The aim of this pilot study was to explore intrapatient mixed metabolic response and early 18F-FDG PET response evaluation using predefined quantification strategies in patients with advanced KRAS wild-type colorectal adenocarcinoma (mCRC) treated with cetuximab.
A 18F-FDG PET was performed at baseline and after 2 cycles of cetuximab. Metabolic response was categorized using thresholds suggested in PERCIST. Quantitative analysis was done for the sum of all target lesions, ≤ 5 lesions and the metabolically most active lesion per PET. Quantitative data were correlated with clinical benefit, according to RECIST v1.1, after two months of treatment.
In nine evaluable patients the total number of target lesions was 34 (1–8 per patient). Mixed metabolic response was observed in three out of seven patients with multiple target lesions, using TLG. Dichotomised metabolic data of the sum of all or ≤ 5 lesions had a concordance with clinical benefit of 89% using SULmax or SULpeak, and 100% using TLG. Evaluating the metabolically most active lesion, concordance was 89% for all three units. Additionally, the decrease in TLG was significantly correlated with PFS for all three quantification strategies.
Mixed metabolic response was observed in nearly half of the patients with advanced KRAS wild-type mCRC treated with cetuximab. If ≤ 5 target lesions were evaluated using TLG clinical benefit was predicted correctly for all patients. Moreover, decrease in TLG is significantly correlated with the duration of PFS. Validation of these promising preliminary results in a larger cohort is currently on-going.
Citation: van Helden EJ, Hoekstra OS, Boellaard R, Roth C, Mulder ER, Verheul HMW, et al. (2016) Early 18F-FDG PET/CT Evaluation Shows Heterogeneous Metabolic Responses to Anti-EGFR Therapy in Patients with Metastatic Colorectal Cancer. PLoS ONE 11(5): e0155178. https://doi.org/10.1371/journal.pone.0155178
Editor: Rafael Rosell, Catalan Institute of Oncology, SPAIN
Received: February 1, 2016; Accepted: April 23, 2016; Published: May 19, 2016
Copyright: © 2016 van Helden et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its supporting Information files.
Funding: This work was funded by a grant from KWF - Alpe d'Huez (2012-5565). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: HV is a member of the advisory board of Erbitux (Merck). HV also received honoraria from Boehringer Ingelheim and Roche for his consultancy/advisory work. HV received research funding (outside this work) from Amgen, Vitromics Healthcare, Immunovo BV, Roche, Novartis. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. The other authors declare no competing interests.
Early changes in glucose metabolism defined with 18F-fluorodeoxyglucose positron emission tomography / computed tomography (18F-FDG PET) is a potential tool to differentiate between responders and non-responders early after start of anti-cancer treatment [1–5]. The advantage of 18F-FDG PET compared to anatomic evaluation of tumour lesions is that changes in metabolic activity can be assessed shortly after start of therapy , whereas anatomic evaluation can only be performed after 2–3 months of treatment.
Complete visual resolution of radiotracer uptake is fairly straightforward and demonstrated to be a good prognostic marker . Yet, with early response evaluation, especially for treatment with targeted agents like cetuximab, smaller changes in 18F-FDG uptake are expected, even in responding patients. Consequently, comparability of acquisition, quantification and response criteria are crucial for the efficacy of 18F-FDG PET as early response marker. The initial 18F-FDG PET response criteria  do not specify strategies for patients with multiple lesions. In clinical practice and in trials, response to therapy is typically classified at a patient level. In 2009 Wahl et al. proposed the PERCIST guideline , in which response is classified using the lesion with the highest radiotracer uptake per time-point, hypothesizing that this lesion is prognostically most relevant. Other strategies are to evaluate changes in 18F-FDG uptake for the sum of all lesions, or for the sum of ≤ 5 lesions (as in Response Evaluation Criteria In Solid Tumours (RECIST) version 1.1 for anatomic evaluation). The quantification strategy for multiple target lesions may be crucial for patients with a heterogeneous response between tumour lesions. Moreover, mixed metabolic response and the effect on response prediction may be particularly relevant for targeted agents, as response of individual lesions may be correlated with the variation of expression of the target or the presence of a resistance-inducing mutation between lesions. Additionally, the optimal type of quantification unit for response prediction, such as standardized uptake value for lean body mass (SUL)max, SULpeak or total lesion glycolysis (TLG), remains unclear.
