Elsevier

Journal of Proteomics

Volume 76, 5 December 2012, Pages 91-101
Journal of Proteomics

SAA1 is over-expressed in plasma of non small cell lung cancer patients with poor outcome after treatment with epidermal growth factor receptor tyrosine-kinase inhibitors

https://doi.org/10.1016/j.jprot.2012.06.022Get rights and content

Abstract

It has been shown that a proteomic algorithm based on 8 MALDI TOF MS signals obtained from plasma of NSCLC patients treated with EGFR TKIs, is able to predict patients' clinical outcome. In the current study, we identified the proteins originating 4 out of 8 mass signals in the classification algorithm. Plasma samples collected before the beginning of gefitinib therapy were analyzed by MALDI TOF MS and classified according to the proteomic algorithm in good and poor profiles. Two pools of good and poor classified samples were prepared using MARS and ProteoMiner Protein Enrichment kit before 2DE analysis. Proteins differentially expressed between good and poor 2DE samples were excised from gels and analyzed with MALDI TOF MS and LC MS/MS. The identified proteins were validated by Immunodepletion and Western blot analyses. serum amyloid A protein 1 (SAA1), together with its two truncated forms, was over-expressed in plasma of poor classified patients, and was identified as the protein that generates 4 out of the 8 mass signals composing the proteomic algorithm VeriStrat. SAA levels measured by ELISA in 97 NSCLC patients treated with gefitinib correlated with the clinical outcome of the patients. This article is part of a Special Issue entitled: Integrated omics.

Graphical abstract

Highlights

► SAA1 generates 4 out of 8 mass signals composing the predictive algorithm VeriStrat. ► SAA1, together with its two N-terminal truncated forms, is over expressed in NSCLC patients. ► SAA1 plasma levels correlate with the clinical outcome of NSCLC patients. ► SAA1 level may help clinicians to select NSCLC patients who do not respond to EGFR TKIs therapy. ► SAA1 might be related to EGFR TKIs resistance.

Introduction

Lung cancer is the primary cause of cancer-related death in the world [1]. Eighty‐five per cent of lung cancer diagnosis is non small cell lung cancer (NSCLC) [2]. Until 2002, chemotherapy was the main treatment option for NSCLC patients. The introduction of gefitinib and erlotinib brought a significant survival advantage in stage IV NSCLC patients, leading to an overall survival (OS) improvement from 4.9 months to 22 months in selected patients [3], [4]. Both gefitinib and erlotinib are oral drugs that inhibit the tyrosine kinase (TK) domain of the epidermal growth factor receptor (EGFR), which is over-expressed in about 80% of patients with NSCLC [5], [6]. The binding of EGFR TKIs (tyrosine kinase inhibitors) to the EGF receptor blocks EGFR autophosphorylation, thus hindering the signaling pathway activation. This causes cell cycle arrest and apoptosis. Two classes of somatic mutations in the TK domain of the EGFR gene (exon 19 deletions and L858R substitution in exon 21) have been evidenced to have both a predictive and prognostic role in patients treated with EGFR TKIs [7], [8], [9]. Different prospective clinical trials showed that patients carrying EGFR activating mutations have major benefit when treated with EGFR TKIs in terms of response rate (RR) and progression free survival (PFS) vs. either platinum based first line [3], [4], [10], [11], or docetaxel second line chemotherapy [12].

The use of EGFR TKIs in EGFR wild type patients is still a controversial issue. In first line setting, the therapeutic options are clear: unselected and EGFR wild type patients should be treated with chemotherapy [3], while there are no sufficient data that could define the strategy for second line treatment. In unselected and wild type NSCLC population, the efficacy of gefitinib was similar to docetaxel second line chemotherapy as observed in INTEREST, SIGN, and V-15-32 studies [13], [14], [15]. This is probably related to the fact that EGFR TKIs have been compared to platinum doublet in first line, and mono-chemotherapy, less effective, in second line. Moreover, it may be hypothesized that even if the EGFR pathway does not represent the most important driver for tumor cell survival in EGFR wild type patients, it might contribute to tumor cell proliferation, and the inhibition of this pathway might be sufficient to control tumor growth. The crucial point is that for this subgroup of patients, second line therapeutic options are weak and there are no molecular biomarkers able to guide the treatment choice between EGFR TKIs and chemotherapy.

