Purpose: As the interest in antiangiogenesis therapy in oncology is rising, the need for in vivo techniques to monitor such therapy is obvious. Measurement of tumor perfusion using positron emission tomography and H215O potentially is such a technique. The objective of the present study was to assess whether it is feasible to measure perfusion in vivo in non-small cell lung cancer (NSCLC) using H215O and positron emission tomography.

Experimental Design: Fifteen dynamic H215O and [18F]2-fluoro-2-deoxy-d-glucose (18FDG) studies were performed in 10 patients with stage IIIA-N2 NSCLC. Blood flow (BF) data were correlated with simplified methods of analysis (tumor:normal tissue ratio and standardized uptake value) and with glucose metabolism (MRglu).

Results:18FDG data were required for accurate definition of tumor and mediastinal lymph node metastases. There was large intertumor variation in BF. Correlation of simplified methods of analysis with quantitative BF was poor. In addition, BF and MRglu were not correlated.

Conclusion: Measurement of BF in NSCLC using H215O and 18FDG is feasible. Simple uptake analysis, however, cannot be used as an indicator of perfusion. Whether BF can be used for response monitoring needs to be evaluated in a large patient study, where results can be compared with outcome.

For the development of novel cancer treatment strategies, knowledge of tumor biology is essential. Heterogeneity of tumor perfusion has therapeutic consequences for both drug delivery and distribution and for oxygenation status. If tumor blood flow could be monitored noninvasively, more insight in the efficacy of therapy could be obtained. For example, in a tumor with very poor perfusion, it could be predicted that chemotherapeutics would not reach the tumor and thus not affect it, and other therapy options would have to be considered.

Presently, the interest in inhibition of vascular growth of tumors (antiangiogenesis) as a possible therapeutic strategy increases. Monitoring the effects of angiogenesis inhibitors requires a noninvasive technique to assess the perfusion status of the tumor.

Using PET2, it is possible to perform physiological measurements in vivo. Most oncology studies are performed using 18FDG, which is a tracer of glucose metabolism. However, in those studies where information of the perfusion status of the tumor is needed, water labeled with oxygen-15 (H215O) may be used. Initial PET oncology studies using H215O were performed in brain tumors (1, 2, 3, 4, 5, 6, 7). In tumors outside the central nervous system, only a few studies have been reported (8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19), using a variety of different techniques. To the best of our knowledge, no studies have been reported on the use of H215O in lung tumors.

The aim of the present study was to develop a method for the measurement of blood flow in lung tumors. After a brief overview of background and theory, initial results of measuring perfusion in stage IIIA-N2 NSCLC patients are presented.

### Background.

BF can be measured using different tracers and techniques. Of the various tracers, H215O has the advantage that it is freely diffusible and metabolically inert. In addition, because of the very short half-life of 15O, repeat or combined measurements within a single scanning session can be made. Although several techniques to measure BF using H215O have been described, they are all based on the original tracer kinetic model proposed by Kety (20).

The steady-state technique, a noninvasive inhalation technique using the steady-state principle, was first described by Jones et al.(21) and later implemented for PET by Frackowiak et al.(22). In this technique, a static PET scan is performed during continuous inhalation of 15O-labeled carbon dioxide (C15O2). C15O2 is rapidly transferred in the lungs to the waterpool under the influence of carbon anhydrase (23). After an inhalation period of ∼10 min and because of the short half-life of 15O, tissue H215O will reach a dynamic equilibrium in which the diffusion rate from arterial blood into the tissue is balanced by the diffusion rate out of tissue into venous blood and the rate of radioactive decay. The actual scan is started after this equilibration period. Within oncology, the technique has been applied mainly in brain tumor studies (1, 2, 4, 24), in a study in patients with breast carcinoma by Beaney et al.(8), and in 2 patients with hepatic tumors by Taniguchi et al.(19).