Early response evaluation for mCRC patients treated with anti-EGFR therapy is clinically relevant as only half of the patients will have clinical benefit [9–11]. Since there are no other known biomarkers, all patients with RAS wild-type mCRC will receive this treatment until first CT evaluation after 2–3 months (4–7 cycles). By identifying non-responders after just one or two cycles, exposition to ineffective drugs can be avoided and other treatment options can be considered. To our knowledge no metabolic data other than a case report  have been published regarding early response evaluation in patients treated with anti-EGFR antibody monotherapy.
The aim of this pilot study was to investigate intrapatient mixed metabolic response based on early response evaluation with 18F-FDG PET after 2 cycles of cetuximab monotherapy in patients with KRAS wild type mCRC. Additionally, the impact of mixed metabolic response on three quantification strategies (the sum of all target lesions, the sum of ≤ 5 lesions and the metabolically most active lesion) with three quantitative PET metrics (SULmax, SULpeak and TLG) were evaluated. Metabolic parameters were correlated with standard CT evaluation according to RECIST v1.1.
Patients and Methods
Patients with unresectable KRAS wild-type mCRC who had been treated according to standard care (fluoropyrimidines, oxaliplatin and irinotecan) and were candidates for anti-EGFR antibody monotherapy were eligible for this PET imaging study with 89Zr-cetuximab  and 18F-FDG PET. All in- and exclusion criteria are stated in S1 Table. At the time of patient accrual, RAS mutations in KRAS exon 3 and 4 and NRAS exon 2, 3 and 4 were not yet identified as biomarkers for primary resistance, therefore only wild type KRAS exon 2 was required for inclusion. Additional RAS and BRAF (exon 15) mutations were analysed retrospectively. Patients were treated biweekly with cetuximab 500mg/m2. This single centre, two step non-randomized intervention study was reviewed and approved (d.d. 27-06-2011) by the Ethics Committee at VU University Medical Center (VUmc) before the study opened for inclusion. All patients signed informed consent before any study activities were conducted. The authors confirm that all ongoing and related trials for this drug/intervention are registered (clinicaltrials.gov NCT01691391; d.d. 15-05-2012; S1 and S2 Files).
18F-FDG PET/CT’s were performed at baseline (within two weeks prior to the first cetuximab infusion) and after four weeks of treatment. All 18F-FDG PET were conducted according to European Association of Nuclear Medicine guidelines . Briefly, patients fasted 6 hours prior to the radiotracer injection. Patients were injected with 3 MBq/kg (± 10%) 18F-FDG. After 60 min (± 5 min) a PET scan was performed from skull base to mid-thigh. Residual activity in the syringe was measured and subtracted from the injected dose before calculations were done.
Data-analysis of all 18F-FDG PET was done after all data was collected and follow-up ended (January 2015). Target lesions were defined as tumour lesions with a minimal diameter of ≥ 2 cm on CT scan, to minimize potential partial volume effect, and with an uptake above background activity. The background activity was calculated in healthy liver tissue in a 3x3 cm Volume Of Interest (VOI) (using the following formula: (1.5x average SUL liver) + (2x SD average SUL liver)) or in the descending aorta in a 1x1 cm VOI (using the following formula: (2x average SUL descending thoracic aorta) + (2x SD average SUL aorta) in case of hepatic metastases according to PERCIST . Tumour VOI’s were created using a semi-automatic delineation tool which used a 50% standardized uptake value (SUV)max threshold with background correction . If semi-automatic delineation was not possible (mostly due to tissue with high uptake positioned closely to the target lesion), the VOI’s were created manually and shrunk down using the same 50% SUVmax threshold with background correction.
Three quantification units were evaluated: SULpeak (derived from the activity measured in a 1 cm3 sphere within the tumour VOI, placed automatically to ensure that it captured the highest mean radioactivity), SULmax (defined as the voxel with the most radioactivity within the tumour VOI), and TLG (defined as SULmean times metabolically active tumour volume, where this volume was defined with an isocontour VOI of 50% of SUVpeak with background correction) . Quantitative analysis of the PET images was performed using three different strategies: 1. all target lesions with a minimal diameter of 2 cm, evaluated separately (paired for the 2 time points) and summed, 2. the sum of ≤ 5 (≤ 2 per organ) metabolically most active target lesions per PET scan (as is suggested in PERCIST as exploratory approach and as is used for anatomical changes according to RECIST version 1.1 ), 3. the single most metabolically active lesion per PET scan (baseline and on-treatment target lesion can differ, c.f. PERCIST).
