UBC Faculty Research and Publications

The maximum standardized uptake value of 18 F-FDG PET scan to determine prognosis of hormone-receptor… Zhang, Jian; Jia, Zhen; Ragaz, Joseph; Zhang, Ying-Jian; Zhou, Min; Zhang, Yong-Ping; Li, Gang; Wang, Bi-Yun; Wang, Zhong-Hua; Hu, Xi-Chun Jan 31, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12885_2012_Article_3693.pdf [ 327.18kB ]
Metadata
JSON: 52383-1.0221466.json
JSON-LD: 52383-1.0221466-ld.json
RDF/XML (Pretty): 52383-1.0221466-rdf.xml
RDF/JSON: 52383-1.0221466-rdf.json
Turtle: 52383-1.0221466-turtle.txt
N-Triples: 52383-1.0221466-rdf-ntriples.txt
Original Record: 52383-1.0221466-source.json
Full Text
52383-1.0221466-fulltext.txt
Citation
52383-1.0221466.ris

Full Text

RESEARCH ARTICLE Open AccessThe maximum standardized uptake value of18 F-FDG PET scan to determine prognosis ofhormone-receptor positive metastatic breastcancerJian Zhang1†, Zhen Jia1†, Joseph Ragaz2, Ying-Jian Zhang3, Min Zhou3, Yong-Ping Zhang3, Gang Li4, Bi-Yun Wang1,Zhong-Hua Wang1 and Xi-Chun Hu1*AbstractBackground: Whether PET scan maximum standard uptake value (SUVmax) could differentiate luminal A fromluminal B and help predict the survival of metastatic breast cancer (MBC) patients with luminal subtype is stillunknown and need to be investigated.Methods: 305 MBC patients with luminal subtypes were screened with PET/CT. Eligible patients were prospectivelyfollowed up.Results: In total, 134 patients were eligible for this study. SUVmax was significantly related to the number of metastaticsites and presence of visceral metastasis on univariate analysis. SUVmax could not effectively differentiate patients withluminal A from luminal B subtype. Although luminal subtype at diagnosis could predict the relapse-free interval, itcould not predict progression-free survival (PFS) or overall survival (OS) after developing relapse. In contrast, SUVmaxwas predictive of both PFS and OS and this effect was maintained in multivariate COX regression model.Conclusions: SUVmax of MBC did not correlate with molecular subtypes of primary tumor. While molecular subtypemay be a valuable prognostic factor at primary diagnosis of breast cancer, the SUVmax, rather than molecular subtype,does have a potential to predict independently in multivariate analysis for the PFS and OS in patients with metastaticdisease of luminal subtype.Keywords: Metastatic breast cancer, Luminal subtype, PET/CT, SUVmax, PrognosisBackgroundBreast cancer is the most common female cancer. Itaffects almost 1.4 million women worldwide and about459,000 patients die due to this disease every year [1].Approximately 6% of women with breast cancer havemetastatic disease at the time of diagnosis and about20% of patients initially diagnosed with localized diseasewill develop metastatic breast cancer (MBC) [2]. Despitesignificant improvements in the treatment of MBCduring the last decade, it remains an incurable disease,with a median life expectancy of 18–30 months [3].Hormone receptors (HR), estrogen receptor (ER) andprogesterone receptor (PgR), play important roles inbreast cancer development, progression and response totherapy. The traditional classification of breast cancersinto HR-positive and -negative groups helps to guidepatient management. However, despite appropriate endo-crine therapy, some HR-positive tumors recur and/or be-come metastatic. Microarray gene expression analysis(cDNA) has identified two biologically distinct HR-positive subtypes of breast cancer with significantdifferences in patient outcome: luminal A and luminal B[4]. However, cDNA analysis is too complex and costlyand thus not routinely performed to identify breast cancer* Correspondence: huxicun@gmail.com†Equal contributors1Department of Medical Oncology, Department of Oncology, FudanUniversity Shanghai Cancer Center; Shanghai Medical College, FudanUniversity, Shanghai, ChinaFull list of author information is available at the end of the article© 2013 Zhang et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Zhang et al. BMC Cancer 2013, 13:42http://www.biomedcentral.com/1471-2407/13/42subtypes. A clinically relevant subtype classification can beobtained by immunohistochemical (IHC) analysis of thetumor expression of ER, PgR, HER2 or Ki67 [5]. IHCcould also classify two categories of luminal subtypes: lu-minal A (ER and/or PgR-positive, HER2-negative), and lu-minal B (ER and/or PgR-positive, HER2-positive) [5].However, compared to cDNA array, the IHC testing doesnot identify all the luminal B tumors because only 30% to50% are HER2-positive on IHC. Thus, many luminal Btumors on cDNA array would be classified as luminal Aon IHC. In 2009, Cheang et al. modified the IHC defin-ition and found that Ki67 could distinguish on IHC theluminal A versus B subtype more accurately, with theKi67 index cut point of 13.25% [6]. The luminal A subtypewas then defined as HR-positive, HER2-negative breastcancer with Ki67 index < 14%, while luminal B subtypewas defined as also HR-positive, but either HER2-positive,or HER2-negative with Ki67 index ≥ 14%. Compared withluminal A tumors, luminal B tumors have thus higherproliferation and poorer outcomes despite being clinicallyHR-positive. Consequently, the major biological distinc-tion between luminal A and B is the proliferation sig-nature, which includes genes such as CCNB1, MKI67, andMYBL2, with higher expression in luminal B than inluminal A tumors and may be important to breast cancerbiology and prognosis [7,8].Positron emission