Skip to main content

Investigation of atovaquone-induced spatial changes in tumour hypoxia assessed by hypoxia PET/CT in non-small cell lung cancer patients

Abstract

Background

Tumour hypoxia promotes an aggressive tumour phenotype and enhances resistance to anticancer treatments. Following the recent observation that the mitochondrial inhibitor atovaquone increases tumour oxygenation in NSCLC, we sought to assess whether atovaquone affects tumour subregions differently depending on their level of hypoxia.

Methods

Patients with resectable NSCLC participated in the ATOM trial (NCT02628080). Cohort 1 (n = 15) received atovaquone treatment, whilst cohort 2 (n = 15) did not. Hypoxia-related metrics, including change in mean tumour-to-blood ratio, tumour hypoxic volume, and fraction of hypoxic voxels, were assessed using hypoxia PET imaging. Tumours were divided into four subregions or distance categories: edge, outer, inner, and centre, using MATLAB.

Results

Atovaquone-induced reduction in tumour hypoxia mostly occurred in the inner and outer tumour subregions, and to a lesser extent in the centre subregion. Atovaquone did not seem to act in the edge subregion, which was the only tumour subregion that was non-hypoxic at baseline. Notably, the most intensely hypoxic tumour voxels, and therefore the most radiobiologically resistant areas, were subject to the most pronounced decrease in hypoxia in the different subregions.

Conclusions

This study provides insights into the action of atovaquone in tumour subregions that help to better understand its role as a novel tumour radiosensitiser.

Trial registration: ClinicalTrials.gov, NCT0262808. Registered 11th December 2015, https://clinicaltrials.gov/ct2/show/NCT02628080

Background

Lung cancer is the leading cause of cancer-related death in the UK with a five-year overall survival rate only in the region of 16% [1]. Non-small cell lung cancer constitutes 80% of lung cancer cases [2] and is associated with poor survival despite advances in the delivery of multi-modality treatment. Modern treatment of NSCLC is complex and increasingly involves utilisation of several treatment modalities, such as surgery, radiotherapy (RT), systemic therapies (chemotherapy, immunotherapy, and targeted agents), and interventional radiology and palliative care [3]. Modelling using multivariate analysis with Monte Carlo simulation reported that around 77% of patients with lung cancer have an evidence-based indication for RT during their course of disease [4]. Many solid tumours, including NSCLC, are typically dependent upon an abnormal, poorly functioning vasculature for oxygen delivery [5]. Coupled with the high metabolic requirements of many tumours, this leads to an imbalance in oxygen supply and demand and so causes tumour hypoxia. Importantly, various pre-clinical and clinical studies have highlighted that tumour hypoxia enhances resistance to anticancer treatments, particularly RT, and promotes an aggressive tumour phenotype [6].

Oxygen deprivation or hypoxia is defined as an oxygen tension under physiological normoxia (less than or equal to 2.03–3.04 kPa) [7]. Hypoxia is closely associated with several ‘hallmarks of cancer’ such as reprogramming energy metabolism, inducing angiogenesis, and resisting cell death [8]. Gray and colleagues suggested for the first time in the 1950s that hypoxia could influence radiotherapy outcomes as the radioresistance of hypoxic tumour cells increases by a factor of up to three compared to normoxic tumour cells [5, 9, 10].

To reduce tumour hypoxia and therefore enhance radiosensitivity, decreasing cellular oxygen consumption rate (OCR) via metabolic reprogramming has arisen as a promising approach [11]. Mathematical modelling indicated that a 30% reduction in OCR would abrogate severe tumour hypoxia and could stand as a more efficient strategy than enhancing blood oxygen levels or augmenting blood flow [12].

Notably, the well-tolerated anti-malarial drug atovaquone has been shown to decrease OCR in multiple cancer cell lines, alleviate hypoxia in spheroids and xenografted tumours, and induce radiosensitisation [13]. Atovaquone acts as a mitochondrial inhibitor of oxidative phosphorylation at complex III (the cytochrome bc1 complex) of the electron transport chain (ETC) [14]. Moreover, in a translational clinical trial, we recently assessed, through the analysis of gene expression and pharmacodynamics endpoints of tumour hypoxia, the clinical potential of atovaquone as a mitochondrial inhibitor in patients with NSCLC [15]. This study reported for the first time that a mitochondrial inhibitor can be employed to target mitochondrial metabolism and modify the tumour microenvironment through hypoxia reduction in patients so as to potentially improve the efficacy of radiotherapy [15].