In a previous, retrospective study [16], a proteomic algorithm (VeriStrat), was built and validated in order to classify NSCLC patients according to their outcome after EGFR TKI therapy. Serum and plasma were analyzed by matrix assisted laser desorption ionization time of flight mass spectrometry (MALDI TOF MS) before the beginning of treatment with EGFR TKIs. VeriStrat was built in a training set from 3 cohorts of NSCLC patients treated with gefitinib. Eight mass signals (Supplementary Fig. 1), which distinguished patients with stable disease (SD) longer than 6 months from those with progressive disease (PD) in less than 1 month, were used to create the algorithm. This was subsequently blind tested and validated in 2 independent cohorts of patients treated with gefitinib and erlotinib, and in 3 control cohorts of patients treated with chemotherapy and surgery. Patients were classified by VeriStrat in 2 categories of patients: those with better survival outcome (good), and those with worse survival outcome (poor). It was also shown that the algorithm had a predictive rather than a prognostic value, as no differences in terms of survival were observed between VeriStrat good and VeriStrat poor classified patients treated with chemotherapy or surgery. Moreover, thanks to the pre-processing and normalization spectra procedures, a high reproducibility of the technique was observed, since the spectra were acquired by two independent institutions and a concordance rate of 97.1% was shown [16]. Aiming at proving the stability of the VeriStrat proteomic profile in the course of gefitinib treatment, as well as to find a possible correlation between its changes and disease progression, we applied the VeriStrat classification over the course of gefitinib therapy in 111 metastatic NSCLC patients [17]. Samples were analyzed before the beginning of therapy and every 2 months concomitantly with Computed Tomography scan evaluation until withdrawal from treatment for either toxicity or progression. In the majority of cases, the proteomic classification remained stable during therapy, concordantly to the baseline level, while 28% shifted to a poor profile at withdrawal and, in 90% of these cases, patients had the development of new lesions or an early death. Based on these findings, we may suppose that VeriStrat represents the proteomic signal of EGFR TKI resistance, either primary (in the cases of baseline poor patients), or acquired (in the cases of baseline good patients shifted to poor).

The aim of the current study was to establish the identity of the molecular species generating the VeriStrat algorithm in order to identify a plasma biomarker able to select the group of patients who cannot benefit from EGFR TKIs. Moreover, the identification of the proteins differentially expressed in these patients might shed new lights on the biological basis which discriminates patients' clinical benefit to EGFR TKI therapy.

Section snippets

Sample selection

From 2001 to 2004, advanced NSCLC patients (admitted to the San Raffaele Hospital) were treated with gefitinib. Plasma samples were collected before the beginning of the treatment. The collected blood was centrifuged at 2000 g for 10 min, at 4 °C, and plasma was separated and properly stored at − 80 °C until the analysis. All samples were obtained after patients provided written informed consent; analyses were performed under a protocol approved by the local Institutional Review Boards. To overcome

Identification of the proteins differentially expressed among VeriStrat good and VeriStrat poor classified patients

We performed 2DE on total plasma samples (data not shown) and, as expected, only few abundant plasma proteins were detected. In order to reduce the dynamic concentration range and to analyze an increased number of proteins, two low abundant proteome enrichment strategies were applied based on the ProteoMiner and the Agilent Multiple Affinity Removal System (MARS) [20].

