- Email: [email protected]
Serial [18F]-fluoromisonidazole PET during radiochemotherapy for locally advanced head and neck cancer and its correlation with outcome
Radiotherapy and Oncology 117 (2015) 113–117
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
MISO PET in head and neck
Serial [18F]-fluoromisonidazole PET during radiochemotherapy for locally advanced head and neck cancer and its correlation with outcome Nicole E. Wiedenmann a, Sabine Bucher a,c, Michael Hentschel d, Michael Mix b, Werner Vach e, Martin-Immanuel Bittner a, Ursula Nestle a, Jens Pfeiffer f, Wolfgang A. Weber b,g,1, Anca L. Grosu a,⇑,1 a Department of Radiation Oncology, German Cancer Consortium/Partner site Freiburg; b Department of Nuclear Medicine, University Medical Center Freiburg, Germany; c Clinic of Radiology & Nuclear Medicine, University Hospital Basel; d Department of Nuclear Medicine, Inselspital Bern, Switzerland; e Center for Medical Biometry and Medical Informatics; f Department of Otorhinolaryngology – Head and Neck Surgery, University Medical Center Freiburg, Germany; and g Molecular Imaging and Therapy Service, Memorial Sloan-Kettering Cancer Center, New York, USA
a r t i c l e
i n f o
Article history: Received 23 February 2015 Received in revised form 31 August 2015 Accepted 6 September 2015 Available online 29 September 2015 Keywords: Hypoxia Head and neck cancer 18F-MISO PET Radiochemotherapy
a b s t r a c t Purpose: The aim was to assess changes of tumour hypoxia during primary radiochemotherapy (RCT) for head and neck cancer (HNC) and to evaluate their relationship with treatment outcome. Material and methods: Hypoxia was assessed by FMISO-PET in weeks 0, 2 and 5 of RCT. The tumour volume (TV) was determined using FDG-PET/MRI/CT co-registered images. The level of hypoxia was quantified on FMISO-PET as TBRmax (SUVmaxTV/SUVmean background). The hypoxic subvolume (HSV) was defined as TV that showed FMISO uptake P1.4 times blood pool activity. Results: Sixteen consecutive patients (T3–4, N+, M0) were included (mean follow-up 31, median 44 months). Mean TBRmax decreased significantly (p < 0.05) from 1.94 to 1.57 (week 2) and 1.27 (week 5). Mean HSV in week 2 and week 5 (HSV2 = 5.8 ml, HSV3 = 0.3 ml) were significantly (p < 0.05) smaller than at baseline (HSV1 = 15.8 ml). Kaplan–Meier plots of local recurrence free survival stratified at the median TBRmax showed superior local control for less hypoxic tumours, the difference being significant at baseline and after 2 weeks (p = 0.031, p = 0.016). Conclusions: FMISO-PET documented that in most HNC reoxygenation starts early during RCT and is correlated with better outcome. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 117 (2015) 113–117
Hypoxia is a characteristic feature in HNC and has been identified as a factor negatively correlated with prognosis [1–3]. Tumour hypoxia is known to increase radiation resistance, has been identified as a key factor in triggering hypoxia-mediated gene expression and was found to be associated with a more aggressive tumour phenotype [1]. There are a significant number of trials showing that hypoxia-PET may serve as a tool to identify patients potentially at increased risk of local failure [3]. However, the question whether hypoxia-PET may serve as a basis for radiation treatment planning by dose painting in intensity modulated radiation therapy (IMRT) is still open [2–6]. A central problem is the poor knowledge regarding the variation of tumour hypoxia during R (C)T and the possible implication of this variation on radiation treatment planning. There are only limited data on the temporal changes of hypoxia based on PET imaging during R(C)T and its correlation with clinical ⇑ Corresponding author. 1
Equal contribution.
http://dx.doi.org/10.1016/j.radonc.2015.09.015 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
outcome. Eschmann et al. [7] investigated 14 patients with HNC on FMISO-PET before and under radiotherapy and observed a significant tumour reoxygenation under treatment (ca. 45 Gy), suggesting a correlation with a better tumour control but without statistical significance. The group of Leuven [8] assessed the variation of hypoxia on FMISO-PET in 15 HNC patients. They also described a significant re-oxygenation during RCT. Additionally they showed that hypoxia is one of the most important factors affecting RCT treatment resistance: not only the pretreatment level of hypoxia but also the extent during radiotherapy (30 Gy) correlated significantly with disease control. Zips et al. [9] investigated the temporal changes of tumour hypoxia in 25 HNC patients treated with RCT. In an exploratory prospective trial the authors showed that FMISO-PET in the first or second week of RCT has a strong prognostic value for diagnosing patients with high risk of tumour recurrence. Interestingly, PET before treatment was not as strong in comparison. The authors postulate that different treatment related tumour reoxygenation profiles lead to differences in treatment response. A prospective hypoxia imaging study
114
Dynamics of hypoxia in head and neck tumors assessed by [18]F-Miso PET
including 50 patients is on-going. The group of Aarhus [10] investigated on FAZA-PET/CT 40 patients with HNC cancer included in the DAHANCA 24 trial. In 13/40 patients a FAZA-PET/CT was performed about 14 days after the start of RCT. The HSV during treatment became significantly smaller and was located in the same region assessed at the first and second PET investigation. The authors observed a higher recurrence rate in the hypoxic group (4/6 patients) in comparison to the group showing no hypoxia under treatment (2/7 patients), however statistical significance was not reached. Our group assessed the distribution of hypoxia in HNC on FAZAPET [4], showing that in the majority of patients the hypoxic tissue was located in a single confluent region and therefore a HSV for radiation treatment planning could be delineated based on hypoxia-PET. Only in a small number of cases hypoxia was diffusely dispersed or the tumours were not hypoxic. In a more recent study [11] we have evaluated the changes in the location of the HSV on FMISO-PET during RCT and have concluded that in patients with persistent hypoxia after 2 weeks of treatment, the HSV is topographically relatively stable. The goal of this prospective feasibility study was to evaluate the temporal changes of tumour hypoxia during RCT, to correlate these data with the clinical outcome and, considering our results and published data, to discuss their possible impact on the design of new treatment strategies.