Following the observation that atovaquone results in hypoxia reduction in NSCLC patients [15], we hypothesised that atovaquone may affect tumour subregions differently depending on their level of hypoxia. Thus, the aim of the present study was to investigate atovaquone-induced changes in hypoxia in tumour subregions by analysing hypoxia PET-CT scans.

Methods

Patients

Patients with resectable NSCLC were recruited for the open-label, non-randomised, equal-sized two-cohort ATOM clinical trial (NCT02628080) completed at the Oxford Cancer and Haematology Centre (UK) in accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines. Patients in cohort 1 (n = 15) received oral atovaquone (Wellvone, 750 mg/5 mL micronised suspension, GlaxoSmithKline) twice daily. Patients were asked to take atovaquone orally together with fat-containing food to aid absorption. Patients in cohort 2 (n = 15) did not receive atovaquone. Eligible patients were aged ≥ 18 years, had a pathologic or radiological diagnosis of NSCLC, were scheduled for surgical resection, had disease > 2 cm in diameter, and had Eastern Cooperative Oncology Group (ECOG) performance status 0–2. Patients were excluded if taking known ETC inhibitors. Despite a male predominance in untreated patients, the main clinical characteristics were well balanced in the two cohorts. The clinical characteristics of patients are shown in Table 1. For full details regarding trial design, patients, and treatment, the reader is referred to [15].

Table 1 Clinical characteristics of patients

Data acquisition

Change in tumour hypoxic volume (HV) was the primary imaging endpoint measured with hypoxia PET-CT. Single bed position image acquisition centred on the tumour was performed with GE Discovery 690 or 710 PET-CT Scanners (GE Healthcare) for 10 min at 4 hours following the administration of 18F-fluoromisonidazole (FMISO, 29 patients) (University of Cambridge) or 18F-fluoroazomycin-arabinofuranoside (FAZA, 1 patient) (University of Manchester) with an activity of 370 MBq. The same scanner and tracer were used for the two visits of each patient with baseline scans (indicated as ‘pre’) and pre-surgery scans (indicated as ‘post’) for atovaquone-treated and untreated patients. CT images provided attenuation correction and localisation. Patients in cohort 1 had a median length of 13 (IQR, 9–14) days between imaging time points, depending on their planned date for surgery. Patients in cohort 2 had a median length of 14 (IQR 7–14) days between imaging time points.

Image analysis

Tumours on four-hour hypoxia PET-CT images were manually outlined on the CT image by an experienced radiologist and copied to the co-registered PET image. Images were analysed using Hermes Hybrid Viewer Software (Hermes Medical Solutions AB) and MATLAB (version R2021a, MathWorks, Natick, MA, USA). A matrix containing each voxel coordinates (x, y, z) along with the respective radiotracer standardised uptake value (SUV) per voxel was first extracted for every outlined tumour volume.

The background mean SUV (SUVmean) was obtained by outlining blood in the central portion of the descending aorta. To measure the HV, each tumour voxel’s SUV was divided by the background SUVmean to determine the tumour-to-blood ratio (TBR) value per voxel:

$${\text{TBR = }}\frac{{{\text{SUVvoxel}}}}{{\text{SUVmean descending aorta}}}$$
(1)

As described by Koh et al. [16], voxels with a TBR equal to or greater than 1.4 were classified as hypoxic. Volumes of hypoxic voxels and TBRmean values were compared pre- versus post-atovaquone. A decrease in HV equal to or greater than 10% from baseline was described as a meaningful decrease after atovaquone treatment, in accordance with a previous clinical study [17]. The minimum detectable change (MDC) method and FMISO test–retest reproducibility data were used to determine this cut-off [18]. For additional explanation regarding this threshold, the reader is referred to [15], Supplementary Materials and Methods.

Changes in hypoxia following atovaquone treatment were studied in tumours overall and in tumour subregions. In-house MATLAB (version R2021a, MathWorks, Natick, MA, USA) code was first used to calculate the distance of every tumour voxel to the nearest edge of the outlined tumour. Voxel dimensions on PET-CT images were 2.7 × 2.7 × 3.3 mm3. Voxels were then divided into four subregions or distance categories: edge (the outermost shell of voxels), outer (voxels’ centre up to 5.5 mm of the tumour outline), inner (voxels’ centre between 5.5 and 11 mm of the tumour outline), and central (voxels’ centre superior to 11 mm inside the tumour outline), in line with a previous PET study about FMISO uptake in advanced NSCLC [19].