Metabolic data were classified using predefined criteria according to PERCIST as complete metabolic response (CMR), partial metabolic response (PMR, SUL reduction of ≥ 30% and ≥ 0.8 unit; TLG reduction of ≥ 45%), stable metabolic disease (SMD, changes <30% range for SUL and < 45% range for TLG) and progressive metabolic disease (PMD, increase of ≥ 30% SUL and ≥ 0.8 unit; TLG increase of ≥ 45%) . A patient was classified with mixed metabolic response if lesions (from paired metabolic data) within one patient were categorized in different response categories (e.g. CMR, PMR, SMD and PMD as mentioned above).
Since the aim of early response evaluation is the differentiation between patients with and without clinical benefit, metabolic data was dichotomized as metabolic response (reduction of SUL with ≥ 30% and ≥ 0.8 unit; TLG reduction of ≥ 45%) and metabolic non-response (a reduction smaller than the previously mentioned limits or an increase in 18F-FDG uptake). Additionally, progression free survival (PFS) was correlated with the percentage change in 18F-FDG uptake using Spearman's rank correlation coefficient, this test was performed using SPSS version 22 (IBM Corp., Armonk, NY). A p value < 0.05 was considered as significantly relevant.
During treatment with cetuximab, diagnostic computed tomography (CT) scans were performed every 8 weeks until progressive disease or discontinuation of treatment. Response was evaluated and categorised using RECIST version 1.1 as progressive disease (PD), stable disease (SD), partial response (PR) or complete response (CR) . Clinical benefit is defined as SD, PR or CR as best response to treatment. PFS is defined as the period starting at the first treatment until PD. At the time of publication all patients had progressive disease according to RECIST. A physician, blinded for 18F-FDG-PET data, defined the RECIST target lesions.
Apart from the RECIST measurements, all tumour lesions ≥ 2 cm were measured separately (paired) to study heterogeneity in response and correlate with metabolic changes. These anatomic changes for single lesions were categorized using the same thresholds as in RECIST.
Out of 20 patients screened at VUmc, 10 were not eligible (60% KRAS mutation, 30% no extra-hepatic disease, 10% declined participation) and the remaining patients were included (Fig 1). The study was open for inclusion until March 2014, follow-up ended in January 2015, all patients had progressive disease at that time. There were no patients lost in follow-up. At the time of inclusion patients had a median age of 61 years (with a range 50–73 years), 60% were male and the majority of patients had an adenocarcinoma of the rectum or sigmoid (S2 Table). All patients were KRAS (exon 2 / 3) wild-type, in half of the patients additional RAS mutations were tested and proven wild-type. For the other five patients, the tumour tissue quality was not sufficient for retrospective mutation analysis. All patients were tested for mutations in BRAF exon 15, only patient number 5 had this prognostically poor mutation. Five patients had PD at the first CT evaluation (after 8 weeks of treatment), two had PR and three had SD according to RECIST. Median PFS was 8 weeks, with a range of 6–33 weeks (S2 Table).
Due to movement artefacts one patient (#1) was excluded for quantitative analyses. The remaining patients had 34 target lesions with a diameter of ≥ 2 cm and 18F-FDG uptake above background, with a range of 1 to 8 target lesions per patient. Most target lesions were located in the liver, adrenal glands, lungs or lymph nodes.
The background radioactivity was measured in the liver for three patients and for six patients with hepatic metastases this was done in the descending aorta. Three patients had ≥ 20% difference in background activity between baseline and on-treatment PET. One patient had an increased liver uptake of 21% (+0.50 unit), most likely due to a protocol deviation consisting of a longer interval between 18F-FDG administration and PET scanning at baseline. Nevertheless, this patient had an increased accumulation of 18F-FDG in target lesions at the on-treatment PET (i.e. PMD), concordant with the first evaluation CT scan which showed PD. Two patients had a decrease of 25% (-0.60 and -0.63 unit) in SULmean in the background activity of the aorta. However, in healthy liver the difference did not surpass the 20% threshold.
