Background
About one third of patients with non-small cell lung cancer (NSCLC) present loco-regionally advanced disease at the diagnosis [
1,
2], and despite radical treatment with concurrent chemo-radiotherapy (chemo-RT), only 15% of patients will be long-term survivors and 15%–40% will develop loco-regional tumor recurrence [
3,
4]. A higher biologically effective dose of radiotherapy can improve loco-regional control and survival [
5]: however, an escalating radiotherapy dose also results in increasing the risk of toxicity [
6]. For this reason, it is important to carefully select patients for radiotherapy dose intensification. Currently, the response to radiotherapy is not determined until the therapy has been completed. If the individual response to radiotherapy could be evaluated earlier during treatment, a timely therapy modification could be accomplished to better adapt the cure. Molecular imaging offers the potential to characterize the nature of tissues on the basis of its biochemical and biologic features.
18 F-fluoro-2-deoxyglucose (
18 F-FDG) positron emission tomography integrated with computed tomography (
18 F-FDG PET-CT) is largely used in oncology, especially for monitoring the response to treatment. The imaging of changes in glucose metabolism, as reflected by cellular uptake and trapping of
18 F-FDG, can provide a response assessment that is both more timely and more accurate than that provided by standard morphological imaging [
7]. Furthermore, the residual metabolic activity of tumors after radiotherapy, as measured by
18 F-FDG uptake, has been shown to correlate with the pathologic response [
8], and to be a significant prognostic factor for survival in patients with NSCLC [
9‐
11]. Many researchers recommend a delay of 6–8 weeks or longer after radiotherapy before performing the post-treatment PET study because of inflammatory changes with subsequent alterations in
18 F-FDG uptake [
12]. Nevertheless, the confounding effect in the surrounding normal tissue due to the radiation-induced elevation of
18 F-FDG activity in the lung seems to be less relevant when PET is performed during radiotherapy [
13]. The objectives of this study were: to evaluate the metabolic changes on serial
18 F-FDG PET-CT studies performed before, during and after concurrent chemo-radiotherapy in patients with unresectable or locally advanced NSCLC; to correlate the metabolic changes with the delivered radiation dose and with the clinical outcome.
Discussion
18 F-FDG PET detects metabolic modifications which are well known to occur before morphologic ones; therefore functional imaging allows an evaluation of the tumor metabolic response during radiotherapy earlier than morphologic imaging [
13,
16,
17]. Changes in therapy, such as dose-escalation or the addition of another treatment modality may be contemplated for patients who show poor response to the current radiotherapy regimen [
18]. Furthermore, a disease reduction during radiotherapy detected by functional imaging might suggest a radiation dose escalation, whilst remaining within normal tissue constraints [
19]. The role of
18 F-FDG in assessing the response to non-surgical treatment in patients with NSCLC has been extensively investigated. Many Authors performed an
18 F-FDG scan at least two months from the end of radiotherapy or chemotherapy [
10,
20‐
22], while only a few have evaluated the potential role of repeating
18 F-FDG PET earlier, either during or after therapy [
13,
17,
23,
24]. Researchers from the University of Michigan [
13] have performed repeated
18 F-FDG PET in 15 patients with NSCLC stages I to III before, during and after a course of radiotherapy (or chemo-radiotherapy) with conventional fractionation. The Authors have observed a reduction of the peak tumor
18 F-FDG activity at approximately 45 Gy during radiotherapy, with a further reduction on PET performed three months from the end of treatment. The qualitative response during radiotherapy correlated with the overall response post-radiotherapy, and the peak tumor
18 F-FDG activity during radiotherapy correlated with that 3 months post-therapy. Subsequently, the same Authors [
24] have observed that an adaptation of the radiotherapy plan based on
18 F-FDG PET at approximately 45 Gy during radiotherapy, might allow escalating the tumor dose without increasing the normal tissue complication probability. Van Baardwijk et al. [
17] investigated changes in
18 F-FDG uptake in 23 patients with medically inoperable or advanced NSCLC who underwent serial
18 F-FDG PET-CT scans before treatment, 1 week and 2 weeks after the start of treatment and 70 days from the end of an accelerated (1.8 Gy twice a day) radiation treatment with radical intent. While 70 days after the end of radiotherapy, the metabolic activity of the tumor significantly decreased compared to the baseline, during the first 2 weeks of treatment the
18 F-FDG uptake within the tumor changed only moderately (a slight increase during the first week, and a slight decrease during the second week). More recently, Giovacchini et al. [
25] performed four repeated
18 F-FDG PET-CT scans in 6 patients undergoing radical radiation treatment for either locally advanced or medically inoperable NSCLC.