tomography (PET), using the radio-labeled glucose analog 18 F-fluorodeoxyglucose (18 F-FDG),can detect enhanced glycolysis of cancer cells and hasproven valuable in diagnosing, staging, detecting recur-rences, and assessing response to therapy in a multitude ofmalignant disorders [9]. Since 18 F-FDG uptake in cancerusually indicates the degree of tumor proliferation andmetabolism, it was felt important to evaluate whether PETcould be used as a noninvasive diagnostic modality to dif-ferentiate luminal A from luminal B tumors and hencepredicting their behavior and prognosis. The standardizeduptake value (SUV) is a semi-quantitative simplified meas-urement of the tissue FDG accumulation rate, and studiesof the head and neck, lung, esophageal, endometrial,cervical and renal cell cancer have explored the prognosticsignificance of the maximum standardized uptake value(SUVmax) [10-16]. However, the role of the BaselineSUVmax as a prognostic factor for treatment naïve MBCpatients of luminal subtype has not yet been evaluated so far.The main objective of this study was to determinewhether Baseline SUVmax in MBC patients correlateswith validated prognostic markers and their luminalsubtypes, and to establish whether the Baseline SUVmaxcould be used as a noninvasive indicator to differentiateluminal A from luminal B subtypes. In addition, wealso prospectively investigated the impact of BaselineSUVmax on the survival of MBC patients with luminalsubtype.MethodsStudy design and patient populationBetween February 2007 until December 2010, a total of305 MBC patients with luminal subtypes (HR-positive)signed consent for this study and underwent PET/CTexaminations. Baseline information collected, includingPET/CT results, was then used to evaluate whether theindividual was eligible for the study according to inclu-sion and exclusion criteria (Figure 1). Eligible patientsFemale MBC patients with luminal subtypes (n = 305)2007.2 ~ 2010.12ScreeningExcluded (n = 152)29 with history of diabetes mellitus8 with a second primary cancer diagnosed12 with ECOG 3 and life expectancy < 3 months57 with previous treatment in metastatic setting46 with no evaluable lesions documented by abnormal FDG uptakeWithdrew the consent (n = 19)Eligible patients (n= 134)Signed informed consent, underwent PET/CT and baseline information was gatheredFigure 1 Patient screening and inclusion diagram.Zhang et al. BMC Cancer 2013, 13:42 Page 2 of 11http://www.biomedcentral.com/1471-2407/13/42were prospectively followed up at two-month intervalsin Fudan University Shanghai Cancer Center (FUSCC),Shanghai, China. This study was approved by the FUSCCinstitutional ethical review board.Originally we defined the luminal subtypes as definedby Carey et al. [5]. After 2009, we changed the diagnosticcriteria and reviewed all the paraffin sections before.The new criteria were described according to ER, PgR,HER2 and Ki67 status [6]. We defined HR-positive,HER2-negative and Ki67 index <14% as luminal A, HR-positive and HER2-positive (or HER2-negative with Ki67index ≥14%) as luminal B. Her-2/Neu status was definedpositive when over-expressed with 3 plus staining in IHCor amplified with a ratio > 2.2 by fluorescence in situhybridization (FISH). Ki67 was visually scored for percent-age of tumor cell nuclei with positive immunostainingabove the background level by two pathologists.Criteria for inclusion were as follows: female gender, 18 to70years of age, histologically confirmed breast cancer with lu-minal subtypes, eastern cooperative oncology group (ECOG)performance status of 0 to 2, life expectancy of > 3 months,written informed consent for the study participation, ad-equate bone marrow reserve, adequate liver and renalfunction, with no systemic or locoregional therapy in themetastatic setting, and at least one evaluable metastatic le-sion with abnormal FDG uptake.Exclusion criteria included: uncontrolled brain metas-tasis, pregnancy or breast-feeding, history of diabetesmellitus, diagnosis of second primary malignancy, andactive or uncontrolled infection.18 F-FDG PET, image analysis and information collection18 F-FDG was produced automatically by cyclotron(Siemens CTI RDS Eclips ST) using Explora FDG4™module at our single institution. PET/CT was performedusing a PET/CT system (Siemens biograph 16HR). Allpatients were instructed to fast for at least 6 hours be-fore PET imaging. At the time of the tracer injection,patients should have had a blood glucose level of lessthan 7.8 mmol/L. Before and after injection, patientswere kept lying comfortably in a quiet, dimly lit room.There was no significant difference in blood glucose levelsmeasured at the time of the pre- and post-18 F-FDG stud-ies. Image acquisition was started 1 h ± 10 min after intra-venous administration of FDG (7.4 MBq/kg body weight).For the semi-quantitative analysis, a volume of interest(VOI) was drawn with a multimodality computer plat-form (Siemens) for each lesion with the largest uptakeaccording to size and intensity. Tumor size had to be aminimum of 1 cm to minimize partial volume averagingeffects in FDG-PET interpretation. Interpretation of thePET/CT images was based on assessment of the focalFDG uptake and a quantitative evaluation by calculatingthe SUVmax for each lesion instead of using the meanSUV of the lesion, which was more operator-dependent.Two nuclear medicine - CT diagnostic radiologists withat least 5 years of experience and unaware of the clinicalinformation analyzed the data