Statistical analysis

Statistical analyses used IBM SPSS Statistics (version 27). The normality of data was inspected using Shapiro–Wilk test. For each distance category, the chi-square test of homogeneity was used to assess the significance of pre- to post-atovaquone changes in the proportions of voxels assigned to each region. Percentage change in TBRmean and HV was calculated between trial visits for each patient. The binomial method was used to derive 95% confidence intervals (CI) for median changes within cohorts. The statistical significance of the difference of means and medians of TBRmean and HV in each distance category pre- to post-atovaquone was evaluated using the Welch’s t test for samples of unequal variance. For multiple comparisons in t tests, a Bonferroni-corrected P value < 0.05 was considered significant.

Results

Changes of TBR in tumour subregions

Atovaquone-induced changes in hypoxia were initially assessed in tumours overall and in tumour subregions, using four-hour hypoxia PET-CT. TBRmean was the first hypoxia-related score evaluated. TBRmean decreased significantly from baseline in atovaquone-treated patients in tumours overall (1.11 to 1.01, P < 0.01) as well as in the inner (1.32 to 1.20, P = 0.01) and outer (1.37 to 1.26, P < 0.01) subregions (Fig. 1A). However, there was no significant reduction in TBRmean from baseline in the centre subregion (1.19 to 1.12, P = 0.07). Notably, baseline mean TBRmean in the inner [1.32 (95% CI 1.2–1.5)] and outer [1.37 (95% CI 1.2–1.5)] subregions exceeded baseline mean TBRmean in the centre subregion [1.19 (95% CI 1.0–1.3)]. Change in mean TBRmean in the non-hypoxic edge subregion following treatment was negligible [0.50 (95% CI 0.4–0.6) to 0.54 (95% CI 0.5–0.6), P = 0.23]. In contrast, no significant change was observed in TBRmean in untreated patients, neither in tumours overall nor in any tumour subregion (Fig. 1B).

Fig. 1
figure 1

TBRmean scores in tumours overall and tumour subregions at baseline scans (indicated as ‘pre’) and pre-surgery scans (indicated as ‘post’) for atovaquone-treated (A) and untreated (B) patients. TBRmax scores in tumours overall and tumour subregions at ‘pre’ and ‘post’ for atovaquone-treated (C) and untreated (D) patients. Results indicated as mean ± standard error. *, P < 0.05; **, P < 0.01; ns, non-significant

The overall change in TBRmean from baseline was − 9.0% in atovaquone-treated patients against + 3.7% in untreated patients. In atovaquone-treated patients, change in TBRmean from baseline was − 5.8% in centre, − 9.1% in inner, and − 8.0% in outer subregions. In untreated patients, negligible change in TBRmean was observed in these three subregions (0.9%, − 1.6%, and 0.0%, respectively). TBRmean increased from baseline in the edge region in both cohorts similarly, by + 8.0% in atovaquone-treated and by + 4.6% in untreated patients. Thus, atovaquone caused a decrease in TBRmean in tumours overall and in all tumour subregions except the edge region. Interestingly, the reduction in TBRmean in the centre subregion was lower than in the inner and outer subregions. TBRmax values showed a similar trend with no significant changes from baseline in each distance category and overall for untreated patients (Fig. 1D) but was significantly reduced in two distance categories as well as overall for treated patients (Fig. 1C).

Changes of tumour hypoxic volume

Atovaquone-induced changes in tumour HV were then assessed. TBR equal to or greater than 1.4 was first used to define HV. Eleven (73.3%) atovaquone-treated patients had an overall and meaningful (equal to or greater than 10%) decrease in HV from baseline, and the median change was − 28% (95% CI − 58.2 to − 4.4). Meanwhile, only two (13.3%) untreated patients had an overall reduction in HV equal to or greater than 10%, and the median change was + 15.5% (95% CI − 6.5 to 35.5), as previously reported [15]. Following atovaquone treatment, eight patients (53%) also had a meaningful decrease in HV in the centre subregion, with a median change of − 10.4% (95% CI − 54.2 to − 1.05) (Fig. 2A). Nine (60%) patients had a meaningful reduction in HV in the inner subregion, which had the greatest median change of − 33.3% (95% CI − 61.7 to 0.6), as well as in the outer subregion, with a median change of − 23.5% (95% CI − 49.4 to 0.5) (Fig. 2B, C). Lastly, five (33%) patients had a meaningful reduction in HV in the edge subregion, even if the median change was 0.0% (95% CI − 33.7 to 4.6) (Fig. 2D). In contrast, median change in untreated patients was + 11.5% (95% CI − 2.8 to 32.1) in the outer subregion, + 1.5% (95% CI − 7.4 to 7.3) in the inner subregion, and 0.0% in the centre (95% CI − 2.1 to 3.6) and edge (95% CI − 5.1 to 5.7) subregions. A meaningful reduction in HV was observed in only 2 (13.3%) untreated patients in the inner, outer, and edge subregions (Additional file 1: Figure S1).

