The aim of the present paper is to review the scientific literature concerning the usefulness of 18F-FDG PET/CT in the evaluation of response to chemotherapy in patients affected by liver metastases from colorectal cancer.
18F-PET/CT is a well-established imaging modality in oncology, widely used to stage, restage, and follow up several malignancies, including colorectal cancer (CRC) [
18F-FDG PET/CT is a hybrid technique, which associates the molecular imaging PET with CT. In spite of the examination’s result, which resembles a particular CT scan, the major strength of this technique originates from its molecular part, that is, the possibility to evaluate the total amount of tumor metabolism through its consumption of radioactive glucose. The specific power of 18F-FDG PET/CT to measure tumors’ metabolism, and therefore its induced therapy alteration, makes it a theoretically ideal marker of treatment responsiveness.
Here we present a selected review of the scientific literature concerning the usefulness of 18F-FDG PET/CT in the evaluation of response to systemic therapy in patients affected by liver metastases from colorectal cancer (CRCLM).
Computed Tomography (CT) is still the first choice method to evaluate the response of CRLM both in routine clinical practice and in clinical trials. The use of CT scan to evaluate the responsiveness of a neoplasm to a certain antineoplastic agent is based on the assumption that if the treatment is effective, its cytostatic/cytotoxic action will eventually induce a lowering in neoplastic mass, which is measurable by conventional radiologic equipment. Following this statement, many scientific organizations devoted to cancer treatment developed criteria aimed to define a generally accepted base to measure the response in cancer treatment [
RECIST is an international standard criteria based on a simplification of former methods (WHO, ECOG). The prerequisite for RECIST criteria evaluation is based on the presence of quantifiable disease; that is, it is mandatory that at least one lesion be measurable. These organizations offer a simplified, conservative extraction of imaging data for wide application in clinical trials. They presume that linear measures are an adequate substitute for 2D methods. RECIST records four response categories, that is, (a) complete response (CR, disappearance of all target lesions), (b) partial response PR = 30% decrease in the sum of the longest diameter of target lesions, (c) progressive disease (PD = 20% increase in the sum of the longest diameter of target lesions), and (d) stable disease (SD = small changes that do not meet the above criteria).
CT evaluation in all its multiform applications has gained an extensive use due to its reproducibility and widespread distribution. However, the geometrical mechanism at the base of morphologic imaging response is forcibly subsequent to previous molecular, biochemical, and, ultimately, metabolic changes. This characteristic has a medium to low impact on the assessment of cytotoxic drugs but may result in a reduced accuracy when dealing with the new cytostatic biological class of pharmaceuticals [
Moreover, most of the studies suggest a CT evaluation after at least three cycles of chemotherapy.
Indeed, the tumor’s shrinkage may be minimal even when treatment is effective, particularly with cytostatic drugs. The difficulties of CT imaging to evaluate the lack of response after one or two cycles make it impossible to redirect patients towards a more effective strategy, with obvious additional clinical and financial costs.
Along with these concerns about the chronological limits of CT, recent reports hinted at the possibility that a reduction in the density of liver parenchyma on portal venous scans after systemic therapy due to a toxic impact on liver parenchyma could result in reduced tumor-to-liver contrast and to the underestimation of real lesion size [
In contrasts to TC, which is a predominantly morphologic technique, Magnetic Resonance Imaging may include a more significant part of functional information. Magnetic resonance spectroscopy (MRS) and dynamic contrast-enhanced (DCE) and diffusion-weighted (DW) MRI may be used to evaluate molecular, biological, and, eventually, functional modifications induced by treatments. In the path toward a more personalized medicine of CRCLM treatment, DW-MRI has shown promising results as an early predictor of response in patients undergoing chemotherapy.