independently, and a thirdsimilarly qualified physician was asked for opinion incases of discordance. The lesions with positive 18 F-FDGuptake were biopsied (n = 55, including 27 fine needleaspirations and 28 core biopsies), or assessed by furtherimaging and clinical follow-up (n = 79) to establish ma-lignant characteristics.Baseline information of the cohort including SUVmaxand molecular subtypes was collected. For patients whohad multiple metastatic sites, the single lesion with thehighest SUVmax was used for calculation. All the infor-mation of molecular classification was obtained from theinitial tumor sample from the primary surgery, withevaluation of patients’ tumor status performed usingResponse Evaluation Criteria in Solid Tumors (RECIST)v1.0 criteria. All patients were followed up at two-monthintervals, and the data were collected and updated untilFebruary 25, 2012. Relapse-free interval (RFI) wasdefined as the interval between primary tumor and re-currence. Progression-free survival (PFS) was defined asthe length of time from the date of the informed consentto disease progression or death from any cause. Overallsurvival (OS)1 was defined as the interval between thedate of breast surgery and the date of death from anycause. OS2 was defined as the time from the date of theinformed consent until the date of death from any cause.Statistical methodologyWe present summary statistics for SUVmax as mediansand interquartile ranges (IQRs), because data were notnormally distributed (data not shown). The impact ofdifferent clinical parameters including luminal types onBaseline SUVmax was evaluated by Mann–Whitney Utest (between 2 groups) or Kruskal-Wallis test ( ≥ 3 groups).Receiver operator characteristic (ROC) curves were used toidentify potential SUV cutoffs predictive of different luminalsubtype. An area under the curve of 1.0 would indicate aperfect test, whereas 0.5 would represent a noninformativetest. Kaplan-Meier method was accessed for survival ana-lysis. Prognostic variables identified by univariate analysis,with P < 0.1, were analyzed in the multivariate Cox model.All reported p-values were two-sided. Statistical significancelevels were set at P < 0.05. Statistical analyses were per-formed with SPSS 16.0 (SPSS, Chicago, IL).ResultsPatient and tumor characteristicsOverall, 305 MBC patients with luminal subtypes signedinformed consent documents and underwent screeningconsecutively, of whom 134 were eligible for this studyZhang et al. BMC Cancer 2013, 13:42 Page 3 of 11http://www.biomedcentral.com/1471-2407/13/42and included into the final analysis (median age, 52 years;range, 28–74 years) (Figure 1). The median time fromdiagnosis of primary disease to MBC diagnosis was 32.1 -months (range, 0.5–245.9 months), and most patients(64.2%) relapsed after 2 years. Out of all eligible patients,75 (56.0%) were luminal A type, 59 (44.0%) were luminalB type. Visceral metastases were present in 70 patients(52.2%) and non-visceral metastases included only lymphnode involved (13.4%), only bone involved (14.9%), onlyskin and soft tissue involved (3.7%), and mixed (20.1%).Before the PET/CT procedure, 2 patients (1.5%) did notreceive any systemic treatment, 4 (3.0%) received adju-vant or neoadjuvant (adj/neo) chemotherapy only, 6(4.5%) received adj/neo hormonal therapy only, 101(75.4%) received adj/neo chemotherapy plus hormonaltherapy, and 21 (15.7%) received regimens includingtargeting agents in the adj/neo setting. Other baselinecharacteristics are provided in Table 1. The median follow-up time of this cohort after inclusion was 26.6 months(range, 14.17–51.2 months).Factors associated with baseline SUVmaxThe current study evaluated the influence of age, men-struation status, tumor histology, luminal subtype, typeof neo/adjuvant therapy, RFI, number of metastatic sites,and visceral metastasis on Baseline SUVmax withMann–Whitney U test or Kruskal-Wallis test. If patientshad multiple metastatic lesions, the maximum one ofSUVmax values of these lesions was selected for statis-tical analysis. The results showed that SUVmax was sig-nificantly higher only in patients with more metastaticsites (P = 0.002) or with presence of visceral metastasis(P = 0.009). In patients without visceral metastases,SUVmax of patients with bone involved only had a trendto be lower than the others (P = 0.063) (Table 1).Evaluation of baseline SUVmax to differentiate luminalsubtypesThe ROC curve was obtained by plotting a graph, inwhich the vertical axis showed sensitivity and the hori-zontal axis showed the false-positive rate. The areaunder ROC curve was 0.516 (SE 0.052) (SE is the stand-ard error of the area estimate), which indicated that theBaseline SUVmax in the metastatic setting did not ef-fectively separate patients with luminal A subtype fromthose with luminal B subtype (Figure 2).Baseline SUVmax and luminal subtypes as prognosticvariablesAll patients in the study experienced disease relapse anddeveloped metastases after primary breast surgery. Univariateanalysis showed that luminal subtype was significantlyassociated with RFI (P < 0.001) and OS1 (P = 0.011), but notwith PFS (P = 0.550) or OS2 (P = 0.233) (Table 2 andFigure 3A-D). Age, menstruation status, and tumor histologyhad no significant effect on PFS and OS2.The univariate analysis also indicated that RFI ≤ 2 years(P = 0.003 and P = 0.017, respectively), more metastaticsites (P = 0.002 and P = 0.032, respectively), presence ofvisceral metastasis (P = 0.035 and P = 