Image analysis PET data were iteratively reconstructed to voxels of 4.3 4.3 3.4 mm3 using an ordered-subset expectation–maximi zation (OSEM) reconstruction algorithm. Scatter and attenuation correction was applied. Hypoxic volumes were determined using PMOD software (PMOD-Group, Switzerland). Tumour volume (TV) delineation, separately for primary tumour and pathological lymph nodes, was performed on FDG-PET/CT/MRI co-registered images. In accordance with the literature, the TV was determined on FDG-PET by a 40% threshold of the maximum standardized uptake values (SUVs) within TV. The intensity of FMISO uptake was assessed quantitatively by calculating the ratio of the maximum SUV in the tumour (TV) to the mean SUV in contralateral neck musculature (TBRmax). By visual assessment, a TBRmax of P1.4 was considered most appropriate to define a tumour as being hypoxic. For statistical analysis, median TBRmax values were applied to split according to oxygenation status (median TBRmax for FMISO 1, 2, 3: 2.02, 1.26, 1.24). The HSV was determined by normalizing against blood activity concentration by counting all voxels within the TV for which the tumour to blood ratio on FMISO-PETs was P1.4 [12]. The blood activity concentration was determined from a region of interest in the left ventricle. The HF was defined as the ratio of HSV/TV.
Statistical analysis Materials and methods Patients’ characteristics Sixteen consecutive patients undergoing definitive RCT for locally advanced HNC (squamous cell carcinoma) were enrolled. Inclusion criteria were histologically confirmed T3/4, N+, M0 squamous cell carcinomas of the oral cavity, oropharynx, hypopharynx, and larynx and age P18.
Statistical analyses were performed using SPSS (v. 15.0.1.) and Stata/IC 12. p-values <0.05 were considered statistically significant. Comparison of continuous outcomes between subgroups is based on paired Wilcoxon-tests. The Wilcoxon matched-pairs signedranks test was used to compare TBRmax and HSVs between PET scans. The Mann–Whitney-U test was applied to compare mean TBRmax between recurrent and non-recurrent tumours. The association of various PET parameters with the time to recurrence was assessed using Kaplan–Meier plots based on splitting the values at the median, together with p-values from a log-rank test.
Study design A prospective feasibility FMISO-PET imaging monocentric study was conducted. The institutional review board approved the study and all participants provided written informed consent. Tumour hypoxia was assessed by FMISO-PET prior to (FMISO1) and during the course of RCT, at weeks 2–3 (FMISO2) and weeks 5–6 (FMISO3). All patients received planning CT, MRI (MRI1) and [18F]fluorodeoxyglucose (FDG)-PET imaging prior to RCT. MRI was repeated in treatment weeks 5–6 (MRI2). Radiation treatment was delivered as conformal IMRT (2 Gy/d, 5/week; total dose to macroscopic tumour 70 Gy). Cisplatin was administered in weeks 1, 4, 7 (100 mg/m2). For followup, patients were evaluated clinically and by MRI every 3 months.
PET/CT [18F]-MISO and [18F]-FDG production met standard quality criteria. For FMISO-PET, one bed position, covering the head and neck region examined 150 min after injection of approx. 400 MBq [18F]MISO. Scan duration was 35 min (3 frames at 10 min, followed by a 5 min transmission scan). For quantitative analysis, three attenuation corrected frames were summed after excluding frames with patient movement. Planning CT images were acquired in 2-mm slice thickness. For FDG-, FMISO-PET and CT, patients were immobilized identical to the radiation position with a head and neck mask.
Results 15/16 patients (94%) showed hypoxic tumour tissue on the baseline FMISO-PET scan: 11 in the primary tumour only, 2 in lymph nodes only, 2 in both. The number of patients with hypoxic lesions decreased to 5/14 (36%) at the second and to 3/11 (27%) at the third FMISO-PET investigation. Mean TBRmax significantly decreased from 1.94 (pre-treatment) to 1.57 (week 2, p = 0.035) and 1.27 (week 5, p = 0.003) (Table 1, Fig. 1). The median TBRmax dropped from 2.02 (pre-treatment) to 1.26 (week 2) and 1.24 (week 5). Mean and median TV assessed by FDG-PET/CT/MRI and HSV, HF on FMISO-PET at baseline and in week 2 and 5 are presented in Table 2. Mean follow-up was 31 months (median 44, range 1–53 months). The time course of FMISO TBRmax was strikingly different for patients with and without recurrence (Figs. 1 and 2, Table 1). While patients without tumour recurrence demonstrated a rapid and continuous decrease in the TBRmax, the decline was slower in patients with recurrence and some patients even showed an increase in tumour hypoxia at the time of the second FMISO-PET. No local recurrence occurred if there was at least one negative FMISO scan (TBRmax < 1.4, Fig. 1). For the three consecutive FMISO-PET scans, patients with tumour recurrence were found to have higher mean TBRmax compared to patients without tumour recurrence: mean TBRmax over all time points: 2.15 vs. 1.45, (Mann–Whitney-U-Test, p = 0.002).