Advantages of the steady-state technique are its simple implementation, even for obtaining functional blood flow images, and the possibility to improve the statistical quality of the data simply by increasing the duration of the scan. An important disadvantage of the steady-state technique is the underestimation of BF in heterogeneous tissue (25, 26). Tumor is known to be nonhomogeneous, being an admixture of clusters of tumor and normal cells, vascular structures, and necrotic tissue. The steady-state technique assumes that the partition coefficient of water equals 1, which might not be valid in tumors (25), and because of the equilibration period prior to scanning, rather inefficient use is made of the administered radiation dose.

As an alternative to the steady-state technique, the so-called autoradiographic method was developed for the first (slow) generation of PET scanners (27, 28, 29). In this method, the integral counts over the first period after H215O injection (i.e., single frame study) are collected. BF is calculated using the measured arterial input function (multiple samples), assuming a fixed partition coefficient of water. The method was validated for the brain using a 40-s integration period after arrival of H215O in the brain. An advantage over the steady-state method is the much shorter acquisition time, however, at the cost of decreased statistical quality. In addition, there is a more linear relationship between counts and BF than in the steady-state technique. A disadvantage over the steady-state technique is increased sensitivity to the presence of arterial blood. In addition, results were found to be dependent on the integration (acquisition) time, possibly because of delay and dispersion of the arterial input function. The method shares an important disadvantage with the steady-state technique, at least in oncological applications, in that a value for the partition coefficient of water has to be assumed. To the best of our knowledge, the method has not been applied to tumors outside of the brain.

With the introduction of fast multiring PET scanners, dynamic blood flow methods were developed and validated for the brain (30), myocardium (31, 32, 33, 34), and tumors (12) outside the central nervous system. Reproducibility in the brain was found to be better than 10% (33). The main advantage of this dynamic technique is that no value for the partition coefficient or volume of distribution of water (Vd) has to be assumed, thereby significantly increasing the accuracy of the BF measurements. In addition, it has been demonstrated that the sensitivity to tissue heterogeneity is low (26), and that the flow estimates are independent of scan duration. Finally, when necessary, it is possible to account for contamination of the tissue signal within arterial blood activity.

To quantify BF an arterial input function is required. The most accurate method to determine this input function is continuous on-line arterial blood sampling. In patients for whom repetitive scans are needed, however, arterial cannulation should, whenever possible, be avoided. For studies of the myocardium and in breast tumor patients, input curves obtained from both the left atrium and left ventricle have been used and validated (34, 35, 36).

Several studies have been performed using a H215O bolus injection and PET to assess blood flow in different tumor types (9, 10, 11, 12, 13, 14, 15, 16, 18). Some of these studies used the SUV or differential absorption ratio (9, 13, 15, 16, 18) as a measure of tumor BF. Although SUV is used extensively as a measure of glucose metabolism in 18FDG studies, its use is controversial (37). Criticism of the use of SUV would be even more relevant to H215O studies (e.g., dependence on scan duration and measurement time), and no comparison with measured BF values has been made.

As part of an ongoing response monitoring study, 15 dynamic H215O scans were performed in 10 patients using dedicated PET scanners ECAT EXACT HR+ (Siemens/CTI). All patients (7 men, 3 women; mean age, 66.5 years) were clinically staged as having stage IIIA-N2 NSCLC. Patient characteristics are listed in Table 1. The study protocol was approved by the medical ethics committee of both hospitals participating in this study. All patients gave written informed consent.

First, a 10-min transmission scan over the tumor area was performed, followed by injection of H215O using an automated injector (Med-Rad multilevel CT injector), simultaneously starting a 10-min dynamic emission scan (12 × 5 min, 12 × 10 min, 6 × 20 min, and 10 × 30 min). Injection (1000 MBq dissolved in 2.5 ml) was given as a bolus (15 s; 10 ml/min), followed by a flush of 2 min. Ten min after the H215O scan (to allow for radioactive decay of 15O), 370 MBq of 18FDG were injected, and a second (60-min) dynamic emission scan with progressive frame lengths was started.