18 F-FDG PET-CT scans were performed before, during radiotherapy at the delivered dose of 50 Gy, and after approximately one month and 3 months from the end of radiotherapy. Radiotherapy induced a progressive decrease in glucose metabolism that was greater 3 months after the end of treatment, but could even be detected during the treatment itself.
In this study, we have evaluated the metabolic changes on serial
18 F-FDG PET-CT performed before, during and after concurrent chemo-radiotherapy in patients with unresectable or locally advanced non-small-cell-lung-cancer (NSCLC). We have also correlated the metabolic changes with the delivered radiation dose, and with the clinical outcome.
18 F-FDG PET-CT studies were performed earlier both during treatment, at a median time of 17 days from the start (median dose of 23.4 Gy), and after treatment at a median time of 30 days. First of all, we observed that at pre-RT PET-CT the values of SUVmax of the tumors were much higher than those reported in literature [
13,
17], similar only to those reported by Giovacchini et al. [
25]. The enhanced trapping of
18 F-FDG into the tumor cells can be due to either biological mechanisms, such as the up-regulation of glucose transporters and hexokinase enzymes, tumor aggressivity, hypoxia, etc., or to modifications induced by previous treatment [
26‐
29]. Up to now, however, it is not known which of these mechanisms is responsible for the variable levels of
18 F-FDG uptake. In our study, the large tumor size, containing small areas of necrosis, and the relatively small number of patients who received previous treatment could explain the higher values of SUV at pre-RT PET-CT. From our data, we can observe that the tumor metabolic activity significantly decreased early during chemo-RT, and decreased even more at the end of treatment. The metabolic reduction was significant for all parameters, but more significant for the SUVmax. It can be argued that chemo-radiotherapy works better and faster in the cellular metabolism, when considering SUV values, and “relatively” less and more slowly in the tumor volume, based on MTV values. In fact MTV may be considered a “functional volume” and, as such, reduces its activity later in comparison with SUV. Therefore, SUVmax is the more sensitive parameter to show an earlier metabolic modification induced by the treatment. The difference in SUVmax reduction between pre-RT PET-CT and during-RT PET-CT was significantly higher (p = 0.0001) than that observed between during-RT PET-CT and post-RT PET-CT (p = 0.005). This interesting finding allow us to speculate that the tumor cells have a prompt response to treatment. On the contrary, the MTV reduction was similar between pre-RT PET-CT and during-RT PET-CT, as well as between during-RT PET-CT and post-RT PET-CT (p = 0.002, respectively), supporting the concept that MTV is a functional volume and its response to therapy is slower. Regarding the lymph nodes, their metabolic activity tends to decline during chemo-radiotherapy, decreasing significantly only at the end of treatment. This finding suggests the stronger lymph node resistance to therapy: in fact it is well known that the neoplastic lymph nodes are a negative prognostic factor especially in patients with NSCLC [
30].