115
N.E. Wiedenmann et al. / Radiotherapy and Oncology 117 (2015) 113–117
Table 1 TBRmax for FMISO1–3, tumour response on MRI, length of follow-up (months), local recurrence status and cause of death, if applicable. Yellow = TBRmax P 1.4, PR = partial remission, SD = stable disease, CR = complete remission, n.a. = not available.
Kaplan–Meier plots of local recurrence free survival showed a significantly lower local tumour control probability for more hypoxic vs. less hypoxic tumours for scans at baseline (FMISO1, median TBRmax = 2.02), p = 0.031 and scans after week 2 (FMISO2, median TBRmax = 1.26), p = 0.016 (Fig. 2). Representative examples are shown in Fig. 3. No correlation was found between the FDG TBRmax and the FMISO TBRmax at baseline (Pearson correlation coefficient = 0.108, p = 0.714). Tumour FDG was not correlated with survival: neither the TBRmax of the primary tumour, nor of lymph nodes, nor the TV were significantly correlated with survival (p = 0.568, p = 0.935 and p = 0.697, respectively). Discussion
Fig. 1. Individual plots of FMISO TBRmax plotted against time for recurrent tumours (red) vs. non-recurrent tumours (black).
In this prospective feasibility study we assess tumour hypoxia on serial FMISO PET scans during RCT and analyse the correlation between different patterns of tumour oxygenation and the risk of local tumour recurrence. Two main findings are made in this study: the first observation is that tumour hypoxia and its temporal changes upon RCT are different among patients. The presence of baseline hypoxia and subsequent resolution of hypoxia upon treatment was the predominant pattern of tumour oxygenation that was observed. However, there were also patients with hypoxia increasing up to week two or with complete absence of tumour hypoxia. Here we define ‘reoxygenation’ as the improvement of tumour oxygenation status upon radiation treatment. Specifically, reoxygenation refers to reduction of the extent of hypoxic volume and reduction of the level of hypoxia over time. The second observation of this study is that the changes of hypoxia assessed by PET during RCT are significantly correlated with outcome. Our observation of a high incidence of baseline tumour hypoxia and subsequent temporal changes with a tendency for early onset
Table 2 TV measured on FDG-PET/CT/MRI before RCT and the mean and median HSV, measured on FMISO PET in weeks 0, 2 and 5 (HSV1, HSV2, HSV3). HF (HSV/TV) for HSV1–3 is designated by HF1, HF2, and HF3, respectively.
Mean (± STD) Median (min–max)
TV (ml)
HSV1 (ml)
HSV2 (ml)
HSV3 (ml)
HF1
HF2
HF3
27.4 (± 22.8) 20.3 (9.4–98.6)
15.8 (±18.4) 11.3 (1–73.8)
5.8 (±13.1) 0.1 (0–44.7)
0.3 (±0.8) 0 (0–2.3)
54.3%
11.8%
1.2%
116
Dynamics of hypoxia in head and neck tumors assessed by [18]F-Miso PET
red: blue:
TBRmax > median TBRmax < median
number at risk months FMISO1 TBRmax < 2.02 TBRmax ≥ 2.02 FMISO2 TBRmax < 1.26 TBRmax ≥ 1.26 FMISO3 TBRmax < 1.24 TBRmax ≥ 1.24
0
12
24
36
48
60
8 8
6 4
5 4
4 4
4 2
0 0
7 7
6 3
6 2
5 2
4 0
0 0
5 6
4 4
3 4
2 4
2 2
0 0
Fig. 2. Kaplan–Meier plots: Local control probability plotted against time after treatment. Patients are stratified at the median TBRmax for FMISO 1, 2 and 3. Red: TBRmax > median, blue: TBRmax < median. FMISO1: p = 0.031, FMISO2: p = 0.016, FMISO3: p = 0.071.
Fig. 3. Representative examples: regression of hypoxia (above, patient#15) vs. increase (week 2) followed by incomplete reduction (week 6) of hypoxia with residual hypoxia remaining (below, patient#16).