### Data Analysis.

All dynamic scan data were corrected for dead time, decay, scatter, randoms, and photon attenuation and were reconstructed using FBP with a Hanning filter (cutoff, 0.5 cycles/pixel). This resulted in a transaxial spatial resolution of ∼7 mm of full width at half maximum.

ROI were defined automatically over both tumor and MLN metastases by applying a threshold of 50% of the maximum 18FDG pixel value within the lesion. For this purpose, the last three frames (i.e., 45–60 min after injection) of the sinograms of the 18FDG scan were summed and reconstructed using ordered subset expectation maximization with 2 iterations and 12 subsets. This was followed by postsmoothing of the reconstructed image using a 5-mm full width at half maximum Gaussian filter to obtain the same resolution as the dynamic images reconstructed with FBP. The ordered subset expectation maximization reconstructions were used because of their superior image quality, thus facilitating ROI definition. For quantification, however, the more accurate FBP reconstructed images were used (38), and therefore, tumor ROI were copied to the (FBP reconstructed) dynamic H215O data to create tumor time activity curves. Furthermore, ROI were defined manually over the aortic arch, left ventricle, and left atrium on a summed (FBP) image of the first min of the H215O scan. These data were used to create an image-derived input function.

Time activity curves were analyzed using standard nonlinear regression techniques and the single compartment model, both with and without an arterial blood volume component (see Appendix , Eqs. Cand D, respectively), weighting data for acquired counts and frame duration. In addition, to assess the validity of data presented in the literature, T/N ratios from 0 to 5 min and the SUV corrected for body surface area (SUVBSA) from 0 to 5 and from 4 to 5 min were evaluated.

Glucose consumption (MRglu) was obtained from the 18FDG data using nonlinear regression and the standard two-tissue (3k) compartment model (39) with three rate constants and a blood volume component (40) and an image derived input function (41).

### Statistics.

The presence of an arterial blood volume component in the tumor curves was assessed by comparing the residual sum of squares with and without such a component using the Akaike (42) and Schwarz (43) criteria. The correlation of T/N and SUVBSA values with both BF and Vd values was assessed, using the Pearson bivariate correlation; P < 0.01 was considered to be significant. The correlation of MRglu with BF values was assessed, using Pearson bivariate correlation; again, P < 0.01 was considered to be significant.

In Fig. 1, transaxial images at the level of tumor and MLN metastases are shown for both H215O and 18FDG data. It can be seen that the tumor could have been identified and defined on the H215O scan itself. However, for the definition of MLN metastases, the 18FDG scan was needed. Therefore, all tissue ROI were defined on the 18FDG scan.

ROI to generate the image-derived input function were defined on summed dynamic H215O data. For the input function, as many time-activity curves as possible were averaged (weighted to ROI size) to obtain optimal statistics. In 8 scans, the curves from the aortic arch could be used. ROI defined in both the left ventricle and atrium could be used in 13 scans. An example of an average blood time-activity curve (input function) is given in Fig. 2 A.

Examples of fits with and without an arterial blood volume component are given in Fig. 2 B. In 14 of 15 scans (93%), incorporating an arterial blood volume component provided significantly better fits according to the Akaike criterion. This was 12 of 15 (80%) according to the Schwarz criterion. Therefore, all results presented here are for the model with arterial blood volume component (i.e., according to Eq. D).

Individual values of T/N, SUVBSA, BF, and Vd are presented in Table 2. The mean ± SD for blood flow in tumor was 0.59 ± 0.37 ml/ml/min, with a range of 0.15–1.29 ml/ml/min. For MLN metastases, this was 0.48 ± 0.23 ml/ml/min with a range of 0.21–0.86 ml/ml/min. The mean ± SD value for Vd was 0.63 ± 0.10 with a range of 0.45–0.78 for tumor and 0.74 ± 0.17 with a range of 0.47–0.99 for MLN metastases, respectively.