0.393, respect-ively), chemotherapy as the first-line therapy after PET/CT (P = 0.037 and P=0.019, respectively) and higherBaseline SUVmax (P = 0.002 and P = 0.009, respectively;Figure 3E-F) were significantly associated with shorterPFS and OS2 (Table 2). Here, the patients with differentSUVmax were classified into three groups based ontertiles of SUVmax. Tertiles (as opposed to quartiles,quintiles, etc.) were chosen to balance the flexibilitygained by adding more groups with the need to keepgroup sizes sufficiently large for subgroup analyses.Cox regression analysis showed that Baseline SUVmax,RFI, and number of metastatic sites were three inde-pendent prognostic factors for PFS, while the significantpredictors of OS2 in the regression model were BaselineSUVmax and RFI. Hazard ratios (HRs) for thesefactors are reported in Table 3. The significant prog-nostic effect of SUVmax on PFS and OS wasmaintained after correcting for tumor phenotype andvariables with P < 0.1 on univariate analysis. By using thetertile with the lowest SUVmax as the reference group,patients in the highest tertile of SUVmax had the shortestPFS (HR = 2.06; 95% CI, 1.23-3.45) and OS (HR = 3.54;95% CI, 1.66-7.55).DiscussionGiven the fact that human breast cancer depends on HRsignaling in regards to response to endocrine therapies,breast cancers have traditionally been sub-classified intoHR-positive (or “luminal”) and HR-negative diseases. Asidentified, even the HR-positive or luminal cancers com-prise a spectrum of tumors with varying degrees ofproliferation and levels of genetic aberrations. Thus, “lu-minal type” of HR positive tumors can be further dividedinto subclass A and B with luminal B being higher grade,having higher proliferation index and a poorer prognosisindependent on hormonal therapy. Since a significantcorrelation between FDG uptake in breast cancer byPET scan and proliferation index has been observed[17], and the tumor proliferation, as defined bymicroarray-based gene signatures or Oncotype DXtesting (the 21 gene assay, Genomic Health), has beenshown to be one of the strongest predictors of outcomefor patients with HR-positive disease [18], it was import-ant to establish whether SUVmax as identified by PETscanning could noninvasively differentiate luminal Afrom luminal B tumors, and whether SUVmax couldpredict the outcome of MBC patients with luminalsubtypes.Zhang et al. BMC Cancer 2013, 13:42 Page 4 of 11http://www.biomedcentral.com/1471-2407/13/42The evaluation of Baseline SUVmax of metastatic sitesafter relapse may provide important information abouttumor proliferation and metabolism that could be of prog-nostic significance. In this regard, a study of the associ-ation of Baseline SUVmax with other well-establishedprognostic factors could be an important first step to-wards establishing the relevance of FDG-PET in the prog-nostic characterization of MBC. Our study found thatBaseline SUVmax of MBC was significantly associatedonly with number of metastatic sites and presence of vis-ceral metastasis. This finding could be the result ofaccelerated glucose metabolism and related increasedmetabolic activity of those more aggressive metastatictumor phenotypes. However, the location of metastatic le-sion could also influence the SUVmax. For example, bonemetastasis of breast cancer is often osteoblastic andTable 1 Patient characteristics and SUVmax comparisons between or among groupsCharacteristics No. of patients(n=134)Baseline SUVmaxMedian IQR P value*Age≤ 50 63 (47.0%) 6.60 5.10–9.10 0.616> 50 71 (53.0%) 6.85 4.80–9.80Menstruation statusPre-menopausal 53 (39.6%) 6.60 5.10–9.00 0.685Post-menopausal 81 (60.4%) 6.85 4.85–9.75HistologyIDC 124 (92.5%) 6.95 5.03–9.66 0.131ILC 6 (4.5%) 6.33 5.15–9.90Others 4 (3.0%) 4.65 3.10–5.68Luminal subtypeLuminal A 75 (56.0%) 6.75 5.10–9.20 0.744Luminal B 59 (44.0%) 7.00 4.80–10.00Adjuvant/neoadjuvant therapyOnly CT (± RT) 4 (3.0%) 6.68 5.76–9.05 0.887Only HT (± RT) 6 (4.5%) 6.15 3.74–9.85CT + HT (± RT) 101 (75.4%) 6.75 5.15–9.25Therapy with TT 21 (15.7%) 7.00 4.75–10.25No 2 (1.5%) 5.50 4.50–6.50Relapse-free interval≤ 2 years 48 (35.8%) 6.25 4.81–9.20 0.592> 2 years 86 (64.2%) 7.00 5.18–9.59No. of metastatic sites1 53 (39.6%) 6.10 4.60–9.08 0.0022 37 (27.6%) 6.25 4.70–8.15≥ 3 44 (32.8%) 8.40 6.21–10.73Visceral metastasisYes 64 (47.8%) 7.60 5.53–10.28 0.009No 70 (52.2%) 6.32 4.80–7.73Only bone 20 (14.9%) 4.85 3.58–7.10 0.063OthersOnly lymph node 18 (13.4%) 6.37 5.20–7.78Only skin & soft tissue 5 (3.7%) 10.1 4.45–10.30Mixed 27 (20.1%) 6.70 5.10–7.70SUVmax: the maximum standardized uptake value; No.: number; IQR: interquartile range; IDC: invasive ductal carcinoma; ILC: invasive lobular carcinoma; CT:chemotherapy; RT: radiotherapy; HT: hormonal therapy; TT: target therapy.