N.E. Wiedenmann et al. / Radiotherapy and Oncology 117 (2015) 113–117
of reoxygenation and heterogeneity among individual patients are in agreement with published preclinical data. A substantial proportion of baseline tumour hypoxia and significant reoxygenation after R(C)T have also been described in the preclinical setting using mouse tumour models [13,14] as well as in patients by pO2 histographic measurements [15] before hypoxia-PET became available. Accordingly, tumour hypoxia and tumour reoxygenation are well-known phenomenons in radiation biology. Suggested mechanisms for tumour reoxygenation within fractionated radiation treatment include improved blood perfusion, a decrease of tumour oxygen consumption rate mostly due to a reduced cell number and impaired proliferation, and tumour shrinkage with subsequent improved oxygen diffusion to formerly hypoxic cells. The incidence of baseline hypoxia in our study is comparable with published data, however on the upper limit [6–9,12]. The pattern of tumour oxygenation showed individual patient differences with early reoxygenation as the predominant phenomenon. These findings are in line with the trial of Zips et al. [9] as well as other imaging studies [7,8,10,16,17]. In our study, the median TBRmax for FMISO2 nearly equals the median TBRmax for FMISO3 (1.26 vs. 1.24), reflecting complete reoxygenation taking place early. A possible mechanism at this relatively early time point is a therapy (irradiation + 1 cycle cisplatin) induced reduction of oxygen consumption rather than tumour shrinkage. The chemotherapy regimen included three cycles of cisplatin total (100 mg/kg/d). Thereby one third of the total amount of cisplatin had already been administered before imaging FMISO2, which is a higher total dose than administered by more continuous protocols as used by Zips et al. and could be a factor contributing to a lower TBRmax in week 2 compared to Zips et al. The second observation of this study is the correlation between tumour hypoxia and outcome. Kaplan–Meier plots for tumours stratified at the median TBRmax (2.02, 1.26, 1.24) show superior local recurrence free survival for better oxygenated tumours for all time points (Fig. 2). Statistical significance is reached for the baseline examination (FMISO1, p = 0.031) and the examination obtained after week 2 (FMISO2, p = 0.016). A strong correlation between image parameters and local control was also found by Zips et al. for the week 2 time point, while for the baseline examination significance is only reported on univariate, not at multivariate analysis [9]. The strong correlation between imaging parameters of hypoxia and patient outcome indicates that FMISO-PET may be able to differentiate between patients with high and low risk of local recurrence/failure, i.e. to select early responders from non-responders. This might enable treatment intensification for selected patients or even hypoxia-targeted IMRT regimens. Whether the initial TV, HSV1, 2 or 3 should be boosted remains to be elucidated. Conflict of interest None. Acknowledgements Acknowledgement of grant or other financial support, meeting presentation, or assistance with manuscript preparation or data collection is done:
117
This work was supported by the German Cancer Consortium (DKTK), Partner site Freiburg. The authors thank Prof. Peter Vaupel for critical input on the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2015.09. 015. References [1] Vaupel P, Höckel M, Mayer A. Detection and characterization of tumor hypoxia using pO2 histography. Antioxid Redox Signal 2007 Aug;9:1221–35. [2] Overgaard J. Hypoxic modification of radiotherapy in squamous cell carcinoma of the head and neck – a systemic review and meta-analysis. Radiother Oncol 2011;100:22–32. [3] Horsman MR, Mortensen LS, Petersen JB, et al. Imaging hypoxia to improve radiotherapy outcome. Nat Rev Clin Oncol 2012;9:674–87. [4] Grosu AL, Souvatzoglou M, Röper B, et al. Hypoxia imaging with FAZA-PET and theoretical considerations with regard to dose painting for individualization of radiotherapy in patients with head and neck cancer. Int J Radiat Oncol Biol Phys 2007;69:541–51. [5] Thorwarth D, Alber M. Implementation of hypoxia imaging into treatment planning and delivery. Radiother Oncol 2010;97:172–5. [6] Nehmeh SA, Lee NY, Schröder H, et al. Reproducibility of intratumor distribution of (18)F-fluoromisonidazole in head and neck cancer. Int J Radiat Oncol Biol Phys 2008;70:235–42. [7] Eschmann SM, Paulsen F, Bedeshem C, et al. Hypoia-imaging with 18FMisonidazole and PET: changes of kinetics during radiotherapy of head-andneck cancer. Radiother Oncol 2007;83:406–10. [8] Dirix P, Vandecaveye V, De Keyzer F. Dose painting in radiotherapy for head and neck squamous cell carcinoma: value of repeated functional imaging with 18F-FDG PET, 18F-fluoromisonidazole PET, diffusion-weighted MRI and dynamic contrast-enhanced MRI, and dynamic contrast-enhanced MRI. J Nucl Med 2009;50:1020–7. [9] Zips D, Zöphel K, Abolmaali N, et al. Exploratory prospective trial of hypoxiaspecific PET imaging during radiochemotherapy in patients with locally advanced head-and-neck cancer. Radiother Oncol 2012;105:21–8. [10] Mortensen LS, Johansen J, Kallehauge J, et al. FAZA PET/CT hypoxia imaging in patients with squamous cell carcinoma of the head and neck treated with radiotherapy: results from the DAHANCA 24 trial. Radiother Oncol 2012;105:14–20. [11] Bittner MI, Wiedenmann N, Bucher S, et al. Exploratory geographical analysis of hypoxic subvolumes using 18F-MISO-PET imaging in patients with head and neck cancer in the course of primary chemoradiotherapy. Radiother Oncol 2013;108:511–6. [12] Rasey JS, Koh WJ, Evans ML, et al. Quantifying regional hypoxia in human tumors with positron emission tomography of [18F]fluoromisonidazole: a pretherapy study of 37 patients. Int J Radiat Oncol Biol Phys Sep 1996;36:417–28. [13] Yaromina A, Kroeber T, Meinzer A, et al. Exploratory study of the prognostic value of microenvironmental parameters during fractionated irradiation in human squamous cell carcinoma xenografts. Int J Radiat Oncol Biol Phys 2011 Jul 15;80:1205–13. [14] Ljungkvist AS, Bussink J, Kaanders JH, et al. Dynamics of hypoxia, proliferation and apoptosis after irradiation in a murine tumor model. Radiat Res 2006 Mar;165:326–36. [15] Stadler P, Becker A, Feldmann HJ, et al. Influence of the hypoxic subvolume on the survival of patients with head and neck cancer. Int J Radiat Oncol Biol Phys 1999;44:749–54. [16] Servagi-Vernat S, Differding S, Hanin FX, et al. A prospective clinical study of 18 F-FAZA PET-CT hypoxia imaging in head and neck squamous cell carcinoma before and during radiation therapy. Eur J Nucl Med Mol Imaging 2014;41:1544–52. [17] Bollineni VR, Koole MJ, Pruim J, et al. Dynamics of tumor hypoxia assessed by 18F-FAZA PET/CT in head and neck and lung cancer patients during chemoradiation: possible implications for radiotherapy treatment planning strategies. Radiother Oncol 2014 Nov;113:198–203.
