LDH correlation with survival in advanced melanoma from two large, randomised trials (Oblimersen GM301 and EORTC 18951)

https://doi.org/10.1016/j.ejca.2009.04.016Get rights and content

Abstract

Purpose

In a randomised study (GM301; dacarbazine with/without oblimersen), patients with advanced melanoma were stratified based on performance status, metastatic site and lactate dehydrogenase (LDH). Progression-free survival and response and durable response rates showed a highly significant difference favouring dacarbazine–oblimersen and a nearly significant survival difference. All efficacy parameters significantly favoured dacarbazine–oblimersen in patients with normal baseline LDH [⩽1.1 × upper limit of normal (ULN)]. Each stratification factor was assessed for an interaction with treatment on survival and an interaction was detected only for LDH.

Experimental design

Baseline LDH values in Study GM301 treatment groups were combined and analysed using cutoffs above and below 1 × ULN. Baseline LDH in EORTC study 18951 (dacarbazine, cisplatin, interferon-alfa-2b with/without interleukin-2 in advanced melanoma) was independently analysed using the same approach. In Study GM301, the relation between treatment effect and LDH, treatment effect and tumour size, LDH and tumour size and LDH and disease site were determined.

Results

In Study GM301 (N = 760) and Study 18951 (N = 325), LDH was within the upper range of normal for a large number of patients. This was not exhibited in the general population, suggesting such values may be elevated rather than normal in melanoma. A highly ordered and monotonic relationship was apparent between LDH and survival: survival worsened as LDH became more elevated, even when LDH remained within normal range. LDH and tumour size were poorly correlated; elevated LDH was not associated with any one disease site. LDH was highly predictive of oblimersen effect.

Conclusion

In designing studies, LDH should be considered, regardless of tumour size or disease site.

Introduction

An enhanced understanding of key prognostic variables in melanoma has emerged in the last 10 years, primarily due to new biologic techniques and large-scale, population-based analyses.1 As a result, clinical trials now often include risk factors for stratification and identification of patient subpopulations that may achieve maximum therapeutic benefit.2, 3, 4

Serum lactate dehydrogenase (LDH) level is one of the most useful independent prognostic factors in metastatic melanoma.1 Numerous studies investigating various prognostic factors confirm the correlation between increased LDH and decreased survival among patients with advanced melanoma.5, 6, 7, 8, 9, 10, 11, 12, 13 This association is evident even after accounting for the number of metastases and location of distant metastatic involvement.14, 15

Solid tumours require a blood supply to grow and metastasise and are highly adaptable in meeting this need, as well as thriving in a hypoxic state. Increased serum LDH is believed to be related to the hypoxic environment of tumour cells. Unlike normal cells that produce the majority of ATP from glucose through oxidative phosphorylation, many cancer cells produce ATP by converting glucose to lactate,16 a process essential to the hypoxic environment in which many cancer cells exist. LDH is composed of 2 peptides – LDH-A and LDH-B. It is LDH-A (also known as LDH-5 in its tetrametric form) that is upregulated in the hypoxic environment. This enzyme catalyses the conversion of pyruvate to lactate, probably via the activity of the HIF1α transcription factor.17, 18 As LDH is not a secreted enzyme, the finding of elevated LDH in the serum of patients with advanced melanoma is probably due, at least in part, to melanoma-cell necrosis or apo-necrosis and the spillage of LDH, which likely occurs when part of a tumour outgrows its blood supply. The remaining poorly perfused, hypoxic tumour cells may be highly refractory to cytotoxic chemotherapy in light of abnormalities in their vasculature and microenvironment,19 and they may eventually recover and proliferate.