* Mann–Whitney U test (between 2 groups) or Kruskal-Wallis test (≥3 groups).Zhang et al. BMC Cancer 2013, 13:42 Page 5 of 11http://www.biomedcentral.com/1471-2407/13/42osteoblastic bone metastatic foci usually show low FDGuptake, regardless of biologic behavior of tumor cells. Inpatients with visceral metastases (± non-visceral metasta-ses) of this study, almost all the SUVmax were obtainedfrom visceral lesions. Only in patients without visceral me-tastases, the assessment of influence on SUVmax by meta-static locations such as bone, lymph node, skin or softtissue might be important. However, we did not find anysignificant differences among these locations (P = 0.235),even between patients with bone involved only and theothers (P = 0.063) in terms of median SUVmax.Our further investigation using the ROC curveindicated that the Baseline SUVmax in the metastaticsetting could not differentiate luminal A from luminal Bsubtypes effectively. This might be because the differ-ence between luminal A and B tumors in proliferationand metabolism was not substantial enough to influenceSUVmax. Therefore, the SUVmax cannot be used as anoninvasive indicator to differentiate luminal A from lu-minal B subtypes. However, it should be noted that theassociation between Baseline SUVmax and luminalsubtypes may be confounded by the fact that relapsed ormetastatic lesions may have a different HR or HER2 sta-tus from that of the primary tumor [19-24] and that corebiopsies in our study were performed only in a smallproportion of patients at the time of relapse. Thus, inorder to clarify whether SUVmax will have a role in lu-minal subtypes differentiation, it may be necessary toconduct another large-sample study to correlate SUVmaxwith core biopsies of the PET scan tested metastaticlesions. However, from 28 patients with core biopsies inour study, only 14.3% had discordant molecular subtypeswith the primary lesions (Additional file 1: Figure 1).When Baseline SUVmax was used to differentiate the lu-minal subtypes after relapse in these patients, the areaunder ROC curve was 0.551 (SE 0.122) and still indicatedno help to differentiate luminal A from luminal B(Additional file 2: Figure 2).Sorlie et al. demonstrated that breast cancer can beclassified into 5 different subtypes according to cDNAmolecular profiles, and that these molecular subtypessignificantly influence patient’s prognosis [4]. Subse-quently, the study of Munoz et al. [25] showed a signifi-cantly unfavorable prognosis for luminal B patients incomparison to those with luminal A subtype in terms ofRFI and OS1. Our data confirmed these results. How-ever, in our study, we did not find different luminalsubtypes predicting the PFS or OS2 after relapse, aphenomenon which could also be result of transform-ation of the original HR or HER2 status. Another reasonfor this could be a suboptimal accuracy in the measure-ment of our luminal subtypes, with a sensitivity of 77%(95% CI, 0.64-0.87) and a specificity of 78% (95% CI,0.68-0.87) [6]. Hence, it was necessary to explore newindicator to determine prognosis of the MBC patientswith luminal subtypes, and PET scan was of particularattraction due to its non-invasive nature.Our study is one of the first to show that the BaselineSUVmax of PET scan has significant association withprognosis and outcome of MBC patients with luminalsubtypes in terms of PFS and OS2, with the multivariateCOX regression analysis confirming the SUVmax an in-dependent prognostic factor.Although some studies have examined PET/CT im-aging as a predictor of treatment response in theprimary breast cancer lesion [26-33], significantly lessis known about how baseline PET/CT imaging can beused as a prognostic tool by quantifying radiotraceraccumulation in metastases. A very recently publishedstudy showed that only in patients with newly diagnosedMBC to bone was Baseline SUVmax tertile signifi-cantly associated with OS on both univariate analysis(HR = 3.13) and multivariate analysis (HR = 3.19) [34].However, this study was a retrospective one and did notfocus on the patients with luminal subtypes.Our study was prospectively performed with importantfindings for luminal type breast cancer patients withnewly diagnosed metastases. Baseline SUVmax wasfound significantly related to the number of metastaticsites and presence of visceral metastasis but could noteffectively differentiate patients with luminal A from lu-minal B subtype. Most importantly, although luminalFigure 2 The receiver operator characteristic (ROC) curve forSUVmax in the differential diagnosis of luminal A subtype fromluminal B subtype. The curve describes the association betweensensitivity and specificity at different thresholds. The area under thecurve (AUC) was 0.516.Zhang et al. BMC Cancer 2013, 13:42 Page 6 of 11http://www.biomedcentral.com/1471-2407/13/42Table 2 Univariate analysis of prognostic factors affecting RFI, OS1, PFS and OS2Factors Median RFI(months)Pvalue*Median OS1(months)Pvalue*Median PFS(months)Pvalue*Median OS2(months)Pvalue*Age≤ 50 NA† NA† 10.0 0.868 36.5 0.966> 50 10.9 NRMenstruation statusPre-menopausal NA† NA† 9.5 0.985 36.5 0.565Post-menopausal 11.3 NRHistologyIDC 31.5 0.550 130.8 0.454 10.4 0.513 36.5 0.560ILC 36.4 NR 4.4 NROthers 39.1 72.8 17.2 32.6Luminal subtypeLuminal A 41.2 <0.001 222.1 0.011 11.6 0.550 36.5 0.233Luminal B 24.2 71.7 9.4 32.6Adjuvant/neoadjuvanttherapyOnly CT (± RT) 17.7 0.138 NR 0.927 5.1 0.463 NR 0.639Only HT (± RT) 17.2 NR 23.2 NRCT + HT (± RT) 36.1 NR 11.3 NRTherapy with TT 28.9 NR 8.4 NRNo 28.9 NR 13.7 NRRelapse-free interval≤ 2 years NA NA 8.2 0.003 22.8 0.017> 2 years 12.9 NRNo. of metastatic sites1 NA NA 13.0 0.002 NR 0.0322 16.5 32.6≥ 3 8.4 25.1Visceral metastasisYes NA NA 9.0 0.035 36.5 0.393No 13.3 NRFirst-line therapy after PET/CTHT NA NA 21.0 0.037 NR 0.019CT (± TT) 9.5 32.6Baseline SUVmax≤ 5.60 (Lowest tertile) NA NA 19.2 0.002 NR 0.0095.60 ~ 8.70 (Intermediatetertile)10.4 35.3> 8.70 (Highest tertile) 8.2 22.6RFI: relapse-free interval; OS: overall survival; PFS: progression-free survival; IDC: invasive ductal carcinoma; ILC: invasive lobular carcinoma; CT: chemotherapy; RT:radiotherapy; HT: hormonal therapy; TT: target therapy; No.: number; SUVmax: the maximum standardized uptake value; NA: not applicable; NR: not reached.