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
MISO PET in head and neck
Serial [18F]-fluoromisonidazole PET during radiochemotherapy for locally advanced head and neck cancer and its correlation with outcome Nicole E. Wiedenmann a, Sabine Bucher a,c, Michael Hentschel d, Michael Mix b, Werner Vach e, Martin-Immanuel Bittner a, Ursula Nestle a, Jens Pfeiffer f, Wolfgang A. Weber b,g,1, Anca L. Grosu a,⇑,1 a Department of Radiation Oncology, German Cancer Consortium/Partner site Freiburg; b Department of Nuclear Medicine, University Medical Center Freiburg, Germany; c Clinic of Radiology & Nuclear Medicine, University Hospital Basel; d Department of Nuclear Medicine, Inselspital Bern, Switzerland; e Center for Medical Biometry and Medical Informatics; f Department of Otorhinolaryngology – Head and Neck Surgery, University Medical Center Freiburg, Germany; and g Molecular Imaging and Therapy Service, Memorial Sloan-Kettering Cancer Center, New York, USA
a r t i c l e
i n f o
Article history: Received 23 February 2015 Received in revised form 31 August 2015 Accepted 6 September 2015 Available online 29 September 2015 Keywords: Hypoxia Head and neck cancer 18F-MISO PET Radiochemotherapy
a b s t r a c t Purpose: The aim was to assess changes of tumour hypoxia during primary radiochemotherapy (RCT) for head and neck cancer (HNC) and to evaluate their relationship with treatment outcome. Material and methods: Hypoxia was assessed by FMISO-PET in weeks 0, 2 and 5 of RCT. The tumour volume (TV) was determined using FDG-PET/MRI/CT co-registered images. The level of hypoxia was quantified on FMISO-PET as TBRmax (SUVmaxTV/SUVmean background). The hypoxic subvolume (HSV) was defined as TV that showed FMISO uptake P1.4 times blood pool activity. Results: Sixteen consecutive patients (T3–4, N+, M0) were included (mean follow-up 31, median 44 months). Mean TBRmax decreased significantly (p < 0.05) from 1.94 to 1.57 (week 2) and 1.27 (week 5). Mean HSV in week 2 and week 5 (HSV2 = 5.8 ml, HSV3 = 0.3 ml) were significantly (p < 0.05) smaller than at baseline (HSV1 = 15.8 ml). Kaplan–Meier plots of local recurrence free survival stratified at the median TBRmax showed superior local control for less hypoxic tumours, the difference being significant at baseline and after 2 weeks (p = 0.031, p = 0.016). Conclusions: FMISO-PET documented that in most HNC reoxygenation starts early during RCT and is correlated with better outcome. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 117 (2015) 113–117
Hypoxia is a characteristic feature in HNC and has been identified as a factor negatively correlated with prognosis [1–3]. Tumour hypoxia is known to increase radiation resistance, has been identified as a key factor in triggering hypoxia-mediated gene expression and was found to be associated with a more aggressive tumour phenotype [1]. There are a significant number of trials showing that hypoxia-PET may serve as a tool to identify patients potentially at increased risk of local failure [3]. However, the question whether hypoxia-PET may serve as a basis for radiation treatment planning by dose painting in intensity modulated radiation therapy (IMRT) is still open [2–6]. A central problem is the poor knowledge regarding the variation of tumour hypoxia during R (C)T and the possible implication of this variation on radiation treatment planning. There are only limited data on the temporal changes of hypoxia based on PET imaging during R(C)T and its correlation with clinical ⇑ Corresponding author. 1
Equal contribution.
http://dx.doi.org/10.1016/j.radonc.2015.09.015 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
outcome. Eschmann et al. [7] investigated 14 patients with HNC on FMISO-PET before and under radiotherapy and observed a significant tumour reoxygenation under treatment (ca. 45 Gy), suggesting a correlation with a better tumour control but without statistical significance. The group of Leuven [8] assessed the variation of hypoxia on FMISO-PET in 15 HNC patients. They also described a significant re-oxygenation during RCT. Additionally they showed that hypoxia is one of the most important factors affecting RCT treatment resistance: not only the pretreatment level of hypoxia but also the extent during radiotherapy (30 Gy) correlated significantly with disease control. Zips et al. [9] investigated the temporal changes of tumour hypoxia in 25 HNC patients treated with RCT. In an exploratory prospective trial the authors showed that FMISO-PET in the first or second week of RCT has a strong prognostic value for diagnosing patients with high risk of tumour recurrence. Interestingly, PET before treatment was not as strong in comparison. The authors postulate that different treatment related tumour reoxygenation profiles lead to differences in treatment response. A prospective hypoxia imaging study
114
Dynamics of hypoxia in head and neck tumors assessed by [18]F-Miso PET
including 50 patients is on-going. The group of Aarhus [10] investigated on FAZA-PET/CT 40 patients with HNC cancer included in the DAHANCA 24 trial. In 13/40 patients a FAZA-PET/CT was performed about 14 days after the start of RCT. The HSV during treatment became significantly smaller and was located in the same region assessed at the first and second PET investigation. The authors observed a higher recurrence rate in the hypoxic group (4/6 patients) in comparison to the group showing no hypoxia under treatment (2/7 patients), however statistical significance was not reached. Our group assessed the distribution of hypoxia in HNC on FAZAPET [4], showing that in the majority of patients the hypoxic tissue was located in a single confluent region and therefore a HSV for radiation treatment planning could be delineated based on hypoxia-PET. Only in a small number of cases hypoxia was diffusely dispersed or the tumours were not hypoxic. In a more recent study [11] we have evaluated the changes in the location of the HSV on FMISO-PET during RCT and have concluded that in patients with persistent hypoxia after 2 weeks of treatment, the HSV is topographically relatively stable. The goal of this prospective feasibility study was to evaluate the temporal changes of tumour hypoxia during RCT, to correlate these data with the clinical outcome and, considering our results and published data, to discuss their possible impact on the design of new treatment strategies.