Little progress in the treatment of advanced melanoma has been made over the last several decades. The standard therapy remains dacarbazine. Various agents, including cytotoxic drugs, interferon-α and IL-2 and tamoxifen, have been used in combination with dacarbazine in clinical studies.20 Nevertheless, randomised trials of single-agent dacarbazine versus dacarbazine-containing combination regimens have failed to show a survival benefit with the multiple-drug regimens.20

The largest randomised study in advanced melanoma conducted to date is a multinational, Phase 3 trial (Study GM301) in 771 patients, including 760 for whom baseline serum LDH was recorded. This study was performed to evaluate the potential of the antisense oligonucleotide oblimersen to enhance the efficacy of dacarbazine in a chemotherapy-naïve population.21 Prior to randomisation, patients were stratified according to 3 prognostic factors: Eastern Cooperative Oncology Group performance status (0 versus 1 or 2), liver metastases (present versus absent) and disease site in combination with baseline LDH (non-visceral disease and normal LDH, defined as a baseline serum LDH level  1.1 times the upper limit of normal [×ULN], versus visceral disease [excluding liver metastases] or elevated LDH, defined as a baseline serum LDH level >1.1 × ULN). Every 3 weeks patients received either dacarbazine 1000 mg/m2 intravenously (i.v.) or the same dacarbazine dose preceded by a 5-d continuous intravenous infusion of oblimersen 7 mg/kg/d. Patients were assessed by RECIST22 at the end of every two cycles.

Overall survival (the primary end-point) showed an improvement in the oblimersen–dacarbazine group at 24-month minimum follow-up that approached statistical significance (median 9.0 months versus 7.8 months in the dacarbazine group; P = 0.077). Other efficacy end-points consistently demonstrated a highly significant between-treatment difference favouring the oblimersen–dacarbazine regimen, including progression-free survival (median 2.6 months versus 1.6 months; P < 0.001), overall response rate (13.5% versus 7.5%; P = 0.007) and durable response rate (7.3% versus 3.6%; P = 0.03).21 Particularly noteworthy was a separation of the survival curves beginning at about 6 months, which suggested the existence of a heterogeneous population. Thus, we assessed whether there was a subset of patients who derived increased benefit from the addition of oblimersen to dacarbazine.

Each of the three stratification factors in this study (performance status, metastatic site and LDH) was assessed for an interaction with treatment on survival. An interaction was detected only between LDH and treatment.21 Given the heterogeneity of the population, treatment effect was then analysed in both the normal and elevated LDH subpopulations. The normal LDH population included 508 of the 760 patients, representing a sample size greater than that used in any previous randomised study in advanced melanoma.21 Statistically significantly superior results were observed for multiple efficacy end-points in patients with normal LDH who were treated with the oblimersen–dacarbazine regimen, including overall survival (median 11.4 months versus 9.7 months; P = 0.02), progression-free survival (median 3.1 months versus 1.6 months; P < 0.001), overall response rate (17.2% versus 9.3%; P = 0.009) and durable response rate (9.6% versus 4.0%; P = 0.01). In contrast, the elevated LDH population (N = 252) was neutral with respect to treatment effect; no significant between-treatment differences in efficacy parameters were apparent.21

Per protocol, the cutoff for defining normal LDH and elevated LDH had been established as 1.1 × ULN. Given the interaction between baseline LDH and treatment, two important questions arose: (1) Would a different cutoff to define ‘normal LDH’ have led to different conclusions? and (2) What is the medical relevance of this observation? This article attempts to provide answers to these two questions, as well as the medical basis for our findings.

Section snippets

Patients and methods

LDH values from both treatment groups in Study GM301 were combined, and the distribution of these values was compared to the distribution in the general population (i.e. expected Gaussian distribution) from whom the normal reference range is established.

LDH cutoffs above and below 1 × ULN were used to assess the robustness of the definition of normal LDH.23 Five LDH categories were considered: ⩽0.8 × ULN, >0.8 to ⩽ 1.1 × ULN (cutoff point used for stratification in Study GM301), >1.1 to ⩽2 × ULN (cutoff

Results

With a Gaussian distribution, one would expect 50% of LDH values to fall below the median. For the 760 patients in Study GM301, the median midpoint value between the upper and lower limits of the normal range was 0.75, but the median ratio of baseline LDH to the upper limit of normal range was 0.9, demonstrating a shift to the right in the median by approximately 20%. In fact, in Study GM301, one-third of the values fell between 0.8 and 1.1 × ULN, one-third fell below 0.8 × ULN and one-third fell