* Log-rank test.† The data of age and menstruation status were collected after diagnosis of relapse and informed consent obtained. These data were different from those atdiagnosis of primary breast cancer.Zhang et al. BMC Cancer 2013, 13:42 Page 7 of 11http://www.biomedcentral.com/1471-2407/13/42Figure 3 (See legend on next page.)Zhang et al. BMC Cancer 2013, 13:42 Page 8 of 11http://www.biomedcentral.com/1471-2407/13/42subtype diagnosed according to the initial tumor samplefrom the primary surgery could predict the RFI, it couldnot predict PFS or OS2 after developing relapse or me-tastases. In contrast, Baseline SUVmax as determined onPET scan was predictive of both PFS and OS. In multi-variate analysis using COX regression model, the Base-line SUVmax, RFI, and number of metastatic sites werethree independent prognostic factors for PFS. For OS,the significant predictors were only Baseline SUVmaxand RFI.SUV has many drawbacks as it is dependent onparameters such as the delay between injection andmeasurement, plasma glucose concentration, bodyweight, instrumental factors and partial volume effect(PVE) [35]. SUVmax is defined as the SUV derived fromthe single voxel showing the highest uptake within adefined region of interest (ROI) or VOI. In the absenceof noise, this SUVmax is indeed the least affected byPVE and so is often considered the best measure oftumor uptake. However, in any real imaging situation,noise is always present, making SUVmax variable. An-other drawback of SUVmax is that because it is derivedfrom a single voxel, it may not be an adequate surrogatemarker for true tumor biology and it can be heavilyinfluenced by voxel size [36]. Use of the maximum pixelvalue in a tumor to characterize tumor uptake, however,does make the measurement independent of the obser-ver. This is why, despite its sensitivity to noise and voxelsize, the use of SUVmax is still popular.Several limitations of our study should be addressed.Firstly, as not all MBC patients in our center underwentPET/CT imaging, a selection bias may have played a rolein our patients not being representative of general popu-lation of MBC cases with luminal disease. Secondly,more than a half of lesions with positive 18 F-FDG up-take were not biopsied, and thus metastatic disease wasdiagnosed only with imaging and long-time clinicalfollow-up. Thirdly, not all HER2-positive luminal Bpatients received trastuzumab, which may partly influ-ence applicability of our results to the HER2-positivecases who will have trastuzumab therapy. Lastly, weexamined PET/CT imaging from only 1 time-point and,thus, are unable to comment on the predictive effect ofPET/CT imaging with regard to treatment effect.In spite of these limitations, our study remains the firstto establish the role of PET scanning as a noninvasiveoutcome indicator of luminal A versus luminal B MBCsubtypes.ConclusionsWe conclude that while the Baseline SUVmax in ourstudy of MBC did not correlate with molecular subtypesof primary tumor, the SUVmax, rather than molecularsubtype, emerged as a potential surrogate marker forsurvival with metastatic disease. These data indicate apromise of PET scan use for prognostic assessment ofpatients with MBC in general, with future studiesrequired to clarify the PET scan role in refining, as anon-invasive procedure, the significance and, ultimately,individualized therapeutic options for different molecu-lar subtypes.Table 3 Cox regression* results of PFS and OS2Cox regression results of PFSIndependent prognostic factors Hazard ratio (HR) 95% CI P valueRelapse-free interval≤ 2 years Ref> 2 years 0.51 0.34-0.76 0.001No. of metastatic sites1 Ref2 1.00 0.59-1.68 0.993≥ 3 1.90 1.18-3.08 0.008Baseline SUVmaxLowest tertile RefIntermediate tertile 1.60 0.97-2.66 0.067Highest tertile 2.06 1.23-3.45 0.006Cox regression results of OS2Independent prognostic factors Hazard ratio (HR) 95% CI P valueRelapse-free interval≤ 2 years Ref> 2 years 0.44 0.25-0.78 0.005Baseline SUVmaxLowest tertile RefIntermediate tertile 2.44 1.12-5.32 0.025Highest tertile 3.54 1.66-7.55 0.001* The procedure was carried out with the method of “Forward: LR”.PFS: progression-free survival; OS: overall survival; CI: confidence interval; No.:number; SUVmax: the maximum standardized uptake value; Ref:reference category.(See figure on previous page.)Figure 3 Luminal subtypes and Baseline SUVmax as prognostic variables in survival curves. (A) Relapse-free interval (RFI) curves accordingto Luminal types. (B) Overall survival 1 (OS1) curves according to Luminal types. (C) Progression-free survival (PFS) curves according to Luminaltypes. (D) Overall survival 2 (OS2) curves according to Luminal types. (E) PFS curves according to Baseline SUVmax tertiles. (F) OS2 curvesaccording to Baseline SUVmax tertiles.Zhang et al. BMC Cancer 2013, 13:42 Page 9 of 11http://www.biomedcentral.com/1471-2407/13/42Additional filesAdditional file 1: Figure 1. Summary of molecular subtype differencesbetween the primary and relapsed or metastatic lesion in 28 patientswith core biopsies after recurrence.Additional file 2: Figure 2. The receiver operator characteristic (ROC)curve for SUVmax in the differential diagnosis of luminal A subtype fromluminal B subtype in patients with core biopsies after recurrence (luminalA, 13; luminal B, 12). The curve describes the association betweensensitivity and specificity at different thresholds. The area under the curve(AUC) was 0.551.AbbreviationsSUVmax: Maximum standard uptake value; MBC: Metastatic breast cancer;PFS: Progression-free survival; OS: Overall survival; HR: Hormone receptors;ER: Estrogen receptor; PgR: Progesterone receptor;IHC: Immunohistochemical; PET: Positron emission tomography;FDG: Fluorodeoxyglucose; ECOG: Eastern cooperative oncology group;RECIST: Response Evaluation Criteria in Solid Tumors; VOI: Volume of interest;RFI: Relapse-free interval; IQRs: Interquartile ranges; ROC: Receiver operatorcharacteristic; HR: Hazard ratio; PVE: Partial volume effect; ROI: Region ofinterest.