Mixed metabolic response
To evaluate intrapatient heterogeneous metabolic response between tumour lesions, changes in 18F-FDG uptake after 4 weeks of treatment (2 cycles of cetuximab) and anatomic changes on first CT evaluation (after 8 weeks, i.e. 4 cycles of treatment) were compared for all tumour target lesions separately. Of the nine evaluable patients, seven had multiple target lesions. Intrapatient mixed metabolic response was observed in three out of seven patients when tumour uptake was expressed in TLG. Heterogeneity was most pronounced in patient 2, with metabolically responding, stable and progressive target lesions. Additionally, some heterogeneity in TLG data was observed in patient 3 and 7 (Fig 2). Using SULpeak, intrapatient mixed metabolic response was observed in patient 2 and 3, for SULmax mixed response was observed in patient 2 and 7 (S1 and S2 Figs).
A clear trend is noticeable between change in 18F-FDG uptake and change in size. Intrapatient mixed metabolic response is most pronounced in patient 2; different lesions within this patient are categorized as PMR, SMD and PMD. Patient 3 has lesions in both PMR and SMD categories and patient 7 have lesions in both SMD and PMD categories. In all patients minor heterogeneity in anatomic changes is observed.
Anatomic changes on CT scan of all target lesions (bone lesions, being non-measurable lesions, are excluded) demonstrated minor heterogeneity; different lesions of one patient could be categorized as PR and SD or as PD and SD, but never both progressive and responding lesions within one patient. There was a weak trend between decrease of 18F-FDG uptake and that of tumour lesion size (Fig 2). All ten tumour lesions demonstrating anatomic growth at week 8 had stable or progressive metabolic disease at week 4. Five out of six tumour lesions with partial response based on anatomic size had partial metabolic response. For anatomically stable lesions all three metabolic response criteria were observed.
Early response evaluation: the sum of all target lesions
In Fig 3A the percentage change of the sum of all target lesions is depicted. Metabolic data is dichotomized into metabolic responders (decreased of ≥ 30% SUL or ≥ 45% TLG), or metabolic non-responders (an increase or decrease under the response limit), since this resulted in the best discrimination between patients with and without clinical benefit according to RECIST. Expressing changes in 18F-FDG uptake using SULmax and SULpeak resulted in an 89% concordance with clinical benefit (S3 Fig). However, with TLG early response prediction is accurate for all patients.
Panel A illustrates percentage change in 18F-FDG uptake for the sum of all tumour lesions expressed in SULmax, SULpeak and TLG. Patients without clinical benefit (left side) are all correctly categorised as metabolic non-responders. If uptake is expressed in TLG response prediction was accurate for all patients. For SULmax and SULpeak patient 3 who had clinical benefit, was miscategorised as metabolic non-responder. Panel 3B illustrates percentage change in 18F-FDG uptake for the sum of ≤ 5 lesions (≤ 2 per organ) per PET scan. Metabolic response categories are identical to response prediction based on the sum of all target lesions. Panel 3C illustrates percentage change in 18F-FDG uptake of the metabolically most active lesion per PET scan. Patient 3 is miscategorised as metabolic non-responders for all three quantification units.
Besides the correlation between metabolic response and clinical benefit, metabolic data of the sum of all target lesions is compared to the duration of PFS. As illustrated in Fig 4A, there is a significant correlation (Spearman's rank correlation coefficient; Rs -0.94, p-value < 0.001) between the percentage decrease of the sum of all target lesions expressed in TLG after 4 weeks and the duration of PFS. Also for ≤ 5 lesions and the metabolically most active lesion there is a significant correlation (Fig 4B and 4C; Rs -0.82, p-value = 0.009 and Rs -0.84, p-value = 0.006 respectively).
All patients with clinical benefit have a > 45% decrease in TLG. Notably, there is a significant correlation between reduction in TLG and a longer PFS (Rs -0.94, p-value < 0.001). Panel B illustrates the percentage change of the sum of ≤ 5 target lesions expressed in TLG versus PFS in weeks. Again there is a significant correlation between 18F-FDG decrease and the duration of PFS (Rs -0.82, p-value = 0.009). Panel C illustrates the percentage change of the metabolically most active lesion expressed in TLG versus PFS in weeks (Rs -0.84, p-value = 0.006). Patient number 3 was misclassified as non-responder (red square above dotted line). Yet, this patient fits the line, with a moderate decrease in TLG and a relative short PFS.