Fig. 2
figure 2

Change in tumour HV measured by hypoxia PET-CT. Waterfall plots of percentage change in HV in tumour subregions for atovaquone-treated patients: centre (A), inner (B), outer (C), and edge (D). TBR ≥ 1.4 is used to define HV. A reduction in HV ≥ 10% was considered meaningful (dashed line). All atovaquone-treated patients including those with no change in HV between baseline scan and pre-surgery scan are included. Waterfall plots of percentage change in HV in tumour subregions for untreated patients are presented in Supplementary Fig. 1

In both cohorts, the magnitude of tumour HV at baseline was not in itself predictive of hypoxia reduction to atovaquone. For instance, the two untreated patients that displayed a meaningful reduction in HV in tumours overall had either medium (44.8 mL) or very small (3.5 mL) baseline HV. Similarly, no meaningful decrease in HV was observed in four atovaquone-treated patients, who had large (366.9 mL), medium (45.9 and 59.3 mL), and small (24.6 mL) baseline HVs.

As median HV decreased from baseline in the centre, inner, and outer tumour subregions, analysis of more intense tumour hypoxia was performed. TBR greater than 1.6 and 1.8 were used to define tumour hypoxic voxels. Only tumours that had sufficient HV (predefined as ≥ 1.5 mL) at these TBR thresholds to accurately assess change were included. Using TBR greater than 1.6 and 1.8, atovaquone-induced reduction in HV was even more pronounced in all tumour subregions except in the edge subregion where no median change in HV was consistently observed (Table 2).

Table 2 Summary of tumour hypoxic volume results using 4-h hypoxia PET-CT. Three thresholds were used to define progressively higher hypoxia PET tracer uptake: TBR ≥ 1.4 (a), TBR ≥ 1.6 (b), and TBR ≥ 1.8 (c)

Given that the treatment duration differed between patients, its influence on tumour HV changes was evaluated (Additional file 1: Figure S2). The three patients with the greatest reduction in HV from baseline (− 100%, − 86% and − 69%) received atovaquone for more than 11 days. However, the only two patients who had an increase in HV from baseline were also in this group. Overall, no relationship was observed between changes in HV from baseline and treatment time. There was a high correlation between the hypoxic volume (based on TBR > 1.4) and total tumour volume (examined for pre-only patients). The correlation decreased for higher TBR thresholds. This is shown in Additional file 1: Figure S3.

Changes in the fraction of tumour hypoxic voxels

Following the analysis of atovaquone-induced changes in HV in tumour subregions using different TBR thresholds to define hypoxia, variations in the fraction of tumour hypoxic voxels were investigated (Fig. 3). Using TBR ≥ 1.4 to define hypoxia, globally, the mean fraction of hypoxic voxels decreased significantly from baseline in atovaquone-treated patients (P < 0.001) contrary to untreated patients where no significant change was observed (P = 0.63) (Table 3).

Fig. 3
figure 3

Changes in the fraction of hypoxic voxels from baseline in atovaquone-treated (A) and untreated (B) patients. Tumour hypoxic voxels are divided into three groups (yellow, orange, red) according to their level of hypoxia defined by their TBR value. Statistical significances are only indicated for global changes in the mean fraction of hypoxic voxels between visits (i.e. baseline visit and pre-surgery visit named ‘pre’ and ‘post’, respectively) not all three TBR thresholds. **, P < 0.01; ns, non-significant

Table 3 Summary of the fraction of hypoxic voxels at baseline (‘pre’) and pre-surgery (‘post’) visits for atovaquone-treated (a) and untreated (b) patients overall and in tumour subregions, using four-hour hypoxia PET-CT analysis. Voxels with a TBR ≥ 1.4 were defined as hypoxic. Mean values and IQR are shown for pre and post with P values for the difference pre to post

In particular, with treatment, a significant reduction from baseline in the mean fraction of hypoxic voxels was identified in the inner (P = 0.01) and outer (P = 0.01) subregions. The reduction was non-significant in the centre (P = 0.05) and negligible in the edge subregion (P = 0.86) (Table 3a). In untreated patients, despite a trend towards an increase from baseline in the mean fraction of hypoxic voxels in the centre, outer, and edge subregions, or reduction in the inner subregion, no change was statistically significant (Table 3b).

