, these clinical outcomes in a real-world setting still seem suboptimal, with an ORR of 6.3%, a median PFS of 3.5 months, and a median OS of 10.5 months for first-line therapy. Second-line versus first-line patients treated with combination chemotherapy tended to have lower ORR and shorter OS, PFS, and TTP; DOR was short for first-line and second-line therapies. The high proportion of patients with acral or mucosal subtypes in China may be an important reason for the poor outcomes because these patients tended to have relatively late-stage disease at diagnosis and a more aggressive natural history of disease, were more often at M1c stage, and had elevated lactate dehydrogenase levels.
Subgroup analysis of clinical outcomes by histologic subtypes noted differences between chronically sun-damaged and other subtypes in the first-line setting. However, caution should be used when interpreting these findings given the wide CIs and small sample sizes for the different subtypes.
Data generated from this study provide evidence of the real-world effectiveness of anticancer treatments in the first-line and second-line settings for advanced melanoma in China; however, the study has several limitations. A high proportion of data were missing for some variables of interest, such as ECOG performance status (14.0% of first-line patients and 24.1% of second-line patients). Approximately 6% of second-line patients could not be assessed for ORR because of insufficient information in the EMRs. Variability in follow-up schedules among patients may have had an impact on the precision, interpretation, and generalizability of some of the clinical data. However, to increase the reliability of clinical end-point assessments, each CR/PR case was adjudicated by additional reviewers and escalated for further confirmation by the principal investigator, when needed.
The findings of this study provide insight into the treatment patterns and associated clinical outcomes in patients with advanced melanoma in China. Both first-line and second-line anticancer therapies seem to be associated with suboptimal clinical outcomes in patients with advanced melanoma. These findings demonstrate a high unmet medical need in patients with advanced melanoma patients in China.
A waiver of patient consent was applied for and approved by the medical ethics committee of BCH (Address: No.52 Fu-cheng Road, Haidian District, Beijing 1000142, P.R. China) on November 20, 2017.
Sources of funding
This work was supported by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co. Inc., Kenilworth, NJ.
C.C., X.Y., S. Liu, A.C.D., J.L., J. Ge, H.W., and J. Guo: conceived, designed, or planned the study. C.C., X.Y., L.S., Z.C., X.S., B.L., X.W., L.M., B.T., L.Z., X.B., S. Li, and J. Guo: acquired the data. C.C., X.Y., B.L., and J. Guo: analyzed the data. C.C., X.Y., J.L., H.W., and J. Guo: interpreted the results. C.C., X.Y., S. Liu, and J. Guo: drafted the manuscript with contributions from all authors. C.C., X.Y., A.C.D., L.S., Z.C., X.S., B.L., J.L., J. Ge, X.W., L.M., B.T., L.Z., X.B., S. Li, B.L., H.W., and J. Guo: critically reviewed or revised the manuscript for important intellectual content. B.L., and H.W.: provided statistical expertise.
Conflict of interest disclosure
A.C.D. reports employment at Merck Sharp & Dohme Corp., a subsidiary of Merck & Co. Inc., Kenilworth, NJ. J.L., J.Ge., B.L., and H.W. report employment at MSD Beijing, China. S. Liu reports employment at MSD Beijing, China, and holds stock in the company. The remaining authors declare that they have no financial conflict of interest with regard to the content of this report.
Research registration unique identifying number (UIN)
Medical writing and/or editorial assistance was provided by Doyel Mitra, PhD, of the ApotheCom pembrolizumab team (Yardley, PA, USA). This assistance was funded by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co. Inc., Kenilworth, NJ.
1. Shoo BA, Kashani-Sabet M. Melanoma
arising in African-, Asian-, Latino- and Native-American populations. Semin Cutan Med Surg 2009;28:96–102.
2. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015;136:E359–E386.
3. Chen W, Zheng R, Baade PD, et al. Cancer statistics in China
, 2015. CA Cancer J Clin 2016;66:115–32.
4. Guo J, Qin S, Liang J, et al. Chinese guidelines on the diagnosis and treatment
(2015 edition). Chin Clin Oncol 2016;5:57.
5. Chi Z, Li S, Sheng X, et al. Clinical presentation, histology, and prognoses of malignant melanoma
in ethnic Chinese: a study of 522 consecutive cases. BMC Cancer 2011;11:85.
6. Merkel EA, Gerami P. Malignant melanoma
of sun-protected sites: a review of clinical, histological, and molecular features. Lab Invest 2017;97:630–5.
7. Lian B, Zheng N, Chi A. Curative effect of the 1st line and the 2nd line therapies for 232 patients with advanced mucosal melanoma
and factors effecting prognosis of the patients. Chin J Cancer Biother 2017;24:253–8.
8. Cui C, Mao L, Chi Z, et al. A phase II, randomized, double-blind, placebo-controlled multicenter trial of Endostar in patients with metastatic melanoma
. Mol Ther 2013;21:1456–63.
9. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 2007;4:e296.
10. Agha RA, Borrelli MR, Vella-Baldacchino M, et al. The STROCSS statement: strengthening the reporting of cohort studies in surgery. Int J Surg 2017;46:198–202.
11. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Can 2009;45:228–47.