Early response evaluation: the sum of ≤ 5 target lesions
In Fig 3B, the percentage change of the sum of ≤ 5 lesions (≤ 2 lesions per organ) is depicted. Patients 2, 3, 4 and 7 had > 5 target lesions. Early response prediction based on ≤ 5 target lesions rendered identical response categories for all patients compared to response prediction based on the sum of all target lesions, for all three quantification units. A 100% concordance with clinical benefit is achieved when 18F-FDG uptake is expressed in TLG. As with response prediction based on all lesions, changes in 18F-FDG based on ≤ 5 target lesions demonstrated a clear trend between 18F-FDG decrease and duration of PFS (Fig 4B).
Early response evaluation: the metabolically most active lesion (cf. PERCIST)
Three patients had a different target lesion at baseline PET compared to the on-treatment PET using the metabolically most active lesion per time point. Concordance between PERCIST (using SULpeak) and RECIST response categories was 44% (Fig 5). However, dichotomized metabolic data correctly predicted clinical benefit in 89% (Fig 3C), misclassifying patient 3 as metabolic non-responder, with a decrease of 21% in SULpeak. The metabolically most active lesion per scan expressed in SULmax and TLG demonstrated an identical distribution as with SULpeak (Fig 3C), again misclassifying patient 3 with a decrease of 28% and 25% using SULmax and TLG, respectively (all data was summarised in S3 Table).
A comparison between RECIST and PERCIST categories. Patients are grouped on the x-axis based on RECIST, with different colours to indicate the PERCIST categories. On the y-axis change in SULpeak of the baseline and on-treatment 18F-FDG PET is demonstrated. The majority of patients which had PD based on CT evaluation had SMD. Of the three patients who had prolonged stable disease, two patients had PMR and one patient had SMD. Both patients with PR based on CT evaluation had PMR on the early response evaluation with 18F-FDG PET.
In this study mixed metabolic response was observed in three out of seven patients (43%) with multiple metastatic target lesions, treated with monotherapy cetuximab. Dichotomised metabolic data of early 18F-FDG PET/CT evaluation for the metabolically most active lesion yielded a concordance with clinical benefit of 89% for all three units. The sum of all or ≤ 5 lesions had a concordance with clinical benefit of 89% using SULmax or SULpeak, and 100% using TLG. Additionally, the decrease in TLG was significantly correlated with PFS for all three quantification strategies.
Although it is known that mixed metabolic response to (targeted) therapy occurs [2;18], the interpretation of a heterogeneous tumour biology remains unclear. Similar to our study, Hendlisz et al. reported a high incidence of mixed metabolic response (68%) in patients with mCRC treated with chemotherapy. Multiple lesions were dichotomized into (dominantly) metabolic response versus (dominantly) metabolic non-response (non-response was defined < 15% decrease in SUVmax). All metabolic non-responders had no response according to RECIST , as is the case in this study. Additionally, Hendlisz et al. evaluated mixed metabolic response and the influence on survival data in 79 mCRC patients treated with sorafenib and chemotherapy. Their conclusion was that patients that only have responding lesions have a significantly longer PFS compared to patients with a heterogeneous response (p-value < 0.001) , this is in concordance with our data. In contrast to the study of Hendlisz et al., our patients are treated with cetuximab monotherapy. Heterogeneity in response may be more essential for monotherapy with a targeted agent, since treatment benefit may be correlated with the variation of expression of the target between lesions or the presence of a resistance-inducing mutation. To our knowledge our study is the first to report on FDG evaluation for monotherapy with a targeted agent.
The optimal quantitative strategy for an accurate 18F-FDG PET early response prediction is unknown. In literature, numerous small studies regarding early 18F-FDG PET response evaluation for cancer treatments have been published. Unfortunately, comparability is poor due to differences in quantification methods [20–22], cancer type [1;23–25], type of therapy [19;26;27], treatment line and primary outcome [2;6;20–22;24]. The PERCIST guideline is the first to describe a systematic comparable method to evaluate patients with multiple tumour lesions, it is based on the hypothesis that the lesion with the highest radiotracer uptake per time-point is prognostically most relevant. Yet, in case of mixed metabolic response, only evaluating the metabolically most active lesion might not accurately represent the entire tumour biology. Moreover, recent publications demonstrated that early metabolic response according to PERCIST guidelines might not optimally predict response according to RECIST after 2 to 3 months of therapy [6;28;29].