Finally, the tumor metabolic activity significantly decreased after a cumulative radiotherapy dose of only 23.4 Gy, which is much lower than that reported in literature: 45 Gy and 50 Gy [
13,
25]. On the other hand, van Baardwijk et al. [
17] did not observe any significant decrease in tumor metabolic activity after the delivery of approximately 37 Gy of accelerated radiotherapy (1.8 Gy twice a day). Differences in the radiotherapy fractionation schedule, treatment time, concurrent chemotherapy administration, tumor biology, and absolute pre-RT PET-CT SUVmax values, might have an impact on tumor
18 F-FDG uptake during radiotherapy. Similarly to Kong et al. [
13] and van Baardwijk et al. [
17], we observed a large heterogeneity in the changes in metabolic activity among individual patients during and after radiotherapy: this finding may somehow reflect the difference in radio-responsiveness between the individual tumors. Similarly to these Authors [
13,
17], we also observed a large heterogeneity in the metabolic activity among the individual patients before treatment, suggesting a large cellular heterogeneity in each tumor: for this reason we have also utilized ΔSUV and ΔMTV in order to take into account the “individual” variations during and at the end of treatment, rather than only the “absolute” values such as SUVmax and MTV. In fact, the individual variations allow a better measurement of the effects of the treatment “normalizing” the baseline values, especially when they are highly heterogeneuos.
Regarding the metabolic response after treatment, the EORTC criteria [
16] classifies the metabolic response on the basis of SUV values. The classification into four categories using only a number as cut-off, may sometimes give an incorrect classification. In fact, the number does not take into account some scintigraphic features, such as the distribution and shape of the
18 F-FDG activity. From our data, the high number of partial metabolic responses (>50%) can be also attributed to the use of strict SUV criteria, as proposed by EORTC. Therefore, from a clinical point of view, we strongly support the necessity to integrate the SUV values with qualitative and morphological (CT) analyses of the images since the confounding effects such as inflammation may be present at any time after treatment. A wrong classification might therefore be avoided and a clinical significance could be given to the
18 F-FDG uptake. Both Kong et al. [
13] and van Baardwijk et al. [
17], observed an association between tumor metabolic response during radiotherapy and that post treatment, with different patterns of response during radiotherapy for patients with a complete metabolic response, and patients with a persistence of metabolic activity after the therapy. Our results do not confirm these findings. We found a borderline statistical significant difference (p = 0.05) only in SUVmax of the tumor during treatment: non-responders showed higher value of SUVmax of the tumor at during-RT PET-CT when compared to responders patients. Unfortunately, our group sample is too small to speculate on this finding. Moreover the wide range in the cumulative radiation dose delivered until the moment of during-RT PET-CT acquisition (14.4-34.2 Gy) is likely to have contributed to the heterogeneity in metabolic response, and may represent a major drawback for this study. Indeed, we observed a significant correlation between SUVmax measured during-RT PET-CT and the cumulative dose of radiotherapy delivered at the moment of the scan acquisition. Although this finding is not surprising, and is in accordance with the well-known association between the radiation dose delivered and the probability of cure in NSCLC [
31], this is the first time that it is clearly shown in a clinical setting. Further investigations of this association, i.e. describing tumor activity as a function of pre-treatment activity, radiotherapy dose delivered, and time since the beginning of radiotherapy, may prove to be very interesting. For example, dose-SUV curves could be elaborated using experimental data to extrapolate the total radiation dose required to obtain a complete metabolic response in each patient, thus making it possible to adapt the radiation dose prescription.
A significant correlation between the residual
18 F-FDG uptake within the tumor at the end of treatment (or change in
18 F-FDG uptake in respect to the baseline), and survival end-points has been described by several Authors [
10,
17]. Finally, also in our experience, patients with loco-regionally advanced NSCLC who showed a complete metabolic response at post-RT PET-CT had a longer disease-free survival when compared with those with a persisting
18 F-FDG uptake. The main limitation of this study is the small group of patients. More studies on a larger number of patients are necessary to confirm our findings.
Competing interest
The authors declare no conflict of interest.
Authors’ contributions
MM, MLC, VV, AG: conception and design. MM, MLC, MGS, MF acquisition of data, MM, MLC, FC analysis of data. MM, MLC alignment and drafted the manuscript. All authors read and approved the final manuscript.