Different TBR thresholds were then used to compare changes in the fraction of progressively more hypoxic voxels (Fig. 3A, B). Regarding atovaquone-treated patients, in the 1.4 ≤ TBR < 1.6 category, there was no significant change from baseline in the mean fraction of voxels neither overall (P = 0.96) nor in centre (P = 0.31), inner (P = 0.95), outer (P = 0.91), and edge (P = 0.21) subregions (Table 4a). In the 1.6 ≤ TBR < 1.8 category, a significant reduction from baseline was only observed in tumours overall (P = 0.01) (Table 4b). In the TBR ≥ 1.8 category, a significant reduction from baseline was observed in tumours overall (P < 0.001) and in the inner (P < 0.01) and outer (P = 0.01) subregions (Table 4c). In untreated patients, there was no significant change from baseline in the mean fraction of voxels in any TBR category, neither overall in tumours nor in subregions.

Table 4 Summary of the fraction of hypoxic voxels at baseline (‘pre’) and pre-surgery (‘post’) visits for atovaquone-treated patients, with four-hour hypoxia PET-CT. Three TBR categories were used to define progressively higher hypoxia PET tracer uptake: 1.4 ≤ TBR < 1.6 (a), 1.6 ≤ TBR < 1.8 (b), and TBR ≥ 1.8 (c). Mean values and IQR are shown for pre and post with P values for the difference pre to post

Thus, it seems that atovaquone did decrease the global fraction of hypoxic voxels significantly in tumours overall and in all subregions except the edge subregion. The fraction of the most intensely hypoxic voxels also showed the greatest reduction among tumour hypoxic voxels and in most tumour subregions.

Discussion

Following the recent demonstration that the mitochondrial inhibitor atovaquone decreases tumour hypoxia in patients with NSCLC, this study shows that the reduction in TBR and HV occurred in the outer, inner, and centre tumour subregions. The centre subregion displayed a lower TBR at baseline which is likely to be due to lack of hypoxic PET tracer uptake due to a lack of viable tumour cells (e.g. necrotic regions) and may explain the lesser reduction observed in these regions. Besides, atovaquone did not seem to act at the edge subregion of the tumour. Baseline TBRmean in this subregion was markedly lower than the hypoxia threshold set at 1.4, and although TBRmean did increase during treatment, it was insufficient to classify these voxels as hypoxic, thus explaining the absence of change in median HV in the edge tumour subregion.

The increase in median HV and TBRmean from baseline in tumours of untreated patients, signifying that tumour hypoxia intensified over time, reinforced the reduction observed in atovaquone-treated patients. Moreover, as the nitroimidazole-based tracers FMISO and FAZA used accumulate exclusively in viable cells at oxygen levels responsible for hypoxia-related radioresistance [20], any decrease in uptake should therefore result in an improvement in tumour response to radiation. Indeed, several clinical studies reported that high uptake of these tracers acts as a negative prognostic biomarker for patients undergoing RT [21, 22]. Notably, we observed that the most intensely hypoxic tumour voxels (with a TBR greater than 1.6 or 1.8), defining radiobiologically highly resistant regions, were subject to the most pronounced decrease in hypoxia in tumour subregions. This provides insights into the action of atovaquone as a novel tumour radiosensitiser.

As hypoxia appeared to decrease to a greater extent in the inner and outer tumour subregions than in the centre subregion in atovaquone-treated patients, we hypothesise that blood perfusion may impact the reported treatment outcome. TBR indices at 4 hours post-injection seemed to fall towards tumour centres, and it is possible that TBR could underestimate the extent of hypoxia, especially at tumour centres. These voxels could be poorly perfused (or potentially necrotic), leading to very low rates of FMISO influx which would cause TBR values to remain relatively low despite generating hypoxic conditions which would cause long-term tracer accumulation. Thus, it would be interesting to assess the relationship between tumour blood perfusion and TBR using baseline perfusion CT and hypoxia PET scans.