The correlation between BF and SUVBSA (0–5 min) was 0.60 (P = 0.02) and 0.32 (P = 0.25) for SUVBSA (4–5 min), which were both not significant. The correlation between BF and T/N ratio was 0.58 (P = 0.02). SUVBSA (0–5 min) was also poorly correlated with Vd (r = 0.49, P = 0.06); for SUVBSA (4–5 min), this was 0.37 (P = 0.18).

Values of MRglu are also given in Table 2. The correlation with BF was poor (r = 0.28, P = 0.31), as illustrated in Fig. 3.

Inhibition of angiogenesis in tumors is an important therapeutic aim because it might be a means of preventing progression of disease. Presently, many studies are performed to assess the possibilities of such therapy. It is important to be able to monitor the effects of antiangiogenesis therapy in vivo, preferably in a simple and noninvasive manner. A potential means to monitor these effects is the measurement of tumor perfusion. Hence, there is renewed interest in blood flow measurements using PET and H215O.

The most important finding of the present study in lung tumors is the large variation in BF in different tumors, even in the untreated ones. It is likely that this variation has implications for therapy (drug delivery, oxygenation, and others). Therefore, studies are needed to relate perfusion measurements with response to treatment.

In the present study in lung tumors, ROI for tissue time-activity curves and input curves were defined using different datasets. The image-derived input curve was defined using summed (first min) H215O ROI being defined over the aortic arch, left ventricle, and atrium. The use of the last two has been validated previously for myocardial studies (34). In contrast to 18FDG, only minor spillover effects from the myocardium occur. An input curve derived from the aortic arch has not been described for H215O. When a tumor is located in the apex of the lung, the left ventricle and atrium may not be in the field of view of the scan, resulting in the need for another structure to generate an input function. A disadvantage of the use of the aortic arch is the higher risk of partial volume effects. In the present study in 8 of 15 scans, the input curve from the aortic arch could be used when compared with curves from the left atrium and ventricle. Because H215O distributes rapidly throughout the body and because of the short half-life of 15O, as many ROI as possible are needed to obtain accurate statistics.

Tissue time-activity curves were defined using the last 15 min of the 18FDG data. Although summing of the first 5 min of the H215O data provided an acceptable image for the definition of tumor in most patients, the definition of MLN metastases was difficult if not impossible using H215O data alone (Fig. 1). A disadvantage of the use of tumor ROI defined on 18FDG data are the time gap with the actual H215O scan, thereby increasing the risk of patient movement and thus inappropriate tumor ROI. However, it was not feasible to consistently define an accurate tumor ROI on H215O data alone, and for this reason, 18FDG scanning proved to be essential. Very recently, a number of alternative methods for generating parametric images of tumor BF using H215O have been described (44). In 5 patients with renal cell metastases in the thorax, the lesions could be readily identified in the parametric flow images. Further studies are needed, however, to assess whether those methods would also identify MLN metastases in NSCLC and thus obviate the need for 18FDG in defining ROI. The same applies for parametric methods, which thus far have only been applied to the brain (45, 46, 47, 48) or heart (49). In theory, ROI could also be defined on CT scans. This, however, would require accurate positioning and realignment between CT and PET scans. In practice, this is not possible because patient positioning routinely is different for CT (hands up) and PET (hands down).

In the model used, it is assumed that water is freely diffusible, i.e., that the extraction fraction is 100%. This is likely to be the case for low flow values. For high flow values, however, extraction could be <100%, resulting in an underestimation of flow. This effect is expected to be small because blood vessels supplying a tumor are usually more permeable than normal blood vessels.