An alternative quantitative strategy is the evaluation of multiple target lesions. The sum of multiple tumour lesions could even out possible tumour heterogeneity. Although this study describes a small cohort, a perfect concordance was observed between dichotomized early 18F-FDG response prediction based on multiple tumour lesions with uptake expressed in TLG and clinical benefit. Response evaluation based on the sum of all target lesions yielded the same results as ≤ 5 target lesions, suggesting that evaluating > 5 lesions is not required for correct response prediction.
For the three patients with mixed metabolic response in our study, early response evaluation based on the metabolically most active lesion led to a correct prediction in two patients, only patient 3 was not correctly categorized as responder. Interestingly, of all patients with clinical benefit, patient 3 had the shortest PFS of just 4 months stable disease. One might question whether this patient in fact benefited from cetuximab monotherapy with only having stable disease on one CT scan. In addition, changes in early 18F-FDG response evaluation might predict the length of PFS more accurately (Fig 4) than response according to RECIST for patients treated with targeted agents.
This study demonstrates that early response evaluation with 18F-FDG PET is a promising biomarker, especially if multiple lesions are evaluated using TLG, it was based on a small cohort. As a result, differences in response prediction between the three different strategies were based on observations in a single patient. Thus, validation in a large cohort is needed. In the IMPACT-CRC study, for which patient accrual is currently ongoing, the clinical utility of early response evaluation with 18F-FDG PET and 89Zirconium labelled cetuximab PET, for patients with mCRC treated with cetuximab, will be evaluated (NCT02117466).
Finally, with respect to the commonly used semi-quantitative PET measures (SULmax, SULpeak and TLG): SULmax is less susceptible for inter-observer variability in VOI definition, but more sensitive to background noise, as it is based on only one voxel . Additionally, this unit does not reflect metabolic heterogeneity within a tumour lesion . SULpeak is less susceptible for noise since it comprises of 1 cm3 of the most active part of the tumour VOI . Unlike SUL, TLG is not normalized per unit mass of tumour tissue but includes change in metabolic tumour volume [16;28]. Consequently, TLG might reflect overall tumour lesion biology more accurately compared to SULmax or SULpeak. A potential limitation of TLG is the operator dependency in defining the metabolically active tumour volume. However, repeatability of TLG was shown to be acceptable . In this small cohort, uptake of multiple target lesions expressed in TLG indeed demonstrated a better concordance with clinical benefit compared to SULpeak and SULmax.
In this pilot study interlesional mixed metabolic occurred in 43% of all patients with multiple evaluable tumour (KRAS wild-type) colorectal cancer lesions, treated with cetuximab monotherapy. Early 18F-FDG PET response prediction based on ≤ 5 target lesions (2 ≤ per organ) expressed in TLG resulted in a correctly predicted clinical benefit in all patients. Additionally, there was a significant correlation between the decrease in TLG and PFS. Validation of these promising preliminary results in a larger patient cohort is currently on-going.
S1 Fig. Percentage change in tumour size versus change in 18F-FDG uptake (SULpeak) per tumour lesion.
S2 Fig. Percentage change in tumour size versus change in 18F-FDG uptake (SULmax) per tumour lesion.
S3 Fig. The sum of all target lesions (SULpeak).
Conceived and designed the experiments: OS HV WM. Performed the experiments: EH CR WM. Analyzed the data: EH EM. Wrote the paper: EH OH WM. Manuscript revising and final approval: EH OH RB HV WM.