Discussion

Elevated serum baseline LDH is widely recognised as an important prognostic factor in melanoma. As with all laboratory values, the normal limits for LDH are derived from the general healthy population. Based on our findings, the applicability of the upper limit of the normal range is questionable in the setting of advanced melanoma. The overrepresentation of values close to but not exceeding the upper limit of normal implies that these values may actually be elevated rather than normal in a

Conclusions

The consistency in efficacy demonstrated in Study GM301 has not previously been observed in any other completed, randomised study in advanced melanoma. In this study, low baseline LDH (⩽0.8 × ULN) was found to be predictive of the treatment effect of oblimersen in combination with dacarbazine. A confirmatory study is ongoing. Whether the low LDH cutoff point of 0.8 × ULN is disease- or drug-specific warrants exploration.

In the setting of advanced melanoma, baseline serum LDH is not a simple marker

Trial registration numbers

Study GM301 – NCT00016263; EORTC Study 18951 – NCT00002669

Conflict of interest statement

Sanjiv S. Agarwala – none declared.

Ulrich Keilholz – none declared.

Erard Gilles – employment (Genta; compensated).

Agop Y. Bedikian – none declared.

Jane Wu – employment (Genta; compensated).

Richard Kay – consultant (Genta Incorporated; compensated).

Cy A. Stein – consultant (Genta; uncompensated); honoraria (Genta); expert testimony (Genta, compensated).

Loretta M. Itri – employment (Genta; compensated); stock ownership (Genta).

Stefan Suciu – none declared.

Alexander M.M. Eggermont – consultant

Acknowledgements

Study GM301 was sponsored by Genta Incorporated. EORTC Study 18951 was supported by Chiron BV, Schering Plough and the National Cancer Institute. We thank Janet Ehlert and Steven Novick (both of Genta Incorporated) for writing assistance.

References (26)

  • A. Franzke et al.

    Elevated pretreatment serum levels of soluble vascular cell adhesion molecule 1 and lactate dehydrogenase as predictors of survival in cutaneous metastatic malignant melanoma

    Br J Cancer

    (1998)
  • T.M. Proebstle et al.

    Long-term outcome of treatment with dacarbazine, cisplatin, interferon-α, and intravenous high dose interleukin-2 in poor risk melanoma patients

    Melanoma Res

    (1998)
  • J. Manola et al.

    Prognostic factors in metastatic melanoma: a pooled analysis of Eastern Cooperative Oncology Group Trials

    J Clin Oncol

    (2000)
  • Cited by (124)

    • Synergizing radiotherapy and immunotherapy: Current challenges and strategies for optimization

      2023, Neoplasia (United States)
      Citation Excerpt :

      Importantly, tumor burden is a critical determinant of the response to ICI. A number of approaches have been used to quantify tumor burden, including CT and PET [42] imaging, circulating biomarkers such as ctDNA [42–44], and serum lactate dehydrogenase (LDH) [42,45,46]. In general, a higher tumor burden is associated with inferior clinical outcomes following ICI.

    • Skin cancer biology and its biomarkers: Recent trends and prospective

      2023, Biomarkers in Cancer Detection and Monitoring of Therapeutics: Diagnostic and Therapeutic Applications: Volume 2
    • High serum LDH and liver metastases are the dominant predictors of primary cancer resistance to anti-PD(L)1 immunotherapy

      2022, European Journal of Cancer
      Citation Excerpt :

      Interestingly, several prognostic scoring systems (e.g. Royal Marsden Hospital score, AJCC melanoma staging and classification) use one of these four correlated baseline medical data [24–26]. Elevated LDH is a known poor prognostic marker of cancer and is associated with high tumour burden, intratumour hypoxia, and lactic acid production by cancer cells [12,27]. Accumulated lactic acid blunts tumour immunosurveillance by inhibiting the anti-tumour activity of tumour-infiltrating T cells and NK cells and, at high concentrations, triggers their apoptosis [28–31].

    View all citing articles on Scopus

    Presented in part at the 2007 annual meeting of the American Society of Clinical Oncology and at the 3rd annual International Melanoma Research Congress in 2006.

    View full text