Some studies have shown that baseline apparent diffusion coefficient (ADC) measurements predict therapeutic benefit, with higher ADC metastases responding poorly to chemotherapy [
Another item to be clarified concerns the sensitivity of DW-MRI for metastases with different model of vascular pattern. Hypervascular metastases have shown lower ADC values compared to hypovascular metastases as in most of the cases CRLM are [
18F-fluorodeoxyglucose (FDG) is a glucose analog that is mostly taken up in malignant cells because of their higher glucose metabolism. Standalone 18F-FDG PET first, and PET/CT later on, was shown to be effective for initial staging and follow-up in oncology patients affected by most prevalent tumors (lung cancer, breast cancer, colorectal cancer, and lymphoma). These diagnostic performances warranted a strong role to 18F-FDG PET/CT in the diagnosis and staging work-up of many neoplasms. If we take a look to the more endorsed application of 18F-FDG PET/CT, it is clear that the philosophy of PET/CT use is somehow unbalanced toward anatomic, topographic, and morphologic parameters (the location of the uptake areas, their relationship with the surrounding anatomic structures, etc.). In the first few years of its clinical use, the only 18F-FDG PET/CT molecular characteristics, which have been mainly utilized, were the Standard Uptake Value (SUV) in the prognostic stratification. Indeed, the higher uptake of radioactive glucose is associated with the most biologically aggressive forms of the neoplasms [
Hence, although anatomic interpretation of PET scans has been shown to be the basis of the clinical report, the development of strategies directed to gain quantitative information should allow more objective diagnosis and, above all, the comparisons between serial PET of a patient.
The rationale for the use of 18F-FDG PET/CT in the evaluation of the response to systemic treatments is based on amplified glucose metabolism characteristic of neoplastic cells [
The theoretical construction for 18F-FDG PET/CT tumor’s response has been extensively cleared up in the first decade of this century [
18F-FDG PET/CT may reach the highest outcome for CRCLM in two distinct clinical scenarios, that is, the prognostic stratification after preoperative chemotherapy and the early evaluation of systemic treatment irrespectively of the following treatments. This quite rough schematization is not an end in itself, but it follows the clinics which group patients with CRLM in three separate groups: (1) those with easily resectable disease, (2) those with borderline resectable or high recurrence risk CRLM, and (3) those with inoperable but liver limited CRLM. For the first group of patients the standard of care therapy is surgery, followed by adjuvant therapy if considered. The second group of patients should be treated with systemic neoadjuvant followed by liver surgery. The third group of patients, those with inoperable CRLM, should be offered the most effective systemic therapy with the goal to reach maximal disease response with the intention of conversion to surgical resectability with curative intent [
Surgery together with systemic chemotherapy is the only cure for patients with CRCLM [
The other field in which 18F-FDG PET/CT may play a fundamental role is the early assessment of responsiveness to systemic treatments. This is particularly true for patients treated with the novel molecular targeted drugs which have a predominantly cytostatic effect. These treatments act mainly by halting tumor growth rather than eliminating neoplastic cells. Due to their mechanism of action, therefore, molecular targeted drugs stabilize rather than kill malignant cells. Thus, despite active treatments, conventional imaging may show some changes or even an increase in tumor size, especially in the first phases of therapy, due to inflammatory changes. 18F-FDG PET/CT may bridge the gap between the start of cytostatic treatments effect and the response evaluation. Changes in neoplastic glucose consumption have been noticed as early as 24 hours after a dose of treatment in Gastrointestinal Stromal Tumors [
Keeping in mind the lesson of GIST, some authors evaluated the possible role of 18F-FDG PET/CT in early response assessment of CRCLM with the goal to redirect nonresponding patients towards a more effective treatment. Generally, 18F-FDG PET/CT was carried out after 1 or 2 cycles of therapy, usually scheduled with Folfox-Folfiri plus bevacizumab [
As ever, nuclear medicine is the vocation to give quantitative answers. This statement was true dealing with gamma cameras, but it has largely been emphasized with PET and PET/CT. Many PET/CT parameters have been studied in clinical practice. SUV represents an index for FDG uptake in tissues. This parameter was studied in the last decade of the last century [
SUV is a quantification of normalized radioactivity concentration in PET images. SUV is calculated drawing a bidimensional (ROI) or tridimensional (VOI) region of interest inside the tumor lesion using software.