After low abundant protein enrichment with ProteoMiner, 31 spots with a fold change > 1.8 (p-value < 0.05) were highlighted and

Discussion

The VeriStrat proteomic algorithm has been proposed to have a predictive role in the assistance of the therapeutic choice between EGFR TKIs and chemotherapy in second line EGFR wild type NSCLC patients [18]. In the current study, we identified some of the molecular species generating the MALDI-TOF mass signals composing the VeriStrat algorithm. In VeriStrat poor classified patients, we found increased levels of SAA1 and its two truncated forms, concomitantly with a panel of inflammatory

Conclusions

In conclusion, we proved that SAA1 is one of the proteins that generate the VeriStrat algorithm. Our results indicate for the first time that SAA1 might be related to EGFR TKI resistance. We also demonstrated the clinical correlation between SAA levels and the outcome risk of the patients. These findings open new questions about the role of SAA1 and inflammation in NSCLC patients. It is in fact still not clear if SAA1 has a key role in EGFR TKI resistance, or if it is a marker of poor prognosis

Acknowledgments

We would like to thank the patients who took part in the study.

We thank the member of the Biomolecular Mass Spectrometry and Promifa Facility for fruitful discussions and help with the graphics. We would like to acknowledge Biodesix Co (Colorado, USA) that has performed VeriStrat classification and in particular Heinrich Roder and Julia Grigorieva for scientific collaboration. This study was supported by AIRC grant IG 5153 ‘Mass spectrometry to predict outcome for NSCLC patients treated with

References (34)

  • A. Jemal et al.

    Cancer statistics

    CA Cancer J Clin

    (2008)
  • T.S. Mok et al.

    Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma

    N Engl J Med

    (2009)
  • F.A. Shepherd et al.

    Erlotinib in previously treated non small cell lung cancer

    N Engl J Med

    (2005)
  • A. Kumar et al.

    Structure and clinical relevance of the epidermal growth factor receptor in human cancer

    J Clin Oncol

    (2008)
  • S.V. Sharma et al.

    Epidermal growth factor receptor mutations in lung cancer

    Nat Rev Cancer

    (2007)
  • R. Rosell et al.

    Screening for epidermal growth factor receptor mutations in lung cancer

    N Engl J Med

    (2009)
  • M. Maemondo et al.

    Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR

    N Engl J Med

    (2010)
  • Cited by (65)

    • Baseline and early changes in circulating Serum Amyloid A (SAA) predict survival outcomes in advanced non-small cell lung cancer patients treated with Anti-PD-1/PD-L1 monotherapy

      2021, Lung Cancer
      Citation Excerpt :

      Circulating levels of SAA also were independently associated with risk to develop lung cancer [28], gastric cancer [29], or colorectal cancer [26]. For those with established cancer diagnosis, it has shown that SAA was predictive for OS in large patient cohorts of several cancers types (lung cancer [20], nasopharyngeal carcinoma [30], breast cancer [31], renal cell carcinoma [27], gastric cancer [32], and melanoma [33]), irrespective of treatment modality. As investigated, SAA has complex immune-modulating properties.

    • Application of protein set enrichment analysis to correlation of protein functional sets with mass spectral features and multivariate proteomic tests

      2020, Clinical Mass Spectrometry
      Citation Excerpt :

      VeriStrat results were also associated or trended to association (Supplement Table 5) with chronic inflammatory response and IFN type 1-related pathways. Association of VeriStrat with acute phase reaction and inflammatory response was anticipated, considering that one of the major acute phase reactants – SAA – constitutes a part of the VeriStrat signature and is upregulated in patients classified as VeriStrat Poor [34]. Positive correlations of VeriStrat Poor results with inflammatory proteins were also found using a Luminex immunoassay panel by Fidler et al [27].

    • Acute phase reactant serum amyloid A in inflammation and other diseases

      2019, Advances in Clinical Chemistry
      Citation Excerpt :

      Patients with neoplasia who have a serum A-SAA concentration < 10 μg/mL have a significantly greater chance of survival than patients who have higher levels [194]. In nonsmall cell lung cancer patients undergoing EGF receptor tyrosine-kinase inhibitor therapy, the increase in the SAA1 expression levels in plasma serves as an indicator of a poor prognosis [195]. Similarly, a correlation has been established between SAA1 expression and a poor clinical outcome in patients with conventional renal cell carcinoma [196].

    View all citing articles on Scopus

    This article is part of a Special Issue entitled: Integrated omics.

    1

    These authors equally contributed.

    View full text