Image analysis PET data were iteratively reconstructed to voxels of 4.3 4.3 3.4 mm3 using an ordered-subset expectation–maximi zation (OSEM) reconstruction algorithm. Scatter and attenuation correction was applied. Hypoxic volumes were determined using PMOD software (PMOD-Group, Switzerland). Tumour volume (TV) delineation, separately for primary tumour and pathological lymph nodes, was performed on FDG-PET/CT/MRI co-registered images. In accordance with the literature, the TV was determined on FDG-PET by a 40% threshold of the maximum standardized uptake values (SUVs) within TV. The intensity of FMISO uptake was assessed quantitatively by calculating the ratio of the maximum SUV in the tumour (TV) to the mean SUV in contralateral neck musculature (TBRmax). By visual assessment, a TBRmax of P1.4 was considered most appropriate to define a tumour as being hypoxic. For statistical analysis, median TBRmax values were applied to split according to oxygenation status (median TBRmax for FMISO 1, 2, 3: 2.02, 1.26, 1.24). The HSV was determined by normalizing against blood activity concentration by counting all voxels within the TV for which the tumour to blood ratio on FMISO-PETs was P1.4 [12]. The blood activity concentration was determined from a region of interest in the left ventricle. The HF was defined as the ratio of HSV/TV.
Statistical analysis Materials and methods Patients’ characteristics Sixteen consecutive patients undergoing definitive RCT for locally advanced HNC (squamous cell carcinoma) were enrolled. Inclusion criteria were histologically confirmed T3/4, N+, M0 squamous cell carcinomas of the oral cavity, oropharynx, hypopharynx, and larynx and age P18.
Statistical analyses were performed using SPSS (v. 15.0.1.) and Stata/IC 12. p-values <0.05 were considered statistically significant. Comparison of continuous outcomes between subgroups is based on paired Wilcoxon-tests. The Wilcoxon matched-pairs signedranks test was used to compare TBRmax and HSVs between PET scans. The Mann–Whitney-U test was applied to compare mean TBRmax between recurrent and non-recurrent tumours. The association of various PET parameters with the time to recurrence was assessed using Kaplan–Meier plots based on splitting the values at the median, together with p-values from a log-rank test.
Study design A prospective feasibility FMISO-PET imaging monocentric study was conducted. The institutional review board approved the study and all participants provided written informed consent. Tumour hypoxia was assessed by FMISO-PET prior to (FMISO1) and during the course of RCT, at weeks 2–3 (FMISO2) and weeks 5–6 (FMISO3). All patients received planning CT, MRI (MRI1) and [18F]fluorodeoxyglucose (FDG)-PET imaging prior to RCT. MRI was repeated in treatment weeks 5–6 (MRI2). Radiation treatment was delivered as conformal IMRT (2 Gy/d, 5/week; total dose to macroscopic tumour 70 Gy). Cisplatin was administered in weeks 1, 4, 7 (100 mg/m2). For followup, patients were evaluated clinically and by MRI every 3 months.
PET/CT [18F]-MISO and [18F]-FDG production met standard quality criteria. For FMISO-PET, one bed position, covering the head and neck region examined 150 min after injection of approx. 400 MBq [18F]MISO. Scan duration was 35 min (3 frames at 10 min, followed by a 5 min transmission scan). For quantitative analysis, three attenuation corrected frames were summed after excluding frames with patient movement. Planning CT images were acquired in 2-mm slice thickness. For FDG-, FMISO-PET and CT, patients were immobilized identical to the radiation position with a head and neck mask.
Results 15/16 patients (94%) showed hypoxic tumour tissue on the baseline FMISO-PET scan: 11 in the primary tumour only, 2 in lymph nodes only, 2 in both. The number of patients with hypoxic lesions decreased to 5/14 (36%) at the second and to 3/11 (27%) at the third FMISO-PET investigation. Mean TBRmax significantly decreased from 1.94 (pre-treatment) to 1.57 (week 2, p = 0.035) and 1.27 (week 5, p = 0.003) (Table 1, Fig. 1). The median TBRmax dropped from 2.02 (pre-treatment) to 1.26 (week 2) and 1.24 (week 5). Mean and median TV assessed by FDG-PET/CT/MRI and HSV, HF on FMISO-PET at baseline and in week 2 and 5 are presented in Table 2. Mean follow-up was 31 months (median 44, range 1–53 months). The time course of FMISO TBRmax was strikingly different for patients with and without recurrence (Figs. 1 and 2, Table 1). While patients without tumour recurrence demonstrated a rapid and continuous decrease in the TBRmax, the decline was slower in patients with recurrence and some patients even showed an increase in tumour hypoxia at the time of the second FMISO-PET. No local recurrence occurred if there was at least one negative FMISO scan (TBRmax < 1.4, Fig. 1). For the three consecutive FMISO-PET scans, patients with tumour recurrence were found to have higher mean TBRmax compared to patients without tumour recurrence: mean TBRmax over all time points: 2.15 vs. 1.45, (Mann–Whitney-U-Test, p = 0.002).