The measured radioactivity is then normalized to the average total radioactivity present in the body, which is hypothesized as the injected dose divided by the patient body weight (SUVbw) or the patient lean body mass (SUVlbm or SUL) or body surface area (SUVbsa). SUVmax in a VOI is the most used parameter because it is probably more reproducible due to the self-determining ROI/VOI assessment. SUVpeak is a composite measurement calculating the local average SUV within all the voxels close to the one with the highest radioactivity.
The most frequent source of error affecting SUV family measurement is 18F-FDG extravasation. When 18F-FDG PET/CT is repeated in the same patient to evaluate the outcome of a treatment it is mandatory both to inject the same activity of the radiopharmaceutical and to scan the patient at the same interval.
Other more recent methods to evaluate tumor metabolism are Total Lesion Glycolysis (TLG) [
In summary, 18F-FDG PET/CT brings with it the opportunity to measure a number of parameters, each one mirroring a specific molecular, biologic, and metabolic feature. But how have all these pieces of information been considered in clinical guidelines?
The first attempt to do so was carried out by European Organization Research and Treatment of Cancer (EORTC) in 1999 [
EORTC response criteria for 18F-FDG PET/CT.
CMR | Complete resolution of [18F]-FDG uptake within the tumour volume so that it was indistinguishable from surrounding normal tissue |
|
|
PMR | Reduction of a minimum of 15 ± 25% in tumour 18F-FDG SUV after one cycle of chemotherapy and greater than 25% after more than one treatment cycle |
|
|
SMD | Increase in tumour 18F-FDG SUV of less than 25% or a decrease of less than 15% and no visible increase in extent of 18F-FDG tumour uptake (20% in the longest dimension) |
|
|
PMD | Increase in 18F-FDG tumour SUV of greater than 25% within the tumour region defined on the baseline scan and visible increase in the extent of 18F-FDG tumour uptake (20% in the longest dimension) or the appearance of new 18F-FDG uptake in metastatic lesions |
The cornerstone of nuclear medicine therapy response evaluation could be represented by the paper of Wahl et al. [
A recent paper has compared EORTC and PERCIST criteria in the assessment treatment with irinotecan and cetuximab in CRCLM [
Comparison of the two methods to assess the response and overall survival (OS).
EORTC (patients) | OS (months) | PERCIST (patients) | OS (months) | |
---|---|---|---|---|
CMR | 0 | n.a. | 0 | n.a. |
PMR | 38 | 14.2 | 34 | 14.5 |
SMD | 16 | 6.4 | 20 | 6.9 |
PMD | 7 | 12.2 | 7 | 12.2 |
SMD + PMD | 23 | 7.2 | 27 | 7.9 |
The study evidenced that patients with a PR as assessed by 18F-FDG PET/CT have a significantly better median OS compared with those grouped as SD + PD, regardless of which criteria have been used. For all the other figures of merit, both criteria have very similar results, probably because their measurements are restricted to the most metabolically active part of the patient’s tumor burden. Moreover, the lack of discrepancies provides excellent credentials to the measurement of SUV for treatment response. The authors concluded that PERCIST criterion is somehow preferable because it is more defined and less operator-dependent in nature.
The prognosis of patients affected by CRCLM is heavily influenced by the response to treatments. 18F-FDG PET/CT is a hybrid imaging method, which holds the power of both anatomic and molecular imaging. This double characteristic makes 18F-FDG PET/CT the ideal tool to evaluate the outcome antineoplastic therapies. The results deduced from our analysis of the scientific literature are promising and should spur the researchers towards more structured, possibly multicentric, and prospective trials.
The authors declare that there is no conflict of interests regarding the publication of this paper.