- 1. Skougaard K, Johannesen HH, Nielsen D, Schou JV, Jensen BV, Hogdall EV et al. CT versus FDG-PET/CT response evaluation in patients with metastatic colorectal cancer treated with irinotecan and cetuximab. Cancer Med. 2014; 3(5):1294–1301. pmid:24941936
- 2. Hendlisz A, Golfinopoulos V, Garcia C, Covas A, Emonts P, Ameye L et al. Serial FDG-PET/CT for early outcome prediction in patients with metastatic colorectal cancer undergoing chemotherapy. Ann Oncol. 2012; 23(7):1687–1693. pmid:22112970
- 3. Monteil J, Mahmoudi N, Leobon S, Roudaut PY, El BA, Verbeke S et al. Chemotherapy response evaluation in metastatic colorectal cancer with FDG PET/CT and CT scans. Anticancer Res. 2009; 29(7):2563–2568. pmid:19596929
- 4. de Geus-Oei LF, Vriens D, van Laarhoven HW, van der Graaf WT, Oyen WJ. Monitoring and predicting response to therapy with 18F-FDG PET in colorectal cancer: a systematic review. J Nucl Med. 2009; 50 Suppl 1:43S–54S. pmid:19403879
- 5. Contractor KB, Aboagye EO. Monitoring predominantly cytostatic treatment response with 18F-FDG PET. J Nucl Med. 2009; 50 Suppl 1:97S–105S. pmid:19403880
- 6. Maffione AM, Ferretti A, Grassetto G, Bellan E, Capirci C, Chondrogiannis S et al. Fifteen different 18F-FDG PET/CT qualitative and quantitative parameters investigated as pathological response predictors of locally advanced rectal cancer treated by neoadjuvant chemoradiation therapy. Eur J Nucl Med Mol Imaging. 2013; 40(6):853–864. pmid:23417501
- 7. Young H, Baum R, Cremerius U, Herholz K, Hoekstra O, Lammertsma AA et al. Measurement of clinical and subclinical tumour response using [18F]-fluorodeoxyglucose and positron emission tomography: review and 1999 EORTC recommendations. European Organization for Research and Treatment of Cancer (EORTC) PET Study Group. Eur J Cancer. 1999; 35(13):1773–1782. pmid:10673991
- 8. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: Evolving Considerations for PET response criteria in solid tumors. J Nucl Med. 2009; 50 Suppl 1:122S–150S. pmid:19403881
- 9. Karapetis CS, Khambata-Ford S, Jonker DJ, O'Callaghan CJ, Tu D, Tebbutt NC et al. K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med. 2008; 359(17):1757–1765. pmid:18946061
- 10. Douillard JY, Oliner KS, Siena S, Tabernero J, Burkes R, Barugel M et al. Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer. N Engl J Med. 2013; 369(11):1023–1034. pmid:24024839
- 11. Price TJ, Peeters M, Kim TW, Li J, Cascinu S, Ruff P et al. Panitumumab versus cetuximab in patients with chemotherapy-refractory wild-type KRAS exon 2 metastatic colorectal cancer (ASPECCT): a randomised, multicentre, open-label, non-inferiority phase 3 study. Lancet Oncol. 2014; 15(6):569–579. pmid:24739896
- 12. Krystal GW, Alesi E, Tatum JL. Early FDG/PET scanning as a pharmacodynamic marker of anti-EGFR antibody activity in colorectal cancer. Mol Cancer Ther. 2012; 11(7):1385–1388. pmid:22562987
- 13. Menke—van der Houven CW, Gootjes EC, Huisman MC, Vugts DJ, Roth C, Luik AM et al. 89 Zr-cetuximab PET imaging in patients with advanced colorectal cancer. Oncotarget; Advance Online Publications: Page 2. 2015.
- 14. Boellaard R, O'Doherty MJ, Weber WA, Mottaghy FM, Lonsdale MN, Stroobants SG et al. FDG PET and PET/CT: EANM procedure guidelines for tumour PET imaging: version 1.0. Eur J Nucl Med Mol Imaging. 2010; 37(1):181–200. pmid:19915839
- 15. Cheebsumon P, Yaqub M, van Velden FH, Hoekstra OS, Lammertsma AA, Boellaard R. Impact of [(1)(8)F]FDG PET imaging parameters on automatic tumour delineation: need for improved tumour delineation methodology. Eur J Nucl Med Mol Imaging. 2011; 38(12):2136–2144. pmid:21858528
- 16. Larson SM, Erdi Y, Akhurst T, Mazumdar M, Macapinlac HA, Finn RD et al. Tumor Treatment Response Based on Visual and Quantitative Changes in Global Tumor Glycolysis Using PET-FDG Imaging. The Visual Response Score and the Change in Total Lesion Glycolysis. Clin Positron Imaging. 1999; 2(3):159–171. pmid:14516540
- 17. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009; 45(2):228–247. pmid:19097774
- 18. Huyge V, Garcia C, Alexiou J, Ameye L, Vanderlinden B, Lemort M et al. Heterogeneity of metabolic response to systemic therapy in metastatic breast cancer patients. Clin Oncol (R Coll Radiol). 2010; 22(10):818–827.