Fixed distance categories were used to assign voxels to a tumour subregion, in accordance with a previous study on FMISO PET kinetics in NSCLC [19]. Yet, tumour volumes varied, and for large tumours, a predominance of voxels was assigned to the centre subregion. We tested dividing tumours into five subregions instead of four, but no significant change of TBR in response to atovaquone between each subregion was observed. Alternatively, a method that determines distance categories based on the tumour volume could be employed. By ranking from lowest to highest the voxels’ minimum distance from the tumour outline, and then assigning 25% of the voxels with the lowest distance from the outline in the edge subregion, the next 25% of the voxels in the outer subregion, the next 25% of the voxels in the inner subregion, and the voxels left in the centre subregion, the voxels may be divided into the subregions more proportionally. This method was tried for the first three patients, but again, no significant change in TBR per subregion was observed in response to atovaquone.

Moreover, the resolution of a PET scanner is around 3–4 mm [23], whilst the molecular effect of oxygen ranges from nm to μm [24]. The width of PET image voxels is thus of this length or a little bit smaller, which is notably greater than that of an average hypoxic tumour region, causing PET image voxels to likely include both hypoxic and normoxic regions. This causes relatively low SUV values on PET scans using the radiotracer FMISO compared to FDG, whose uptake within individual voxels is more uniform [25]. Moreover, as the binding function of FMISO is nonlinear, and smooth instead of a simple threshold, several scenarios can engender the same SUVmean in a voxel. As shown by Grimes et al (figure 5, [24]), one would expect roughly the same PET signal whether the entire voxel was at 4.2 mmHg, or the voxel was split in half between oxic (50%) and 1.4 mmHg (50%), or if a voxel was 25% anoxic (and viable) and 75% well oxygenated. Concerns about the interpretation of PET results could thus be raised as the radiobiological response would presumably vary between these scenarios, given the typical oxygen enhancement ratio curve.

Our analysis of hypoxia PET scans shows differences in hypoxia in different NSCLC tumour regions at baseline, which could be useful for RT dose guidance. Hypoxia-based RT dose painting is an appealing approach for cancer therapy [26, 27]. Additional RT is dosimetrically and spatially adapted in dose painting, using tumour response maps obtained by repeat imaging during RT [28]. FMISO PET has previously been investigated for dose painting in head and neck cancer [29]. As retention of hypoxia PET radiotracer exclusively happens within viable cells retaining a functional ETC, tumour subvolumes that are very hypoxic but necrotic will plausibly not receive unnecessary dose escalation. Higher RT dose is, however, likely to only benefit hypoxic cancer cells that display clonogenic replicative potential [26].

Hypoxia PET image repeatability has been demonstrated to be high with modern PET/CT systems using FMISO head and neck cancer (HNC) [30] and NSCLC [31] over 1–2 days and using 18F-EF5 over a median 7-day period [32] Whilst the ATOM study did not formally test the reproducibility of hypoxia PET imaging, within the untreated cohort there were no significant differences between TBRmean between the two scans for the overall tumour or individual tumour regions.

Conclusions

In this study, we provide insights into the action of atovaquone in different subregions of NSCLC tumours. Following the demonstration that atovaquone reduces tumour hypoxia in patients with NSCLC, we show that this reduction mostly occurs in the inner and outer tumour subregions, and to a lesser extent in the centre subregion. Importantly, the most intensely hypoxic tumour voxels, and therefore the most radiobiologically resistant areas, are subject to the most pronounced decrease in hypoxia. This enables a better understanding of the role of atovaquone as a novel tumour radiosensitiser. Further investigation is required to assess whether the PET radiotracer may underestimate the level of hypoxia in the innermost tumour subregion and whether perfusion is perhaps responsible, at least in part, for this observation.

Availability of data and materials

Data are available under reasonable request to the corresponding author.

Abbreviations

CI:

Confidence interval

DNA:

Deoxyribonucleic acid

ETC:

Electron transport chain

18F:

Fluorine-18

FAZA:

18F-fluoroazomycinarabinofuranoside

FDG:

18F-fluorodeoxyglucose

FMISO:

18F-fluoromisonidazole

HV:

Hypoxic volume

IQR:

Interquartile range

MDC:

Minimum detectable change

NSCLC:

Non-small cell lung cancer

OCR:

Oxygen consumption rate

PET:

Positron emission tomography

RT:

Radiotherapy

SCLC:

Small cell lung cancer

SUV:

Standardised uptake value

SUVmean :

Mean SUV

TBR:

Tumour-to-blood ratio

TBRmean :

Mean TBR

References

  1. UK, C.R. Lung cancer statistics. 2021 [cited 2021 01/08/2021]. https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/lung-cancer#heading-One

  2. Molina JR, et al. Non-small cell lung cancer: epidemiology, risk factors, treatment, and survivorship. Mayo Clin Proc. 2008;83(5):584–94.

