MR metabolic characterization of locally advanced breast cancer: – treatment effects and prognosis

Abstract

Brystkreft er den hyppigste kreftsykdommen blant kvinner. Lokalavansert brystkreft utgjør omtrent 10% av alle brystkrefttilfeller og omfatter en heterogen pasientgruppe med ulike prognoser. Pasienter med lokalavansert brystkreft får ofte kjemoterapi før kirurgisk fjerning av tumor, såkalt neoadjuvant kjemoterapi, for å redusere størrelsen på tumoren. Det er stor variasjon i behandlingsrespons for denne pasientgruppen, og det er derfor behov for å utvikle målrettet og individualisert behandling, samt metoder for oppfølging av behandlingsrespons.

Metabolomics er en systematisk analyse av småmolekylære forbindelser (metabolitter) i biologiske prøver. Den metabolske profilen til brystkreftvev er vist å korrelere med viktige kliniske parametere, for eksempel tumorgrad, hormonreseptorstatus og lymfeknutespredning. Magnetisk resonans (MR) spektroskopi er en metode som kan gi en detaljert beskrivelse av metabolittprofilen i vevet. En fordel med denne metoden er at vevsprøven er intakt etter analysen slik at den samme vevsprøven kan analyseres videre med andre metoder, for eksempel immunohistokjemi, genuttrykk- eller proteinanalyse.

En sentral gruppe metabolitter innenfor brystkreftforskning er kolinforbindelser. Disse metabolittene er viktig for celledeling, signaloverføring, lipidmetabolisme og cellens membranstruktur. Laktat er en annen viktig metabolitt som inngår i energimetabolismen. I dette arbeidet ble vevsprøver fra pasienter med lokalavansert brystkreft analysert ved bruk av MR spektroskopi for prediksjon av behandlingsrespons og overlevelse. I tillegg undersøkte vi rollen til glycerophosphodiester phosphodiesterase (GDPD) i regulering av kolinforbindelser

Alle pasienter hadde en effekt av behandlingen som ble gitt, og nesten alle fikk en reduksjon i tumorstørrelse etter behandling. Resultatene viste ingen metabolske forskjeller mellom pasienter med klinisk god eller dårlig behandlingsrespons. Resultatene viser derimot at de metabolske forandringene som skjer under neoadjuvant kjemoterapi er forskjellig i pasienter som overlever mer enn fem år og de som dør før fem år. Høyere nivå av laktat, glycine og kolinforbindelser etter behandling var forbundet med dårlig prognose. Analysene av GDPD5 tyder på at enzymet er involvert i reguleringen av kolinmetabolismen, men dets rolle for bruk i målrettet terapi er fortsatt uklart og nye studier må til for å undersøke dette.

MR metabolomics kan brukes til å undersøke metabolske forandringer under neoadjuvant kjemoterapi behandling og kan identifisere viktige metabolitter for prediksjon av overlevelse hos pasienter med lokalavansert brystkreft.

Breast cancer is the most frequent cancer disease among women globally. Locally advanced breast cancer (LABC) constitutes a heterogeneous group of patients with variable prognosis. Today’s treatment decision is predominately based on clinical assessment, histopathological evaluation, and hormone receptor and lymph node status. So far, these data are not sufficient for designing a proper personalized treatment or accurately predicting treatment response and survival. Molecular characterization of tumors may help stratifying patients for individualized treatment, thereby achieving better prognosis.

Magnetic resonance (MR) metabolomics analyses assess the downstream products of gene and protein expressions, i.e. the metabolites, and have shown to provide both predictive and prognostic information for several types of cancers. Proton high resolution magic angle spinning (1H HR MAS) MR spectroscopy is a non-destructive and high-throughput technique that provides highly resolved MR spectra from biological tissue. Recently, altered cell metabolism is suggested as a new emerging hallmark of cancer. Choline phospholipid metabolism is involved in cell signaling, lipid metabolism, and the structural integrity of the cell membrane. Several MRS studies have suggested the total choline-containing metabolite (tCho) level as an in vivo biomarker for diagnosis and treatment evaluation of breast cancer. Reprogramming of energy metabolism and activation of tumor hypoxic response are commonly observed in cancers, and can be characterized by high lactate production.

In this thesis, multivariate data analyses and metabolite quantification of 1H HR MAS MRS data were performed to investigate the potential of metabolomics for prediction of clinical response and long-term survival in LABC patients receiving neoadjuvant chemotherapy (NAC). In addition, the role of glycerophosphodiester phosphodiesterase (GDPD) in choline phospholipid metabolism of human breast cancer was investigated.

All patients had a metabolic response to NAC and almost all patients had a reduction in tumor size. Our results show no clear differences in metabolic responses to NAC between patients with partial response and stable disease and no significant multivariate models for prediction of clinical response by MR metabolomics data. In general, all patients experienced a decrease in tCho levels. It is possible that a cohort including also patients with progressive disease would reveal clearer differences in the metabolite profiles between the clinical response groups. This thesis demonstrates that MR metabolomics contain prognostic information that is associated with survival status of LABC patients. Increase in lactate levels as a response to NAC was associated with low survival rates (< 5 years), while decreased glycine and choline phospholipid metabolites were associated with long-term survival (≥ 5 years). The observed metabolite profiles consisting of higher levels of lactate, glycine, and tCho post-treatment were predictive of low breast cancer survival rates.

GDPD5 gene expression was correlated with choline phospholipid metabolite levels and with CHKA and PLD1 gene expressions suggesting GDPD5 to have a role in regulation of choline phospholipid metabolism in human breast cancer. However, more studies are needed to investigate the relationship between GDPD5 and tumor malignancy, and also estrogen receptor status, for use as target in breast cancer treatment.

In conclusion, monitoring metabolic responses to NAC by MR metabolomics may have the potential to assist the prediction of survival and help identify new targets for therapeutic treatment of breast cancer.