115
N.E. Wiedenmann et al. / Radiotherapy and Oncology 117 (2015) 113–117
Table 1 TBRmax for FMISO1–3, tumour response on MRI, length of follow-up (months), local recurrence status and cause of death, if applicable. Yellow = TBRmax P 1.4, PR = partial remission, SD = stable disease, CR = complete remission, n.a. = not available.
Kaplan–Meier plots of local recurrence free survival showed a significantly lower local tumour control probability for more hypoxic vs. less hypoxic tumours for scans at baseline (FMISO1, median TBRmax = 2.02), p = 0.031 and scans after week 2 (FMISO2, median TBRmax = 1.26), p = 0.016 (Fig. 2). Representative examples are shown in Fig. 3. No correlation was found between the FDG TBRmax and the FMISO TBRmax at baseline (Pearson correlation coefficient = 0.108, p = 0.714). Tumour FDG was not correlated with survival: neither the TBRmax of the primary tumour, nor of lymph nodes, nor the TV were significantly correlated with survival (p = 0.568, p = 0.935 and p = 0.697, respectively). Discussion
Fig. 1. Individual plots of FMISO TBRmax plotted against time for recurrent tumours (red) vs. non-recurrent tumours (black).
In this prospective feasibility study we assess tumour hypoxia on serial FMISO PET scans during RCT and analyse the correlation between different patterns of tumour oxygenation and the risk of local tumour recurrence. Two main findings are made in this study: the first observation is that tumour hypoxia and its temporal changes upon RCT are different among patients. The presence of baseline hypoxia and subsequent resolution of hypoxia upon treatment was the predominant pattern of tumour oxygenation that was observed. However, there were also patients with hypoxia increasing up to week two or with complete absence of tumour hypoxia. Here we define ‘reoxygenation’ as the improvement of tumour oxygenation status upon radiation treatment. Specifically, reoxygenation refers to reduction of the extent of hypoxic volume and reduction of the level of hypoxia over time. The second observation of this study is that the changes of hypoxia assessed by PET during RCT are significantly correlated with outcome. Our observation of a high incidence of baseline tumour hypoxia and subsequent temporal changes with a tendency for early onset
Table 2 TV measured on FDG-PET/CT/MRI before RCT and the mean and median HSV, measured on FMISO PET in weeks 0, 2 and 5 (HSV1, HSV2, HSV3). HF (HSV/TV) for HSV1–3 is designated by HF1, HF2, and HF3, respectively.
Mean (± STD) Median (min–max)
TV (ml)
HSV1 (ml)
HSV2 (ml)
HSV3 (ml)
HF1
HF2
HF3
27.4 (± 22.8) 20.3 (9.4–98.6)
15.8 (±18.4) 11.3 (1–73.8)
5.8 (±13.1) 0.1 (0–44.7)
0.3 (±0.8) 0 (0–2.3)
54.3%
11.8%
1.2%
116
Dynamics of hypoxia in head and neck tumors assessed by [18]F-Miso PET
red: blue:
TBRmax > median TBRmax < median
number at risk months FMISO1 TBRmax < 2.02 TBRmax ≥ 2.02 FMISO2 TBRmax < 1.26 TBRmax ≥ 1.26 FMISO3 TBRmax < 1.24 TBRmax ≥ 1.24
0
12
24
36
48
60
8 8
6 4
5 4
4 4
4 2
0 0
7 7
6 3
6 2
5 2
4 0
0 0
5 6
4 4
3 4
2 4
2 2
0 0
Fig. 2. Kaplan–Meier plots: Local control probability plotted against time after treatment. Patients are stratified at the median TBRmax for FMISO 1, 2 and 3. Red: TBRmax > median, blue: TBRmax < median. FMISO1: p = 0.031, FMISO2: p = 0.016, FMISO3: p = 0.071.
Fig. 3. Representative examples: regression of hypoxia (above, patient#15) vs. increase (week 2) followed by incomplete reduction (week 6) of hypoxia with residual hypoxia remaining (below, patient#16).