    Article  Google Scholar 

  3. Vinod SK, Hau E. Radiotherapy treatment for lung cancer: current status and future directions. Respirology. 2020;25(Suppl 2):61–71.

    Article  Google Scholar 

  4. Delaney GP, Barton MB. Evidence-based estimates of the demand for radiotherapy. Clin Oncol (R Coll Radiol). 2015;27(2):70–6.

    Article  CAS  Google Scholar 

  5. Wilson WR, Hay MP. Targeting hypoxia in cancer therapy. Nat Rev Cancer. 2011;11(6):393–410.

    Article  CAS  Google Scholar 

  6. Eales KL, Hollinshead KE, Tennant DA. Hypoxia and metabolic adaptation of cancer cells. Oncogenesis. 2016;5(1):e190.

    Article  CAS  Google Scholar 

  7. Hall EJ, Giaccia AJ. Radiobiology for the radiologist. 8th ed. Philadelphia: Wolters Kluwer; 2018.

    Google Scholar 

  8. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74.

    Article  CAS  Google Scholar 

  9. Dhani N, et al. The clinical significance of hypoxia in human cancers. Semin Nucl Med. 2015;45(2):110–21.

    Article  Google Scholar 

  10. Overgaard J. Hypoxic modification of radiotherapy in squamous cell carcinoma of the head and neck—a systematic review and meta-analysis. Radiother Oncol. 2011;100(1):22–32.

    Article  Google Scholar 

  11. Ashton TM, et al. Oxidative phosphorylation as an emerging target in cancer therapy. Clin Cancer Res. 2018;24(11):2482–90.

    Article  CAS  Google Scholar 

  12. Secomb TW, et al. Analysis of the effects of oxygen supply and demand on hypoxic fraction in tumors. Acta Oncol. 1995;34(3):313–6.

    Article  CAS  Google Scholar 

  13. Ashton TM, et al. The anti-malarial atovaquone increases radiosensitivity by alleviating tumour hypoxia. Nat Commun. 2016;7:12308.

    Article  CAS  Google Scholar 

  14. Haile LG, Flaherty JF. Atovaquone: a review. Ann Pharmacother. 1993;27(12):1488–94.

    Article  CAS  Google Scholar 

  15. Skwarski M, et al. Mitochondrial inhibitor atovaquone increases tumor oxygenation and inhibits hypoxic gene expression in patients with non-small cell lung cancer. Clin Cancer Res. 2021;27(9):2459–69.

    Article  CAS  Google Scholar 

  16. Koh WJ, et al. Imaging of hypoxia in human tumors with [F-18]fluoromisonidazole. Int J Radiat Oncol Biol Phys. 1992;22(1):199–212.

    Article  CAS  Google Scholar 

  17. McGowan DR, et al. Buparlisib with thoracic radiotherapy and its effect on tumour hypoxia: a phase I study in patients with advanced non-small cell lung carcinoma. Eur J Cancer. 2019;113:87–95.

    Article  CAS  Google Scholar 

  18. Copay AG, et al. Understanding the minimum clinically important difference: a review of concepts and methods. Spine J. 2007;7(5):541–6.

    Article  Google Scholar 

  19. McGowan DR, et al. (18) F-fluoromisonidazole uptake in advanced stage non-small cell lung cancer: a voxel-by-voxel PET kinetics study. Med Phys. 2017;44(9):4665–76.

    Article  CAS  Google Scholar 

  20. Krohn KA, Link JM, Mason RP. Molecular imaging of hypoxia. J Nucl Med. 2008;49(Suppl 2):129s-s148.

    Article  CAS  Google Scholar 

  21. Eschmann SM, et al. Prognostic impact of hypoxia imaging with 18F-misonidazole PET in non-small cell lung cancer and head and neck cancer before radiotherapy. J Nucl Med. 2005;46(2):253–60.

    PubMed  Google Scholar 

  22. Rischin D, et al. Prognostic significance of [18F]-misonidazole positron emission tomography-detected tumor hypoxia in patients with advanced head and neck cancer randomly assigned to chemoradiation with or without tirapazamine: a substudy of Trans-Tasman Radiation Oncology Group Study 98.02. J Clin Oncol. 2006;24(13):2098–104.