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsConceived and designed the study: XCH, YJZ. Performed the study: JZ, ZJ,MZ, YPZ. Analyzed the data: JZ, ZJ, JR, GL, BYW, ZHW. Wrote the paper: JZ,ZJ, JR, XCH. All authors have read and approved the final manuscript.AcknowledgementsWe thank all the patients who participated in the study. We also thank allthe personnel in the hospital who help the study accomplish successfully.Author details1Department of Medical Oncology, Department of Oncology, FudanUniversity Shanghai Cancer Center; Shanghai Medical College, FudanUniversity, Shanghai, China. 2Faculty of Medicine, School of Population andPublic Health, University of British Columbia, Vancouver, BC, Canada.3Department of Nuclear Medicine, Department of Oncology, FudanUniversity Shanghai Cancer Center; Shanghai Medical College, FudanUniversity, Shanghai, China. 4Department of Medical Oncology, FudanUniversity Shanghai Cancer Center Minhang Branch, Fudan University,Shanghai, China.Received: 30 October 2012 Accepted: 30 January 2013Published: 31 January 2013References1. Youlden DR, Cramb SM, Dunn NA, Muller JM, Pyke CM, Baade PD:The descriptive epidemiology of female breast cancer: an internationalcomparison of screening, incidence, survival and mortality. CancerEpidemiol 2012, 36(3):237–248.2. Brewster AM, Hortobagyi GN, Broglio KR, Kau SW, Santa-Maria CA, Arun B,Buzdar AU, Booser DJ, Valero V, Bondy M, et al: Residual risk of breastcancer recurrence 5 years after adjuvant therapy. J Natl Cancer Inst 2008,100(16):1179–1183.3. Mariani G: New developments in the treatment of metastatic breastcancer: from chemotherapy to biological therapy. Ann Oncol 2005,16(Suppl 2):i191–i194.4. Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, EisenMB, van de Rijn M, Jeffrey SS, et al: Gene expression patterns of breastcarcinomas distinguish tumor subclasses with clinical implications. ProcNatl Acad Sci USA 2001, 98(19):10869–10874.5. Carey LA, Perou CM, Livasy CA, Dressler LG, Cowan D, Conway K, Karaca G,Troester MA, Tse CK, Edmiston S, et al: Race, breast cancer subtypes, andsurvival in the Carolina Breast Cancer Study. JAMA 2006, 295(21):2492–2502.6. Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, Watson M, Davies S,Bernard PS, Parker JS, et al: Ki67 index, HER2 status, and prognosis of patientswith luminal B breast cancer. J Natl Cancer Inst 2009, 101(10):736–750.7. Hu Z, Fan C, Oh DS, Marron JS, He X, Qaqish BF, Livasy C, Carey LA,Reynolds E, Dressler L, et al: The molecular portraits of breast tumors areconserved across microarray platforms. BMC Genomics 2006, 7:96.8. Perou CM, Jeffrey SS, van de Rijn M, Rees CA, Eisen MB, Ross DT,Pergamenschikov A, Williams CF, Zhu SX, Lee JC, et al: Distinctive geneexpression patterns in human mammary epithelial cells and breastcancers. Proc Natl Acad Sci USA 1999, 96(16):9212–9217.9. Alavi A, Lakhani P, Mavi A, Kung JW, Zhuang H: PET: a revolution inmedical imaging. Radiol Clin North Am 2004, 42(6):983–1001.10. Allal AS, Slosman DO, Kebdani T, Allaoua M, Lehmann W, Dulguerov P:Prediction of outcome in head-and-neck cancer patients using thestandardized uptake value of 2-[18 F]fluoro-2-deoxy-D-glucose. Int JRadiat Oncol Biol Phys 2004, 59(5):1295–1300.11. Downey RJ, Akhurst T, Gonen M, Vincent A, Bains MS, Larson S, Rusch V:Preoperative F-18 fluorodeoxyglucose-positron emission tomographymaximal standardized uptake value predicts survival after lung cancerresection. J Clin Oncol 2004, 22(16):3255–3260.12. Sasaki R, Komaki R, Macapinlac H, Erasmus J, Allen P, Forster K, Putnam JB,Herbst RS, Moran CA, Podoloff DA, et al: [18 F]fluorodeoxyglucose uptakeby positron emission tomography predicts outcome of non-small-celllung cancer. J Clin Oncol 2005, 23(6):1136–1143.13. Kitajima K, Kita M, Suzuki K, Senda M, Nakamoto Y, Sugimura K: Prognosticsignificance of SUVmax (maximum standardized uptake value) measuredby [(18)F]FDG PET/CT in endometrial cancer. Eur J Nucl Med Mol Imaging2012, 39(5):840–845.14. Lee YY, Choi CH, Kim CJ, Kang H, Kim TJ, Lee JW, Lee JH, Bae DS, Kim BG: Theprognostic significance of the SUVmax (maximum standardized uptakevalue for F-18 fluorodeoxyglucose) of the cervical tumor in PET imaging forearly cervical cancer: preliminary results. Gynecol Oncol 2009, 115(1):65–68.15. Namura K, Minamimoto R, Yao M, Makiyama K, Murakami T, Sano F, HayashiN, Tateishi U, Ishigaki H, Kishida T, et al: Impact of maximum standardizeduptake value (SUVmax) evaluated by 18-Fluoro-2-deoxy-D-glucosepositron emission tomography/computed tomography (18 F-FDG-PET/CT) on survival for patients with advanced renal cell carcinoma: apreliminary report. BMC Cancer 2010, 10:667.16. Pan L, Gu P, Huang G, Xue H, Wu S: Prognostic significance of SUV onPET/CT in patients with esophageal cancer: a systematic review andmeta-analysis. Eur J Gastroenterol Hepatol 2009, 21(9):1008–1015.17. Shimoda W, Hayashi M, Murakami K, Oyama T, Sunagawa M: Therelationship between FDG uptake in PET scans and biological behaviorin breast cancer. Breast Cancer 2007, 14(3):260–268.18. Geyer FC, Rodrigues DN, Weigelt B, Reis-Filho JS:Molecular classification of estrogenreceptor-positive/luminal breast cancers. Adv Anat Pathol 2012, 19(1):39–53.19. Macfarlane R, Seal M, Speers C, Woods R, Masoudi H, Aparicio S, Chia SK:Molecular alterations between the primary breast cancer and thesubsequent locoregional/metastatic tumor. Oncologist 2012, 17(2):172–178.20. Simmons C, Miller N, Geddie W, Gianfelice D, Oldfield M, Dranitsaris G, ClemonsMJ: Does confirmatory tumor biopsy alter the management of breast cancerpatients with distant metastases? Ann Oncol 2009, 20(9):1499–1504.21. Liedtke C, Broglio K, Moulder S, Hsu L, Kau SW, Symmans WF, Albarracin C,Meric-Bernstam F, Woodward W, Theriault RL, et al: Prognostic impact ofdiscordance between triple-receptor measurements in primary andrecurrent breast cancer. Ann Oncol 2009, 20(12):1953–1958.22. Gutierrez MC, Detre S, Johnston S, Mohsin SK, Shou J, Allred DC, Schiff R,Osborne CK, Dowsett M: Molecular changes in tamoxifen-resistant breastcancer: relationship between estrogen receptor, HER-2, and p38mitogen-activated protein kinase. J Clin Oncol 2005, 23(11):2469–2476.23. Amir E, Miller N, Geddie W, Freedman O, Kassam F, Simmons C, Oldfield M,Dranitsaris G, Tomlinson G, Laupacis A, et al: Prospective study evaluatingthe impact of tissue confirmation of metastatic disease in patients withbreast cancer. J Clin Oncol 2012, 30(6):587–592.24. Amir E, Clemons M, Purdie CA, Miller N, Quinlan P, Geddie W, Coleman RE,Freedman OC, Jordan LB, Thompson AM: Tissue confirmation of diseaserecurrence in breast cancer patients: Pooled analysis of multi-centre, multi-disciplinary prospective studies. Cancer Treat Rev 2011, 38(6):708–714.25. Munoz M, Fernandez-Acenero MJ, Martin S, Schneider J: Prognosticsignificance of molecular classification of breast invasive ductalcarcinoma. Arch Gynecol Obstet 2009, 280(1):43–48.Zhang et al. BMC Cancer 2013, 13:42 Page 10 of 11http://www.biomedcentral.com/1471-2407/13/4226. Martoni AA, Zamagni C, Quercia S, Rosati M, Cacciari N, Bernardi A, Musto A,Fanti S, Santini D, Taffurelli M: Early (18)F-2-fluoro-2-deoxy-d-glucosepositron emission tomography may identify a subset of patients withestrogen receptor-positive breast cancer who will not respond optimallyto preoperative chemotherapy. Cancer 2010, 116(4):805–813.27. Ueda S, Tsuda H, Saeki T, Omata J, Osaki A, Shigekawa T, Ishida J, Tamura K,Abe Y, Moriya T, et al: Early metabolic response to neoadjuvant letrozole,measured by FDG PET/CT, is correlated with a decrease in the Ki67labeling index in patients with hormone receptor-positive primarybreast cancer: a pilot study. Breast Cancer 2011, 18(4):299–308.28. Humbert O, Berriolo-Riedinger A, Riedinger JM, Coudert B, Arnould L, CochetA, Loustalot C, Fumoleau P, Brunotte F: Changes in 18 F-FDG tumormetabolism after a first course of neoadjuvant chemotherapy in breastcancer: influence of tumor subtypes. Ann Oncol 2012, 23(10):2572–2577.29. Kolesnikov-Gauthier H, Vanlemmens L, Baranzelli MC, Vennin P, Servent V,Fournier C, Carpentier P, Bonneterre J: Predictive value of neoadjuvantchemotherapy failure in breast cancer using FDG-PET after the firstcourse. Breast Cancer Res Treat 2012, 131(2):517–525.30. Constantinidou A, Martin A, Sharma B, Johnston SR: Positron emissiontomography/computed tomography in the management of recurrent/metastatic breast cancer: a large retrospective study from the RoyalMarsden Hospital. Ann Oncol 2011, 22(2):307–314.31. Couturier O, Jerusalem G, N’Guyen JM, Hustinx R: Sequential positronemission tomography using [18 F]fluorodeoxyglucose for monitoringresponse to chemotherapy in metastatic breast cancer. Clin Cancer Res2006, 12(21):6437–6443.32. Tateishi U, Gamez C, Dawood S, Yeung HW, Cristofanilli M, Macapinlac HA:Bone metastases in patients with metastatic breast cancer: morphologicand metabolic monitoring of response to systemic therapy withintegrated PET/CT. Radiology 2008, 247(1):189–196.33. De Giorgi U, Mego M, Rohren EM, Liu P, Handy BC, Reuben JM, MacapinlacHA, Hortobagyi GN, Cristofanilli M, Ueno NT: 18 F-FDG PET/CT findings andcirculating tumor cell counts in the monitoring of systemic therapies forbone metastases from breast cancer. J Nucl Med 2010, 51(8):1213–1218.34. Morris PG, Ulaner GA, Eaton A, Fazio M, Jhaveri K, Patil S, Evangelista L, ParkJY, Serna-Tamayo C, Howard J, et al: Standardized uptake value bypositron emission tomography/computed tomography as a prognosticvariable in metastatic breast cancer. Cancer 2012, 118(22):5454–5462.35. Boellaard R, Oyen WJ, Hoekstra CJ, Hoekstra OS, Visser EP, Willemsen AT,Arends B, Verzijlbergen FJ, Zijlstra J, Paans AM, et al: The Netherlands protocolfor standardisation and quantification of FDG whole body PET studies inmulti-centre trials. Eur J Nucl Med Mol Imaging 2008, 35(12):2320–2333.36. Boellaard R: Mutatis mutandis: harmonize the standard! J Nucl Med 2012,53(1):1–3.doi:10.1186/1471-2407-13-42Cite this article as: Zhang et al.: The maximum standardized uptakevalue of 18 F-FDG PET scan to determine prognosis of hormone-receptorpositive metastatic breast cancer. BMC Cancer 2013 13:42.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitZhang et al. BMC Cancer 2013, 13:42 Page 11 of 11http://www.biomedcentral.com/1471-2407/13/42

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0221466/manifest

Comment

Related Items