- 19. Hendlisz A, Deleporte A, Delaunoit T, Marechal R, Peeters M, Holbrechts S et al. The Prognostic Significance of Metabolic Response Heterogeneity in Metastatic Colorectal Cancer. PLoS One. 2015; 10(9):e0138341. pmid:26421426
- 20. Janssen MH, Ollers MC, van Stiphout RG, Riedl RG, van den Bogaard J, Buijsen J et al. PET-based treatment response evaluation in rectal cancer: prediction and validation. Int J Radiat Oncol Biol Phys. 2012; 82(2):871–876. pmid:21377810
- 21. Sun W, Xu J, Hu W, Zhang Z, Shen W. The role of sequential 18(F) -FDG PET/CT in predicting tumour response after preoperative chemoradiation for rectal cancer. Colorectal Dis. 2013; 15(5):e231–e238. pmid:23384167
- 22. Maffione AM, Ferretti A, Chondrogiannis S, Rampin L, Marzola MC, Grassetto G et al. Proposal of a new 18F-FDG PET/CT predictor of response in rectal cancer treated by neoadjuvant chemoradiation therapy and comparison with PERCIST criteria. Clin Nucl Med. 2013; 38(10):795–797. pmid:24107808
- 23. Skougaard K, Nielsen D, Jensen BV, Hendel HW. Comparison of EORTC criteria and PERCIST for PET/CT response evaluation of patients with metastatic colorectal cancer treated with irinotecan and cetuximab. J Nucl Med. 2013; 54(7):1026–1031. pmid:23572497
- 24. Koolen BB, Valdes Olmos RA, Wesseling J, Vogel WV, Vincent AD, Gilhuijs KG et al. Early assessment of axillary response with (1)(8)F-FDG PET/CT during neoadjuvant chemotherapy in stage II-III breast cancer: implications for surgical management of the axilla. Ann Surg Oncol. 2013; 20(7):2227–2235. pmid:23456316
- 25. Gebhart G, Gamez C, Holmes E, Robles J, Garcia C, Cortes M et al. 18F-FDG PET/CT for early prediction of response to neoadjuvant lapatinib, trastuzumab, and their combination in HER2-positive breast cancer: results from Neo-ALTTO. J Nucl Med. 2013; 54(11):1862–1868. pmid:24092940
- 26. Lastoria S, Piccirillo MC, Caraco C, Nasti G, Aloj L, Arrichiello C et al. Early PET/CT scan is more effective than RECIST in predicting outcome of patients with liver metastases from colorectal cancer treated with preoperative chemotherapy plus bevacizumab. J Nucl Med. 2013; 54(12):2062–2069. pmid:24136935
- 27. De BS, Van DN, Smeets P, Ferdinande L, Ceelen W, Mertens J et al. Value of DCE-MRI and FDG-PET/CT in the prediction of response to preoperative chemotherapy with bevacizumab for colorectal liver metastases. Br J Cancer. 2012; 106(12):1926–1933. pmid:22596235
- 28. Farnebo J, Gryback P, Harmenberg U, Laurell A, Wersall P, Blomqvist LK et al. Volumetric FDG-PET predicts overall and progression- free survival after 14 days of targeted therapy in metastatic renal cell carcinoma. BMC Cancer. 2014; 14:408. pmid:24906441
- 29. Liu FY, Yen TC, Wang JY, Yang TS. Early prediction by 18F-FDG PET/CT for progression-free survival and overall survival in patients with metastatic colorectal cancer receiving third-line cetuximab-based therapy. Clin Nucl Med. 2015; 40(3):200–205. pmid:25608159
- 30. de Langen AJ, Vincent A, Velasquez LM, van TH, Boellaard R, Shankar LK et al. Repeatability of 18F-FDG uptake measurements in tumors: a metaanalysis. J Nucl Med. 2012; 53(5):701–708. pmid:22496583
- 31. Lee SJ, Kim JG, Lee SW, Chae YS, Kang BW, Lee YJ et al. Clinical implications of initial FDG-PET/CT in locally advanced rectal cancer treated with neoadjuvant chemoradiotherapy. Cancer Chemother Pharmacol. 2013; 71(5):1201–1207. pmid:23420438
- 32. Lodge MA, Chaudhry MA, Wahl RL. Noise considerations for PET quantification using maximum and peak standardized uptake value. J Nucl Med. 2012; 53(7):1041–1047. pmid:22627001
- 33. Frings V, van Velden FH, Velasquez LM, Hayes W, van de Ven PM, Hoekstra OS et al. Repeatability of metabolically active tumor volume measurements with FDG PET/CT in advanced gastrointestinal malignancies: a multicenter study. Radiology. 2014; 273(2):539–548. pmid:24865311