    Article  Google Scholar 

  23. Bettinardi V, et al. Physical performance of the new hybrid PET∕CT Discovery-690. Med Phys. 2011;38(10):5394–411.

    Article  CAS  Google Scholar 

  24. Grimes DR, Warren DR, Warren S. Hypoxia imaging and radiotherapy: bridging the resolution gap. Br J Radiol. 2017;90(1076):20160939.

    Article  Google Scholar 

  25. Carlin S, Humm JL. PET of hypoxia: current and future perspectives. J Nucl Med. 2012;53(8):1171–4.

    Article  CAS  Google Scholar 

  26. Horsman MR, et al. Imaging hypoxia to improve radiotherapy outcome. Nat Rev Clin Oncol. 2012;9(12):674–87.

    Article  CAS  Google Scholar 

  27. Bentzen SM, Gregoire V. Molecular imaging-based dose painting: a novel paradigm for radiation therapy prescription. Semin Radiat Oncol. 2011;21(2):101–10.

    Article  Google Scholar 

  28. Jentsch C, et al. Validation of functional imaging as a biomarker for radiation treatment response. Br J Radiol. 2015;88(1051):20150014.

    Article  CAS  Google Scholar 

  29. Chang JH, et al. Hypoxia-targeted radiotherapy dose painting for head and neck cancer using (18)F-FMISO PET: a biological modeling study. Acta Oncol. 2013;52(8):1723–9.

    Article  CAS  Google Scholar 

  30. Okamoto S, et al. High reproducibility of tumor hypoxia evaluated by 18F-fluoromisonidazole PET for head and neck cancer. J Nucl Med. 2013;54(2):201–7.

    Article  CAS  Google Scholar 

  31. Grkovski M, et al. Reproducibility of (18)F-fluoromisonidazole intratumour distribution in non-small cell lung cancer. EJNMMI Res. 2016;6(1):79.

    Article  Google Scholar 

  32. Silvoniemi A, et al. Repeatability of tumour hypoxia imaging using [(18)F]EF5 PET/CT in head and neck cancer. Eur J Nucl Med Mol Imaging. 2018;45(2):161–9.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We thank all patients and their families for taking part and the staff at Oxford University Hospitals National Health Service Foundation Trust. This trial was sponsored by the University of Oxford. Trial management was provided by the Oncology Clinical Trials Office at the University of Oxford as part of the UKCRC Oxford Clinical Trials Research Unit.

Funding

This work was funded by the Howat Foundation (C5255/A25069). Additional support was obtained from CRUK & EPSRC Cancer Imaging Centre in Oxford (C5255/A16466), CRUK/MRC Oxford Institute for Radiation Oncology (C5255/A23755), and the NIHR Oxford Experimental Cancer Medicine Centre. DRM was supported by a NIHR/HEE Clinical Lectureship (ICA-CL-2016-02-009) and CRUK National Cancer Imaging Translational Accelerator (C34326/A28684 and C42780/A27066). GSH was supported by a CRUK Clinician Scientist Fellowship (C34326/A19590).

Author information

Authors and Affiliations

Authors

Contributions

PB performed the data analysis, interpreted the data, and wrote the initial manuscript. MS designed the work and acquired the data. REM acquired the data. GSH designed the work, acquired the data, and interpreted the data. DRM designed the work, acquired the data, and interpreted the data. All authors read, edited, and approved the final manuscript.

Corresponding author

Correspondence to Daniel R. McGowan.

Ethics declarations

Ethics approval and consent to participate

Ethical approval was obtained from National Research Ethics Service Committee South Central Oxford B (16/SC/0012). Trial conduct adhered to all regulatory requirements and was in full accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines.

Consent for publication

All patients provided written informed consent.

Competing interests

No conflicts of interest to declare.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Figure 1.

Change in tumour HV measured by hypoxia PET-CT. Supplementary Figure 2. Influence of treatment time on tumour hypoxic volume changes. Supplementary Figure 3. Relationship between baseline tumour hypoxic volume and total baseline tumour volume.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bourigault, P., Skwarski, M., Macpherson, R.E. et al. Investigation of atovaquone-induced spatial changes in tumour hypoxia assessed by hypoxia PET/CT in non-small cell lung cancer patients. EJNMMI Res 11, 130 (2021). https://doi.org/10.1186/s13550-021-00871-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13550-021-00871-x

Keywords

  • Non-small cell lung cancer
  • Tumour hypoxia
  • Atovaquone
  • PET
  • FMISO