N.E. Wiedenmann et al. / Radiotherapy and Oncology 117 (2015) 113–117
of reoxygenation and heterogeneity among individual patients are in agreement with published preclinical data. A substantial proportion of baseline tumour hypoxia and significant reoxygenation after R(C)T have also been described in the preclinical setting using mouse tumour models [13,14] as well as in patients by pO2 histographic measurements [15] before hypoxia-PET became available. Accordingly, tumour hypoxia and tumour reoxygenation are well-known phenomenons in radiation biology. Suggested mechanisms for tumour reoxygenation within fractionated radiation treatment include improved blood perfusion, a decrease of tumour oxygen consumption rate mostly due to a reduced cell number and impaired proliferation, and tumour shrinkage with subsequent improved oxygen diffusion to formerly hypoxic cells. The incidence of baseline hypoxia in our study is comparable with published data, however on the upper limit [6–9,12]. The pattern of tumour oxygenation showed individual patient differences with early reoxygenation as the predominant phenomenon. These findings are in line with the trial of Zips et al. [9] as well as other imaging studies [7,8,10,16,17]. In our study, the median TBRmax for FMISO2 nearly equals the median TBRmax for FMISO3 (1.26 vs. 1.24), reflecting complete reoxygenation taking place early. A possible mechanism at this relatively early time point is a therapy (irradiation + 1 cycle cisplatin) induced reduction of oxygen consumption rather than tumour shrinkage. The chemotherapy regimen included three cycles of cisplatin total (100 mg/kg/d). Thereby one third of the total amount of cisplatin had already been administered before imaging FMISO2, which is a higher total dose than administered by more continuous protocols as used by Zips et al. and could be a factor contributing to a lower TBRmax in week 2 compared to Zips et al. The second observation of this study is the correlation between tumour hypoxia and outcome. Kaplan–Meier plots for tumours stratified at the median TBRmax (2.02, 1.26, 1.24) show superior local recurrence free survival for better oxygenated tumours for all time points (Fig. 2). Statistical significance is reached for the baseline examination (FMISO1, p = 0.031) and the examination obtained after week 2 (FMISO2, p = 0.016). A strong correlation between image parameters and local control was also found by Zips et al. for the week 2 time point, while for the baseline examination significance is only reported on univariate, not at multivariate analysis [9]. The strong correlation between imaging parameters of hypoxia and patient outcome indicates that FMISO-PET may be able to differentiate between patients with high and low risk of local recurrence/failure, i.e. to select early responders from non-responders. This might enable treatment intensification for selected patients or even hypoxia-targeted IMRT regimens. Whether the initial TV, HSV1, 2 or 3 should be boosted remains to be elucidated. Conflict of interest None. Acknowledgements Acknowledgement of grant or other financial support, meeting presentation, or assistance with manuscript preparation or data collection is done:
117
This work was supported by the German Cancer Consortium (DKTK), Partner site Freiburg. The authors thank Prof. Peter Vaupel for critical input on the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2015.09. 015. References [1] Vaupel P, Höckel M, Mayer A. Detection and characterization of tumor hypoxia using pO2 histography. Antioxid Redox Signal 2007 Aug;9:1221–35. [2] Overgaard J. Hypoxic modification of radiotherapy in squamous cell carcinoma of the head and neck – a systemic review and meta-analysis. Radiother Oncol 2011;100:22–32. [3] Horsman MR, Mortensen LS, Petersen JB, et al. Imaging hypoxia to improve radiotherapy outcome. Nat Rev Clin Oncol 2012;9:674–87. [4] Grosu AL, Souvatzoglou M, Röper B, et al. Hypoxia imaging with FAZA-PET and theoretical considerations with regard to dose painting for individualization of radiotherapy in patients with head and neck cancer. Int J Radiat Oncol Biol Phys 2007;69:541–51. [5] Thorwarth D, Alber M. Implementation of hypoxia imaging into treatment planning and delivery. Radiother Oncol 2010;97:172–5. [6] Nehmeh SA, Lee NY, Schröder H, et al. Reproducibility of intratumor distribution of (18)F-fluoromisonidazole in head and neck cancer. Int J Radiat Oncol Biol Phys 2008;70:235–42. [7] Eschmann SM, Paulsen F, Bedeshem C, et al. Hypoia-imaging with 18FMisonidazole and PET: changes of kinetics during radiotherapy of head-andneck cancer. Radiother Oncol 2007;83:406–10. [8] Dirix P, Vandecaveye V, De Keyzer F. Dose painting in radiotherapy for head and neck squamous cell carcinoma: value of repeated functional imaging with 18F-FDG PET, 18F-fluoromisonidazole PET, diffusion-weighted MRI and dynamic contrast-enhanced MRI, and dynamic contrast-enhanced MRI. J Nucl Med 2009;50:1020–7. [9] Zips D, Zöphel K, Abolmaali N, et al. Exploratory prospective trial of hypoxiaspecific PET imaging during radiochemotherapy in patients with locally advanced head-and-neck cancer. Radiother Oncol 2012;105:21–8. [10] Mortensen LS, Johansen J, Kallehauge J, et al. FAZA PET/CT hypoxia imaging in patients with squamous cell carcinoma of the head and neck treated with radiotherapy: results from the DAHANCA 24 trial. Radiother Oncol 2012;105:14–20. [11] Bittner MI, Wiedenmann N, Bucher S, et al. Exploratory geographical analysis of hypoxic subvolumes using 18F-MISO-PET imaging in patients with head and neck cancer in the course of primary chemoradiotherapy. Radiother Oncol 2013;108:511–6. [12] Rasey JS, Koh WJ, Evans ML, et al. Quantifying regional hypoxia in human tumors with positron emission tomography of [18F]fluoromisonidazole: a pretherapy study of 37 patients. Int J Radiat Oncol Biol Phys Sep 1996;36:417–28. [13] Yaromina A, Kroeber T, Meinzer A, et al. Exploratory study of the prognostic value of microenvironmental parameters during fractionated irradiation in human squamous cell carcinoma xenografts. Int J Radiat Oncol Biol Phys 2011 Jul 15;80:1205–13. [14] Ljungkvist AS, Bussink J, Kaanders JH, et al. Dynamics of hypoxia, proliferation and apoptosis after irradiation in a murine tumor model. Radiat Res 2006 Mar;165:326–36. [15] Stadler P, Becker A, Feldmann HJ, et al. Influence of the hypoxic subvolume on the survival of patients with head and neck cancer. Int J Radiat Oncol Biol Phys 1999;44:749–54. [16] Servagi-Vernat S, Differding S, Hanin FX, et al. A prospective clinical study of 18 F-FAZA PET-CT hypoxia imaging in head and neck squamous cell carcinoma before and during radiation therapy. Eur J Nucl Med Mol Imaging 2014;41:1544–52. [17] Bollineni VR, Koole MJ, Pruim J, et al. Dynamics of tumor hypoxia assessed by 18F-FAZA PET/CT in head and neck and lung cancer patients during chemoradiation: possible implications for radiotherapy treatment planning strategies. Radiother Oncol 2014 Nov;113:198–203.