In studies performed in brain, the mean volume of distribution of water was found to be 0.86 with a SD of 0.04 (50). For comparison, the mean value of Vd found in breast tumors was 0.56 with a SD of 0.15 (12). In this study, a mean value of 0.63 for NSCLC was found (Table 2), which is very similar to the value found in breast tumors. It should be noted that the estimation of Vd is susceptible to tissue heterogeneity. For example, in the brain the underestimation attributable to tissue heterogeneity can be as high as 30% (26). In contrast, in the same study, the effect of tissue heterogeneity on flow was <5% (26).

An arterial blood volume component significantly contributing to the counts within the tumor ROI was present in 80–93% of the scans in this study. In addition, for MLN metastases this was 100%. In brain and breast cancer studies (12, 30, 33), the contribution of an arterial blood volume component was found to be negligible. However, in NSCLC, the contribution needs to be taken into account (see, for example, Fig. 2 B).

In this study, there was no significant correlation between BF and SUV or T/N ratio. Although the study was limited to a single tumor type, it is likely that the same applies to other tumors. In other words, unless validated for the tumor under investigation, both SUV and T/N ratios cannot be used as indicators of tumor perfusion.

When comparing MRglu with BF, a poor correlation was found (Fig. 3). This would indicate that perfusion and glucose consumption are not coupled in NSCLC, and therefore, as expected, H215O and 18FDG provide complimentary information on tumor physiology. Whether BF alone, or in combination with 18FDG, may predict response to therapy will have to be investigated in a larger patient study where results can be compared with outcome.

In conclusion, although the uptake of H215O in NSCLC is higher than in normal lung tissue (T/N >1), it is difficult to determine the exact location. This is especially true for possible MLN metastases because of their position close to vascular structures with high tracer concentration. Therefore at present, the study needs to be combined with, for example, an 18FDG study for accurate ROI definition.

When performing dynamic blood flow PET using H215O in NSCLC, the influence of an arterial blood volume component has to be taken into account. The volume of distribution of water is similar to reported values for breast tumors.

An important finding is the large intertumor variation in perfusion. This could indicate that perfusion might be an important parameter for predicting therapy efficacy. The actual additional value of PET using H215O for monitoring antiangiogenesis therapy needs to be evaluated further in a larger serial patient study. In addition, the value of combining H215O and 18FDG studies will need further investigation.

For the present study, flow was measured using the dynamic method based on the tracer kinetic model originally described by Kety (20, 51). The single tissue compartment model is illustrated in Fig. 4, where it is assumed that all concentrations are corrected for decay. The rate of change of activity in tissue is the balance between delivery and washout, or in mathematical terms:

$dC_{\mathrm{t}}/\mathrm{dt}\ {=}\ \mathrm{FC}_{\mathrm{a}}\ {-}\ (\mathrm{F/V}_{\mathrm{d}})\ {\cdot}\ \mathrm{Ct}$

where F is BF or perfusion (ml blood/ml tissue/min), Vd is the partition coefficient or volume of distribution of water (unitless), Ca is the concentration of H215O in arterial blood (kBq/ml), and Ct is the concentration of H215O in tissue (kBq/ml).

The solution of this differential equation is:

$C_{\mathrm{t}}\ {=}\ \mathrm{F}\ {\cdot}\ \mathrm{C}_{\mathrm{a}}\ {\otimes}\ \mathrm{e}^{{-}\mathrm{F/Vd}{\cdot}\mathrm{t}}$

where ⊗ represents the operation of convolution.

The method is based on the following assumptions: (a) flow (F) and Vd are constant during the period of measurement, i.e., the system is in a physiological steady state; (b) H215O is a freely diffusible tracer, i.e., the extraction fraction of water is unity, and no binding of water in tissue occurs; (c) H215O that diffuses into tissue equilibrates instantaneously within the tissue, i.e., there are no concentration gradients in tissue; (d) venous and tissue concentrations do not differ, i.e., the volume of distribution (Vd), or the partition coefficient, is close to 1. This means that the venous blood can be considered to be at tissue concentration and, therefore, venous blood and tissue can be considered to be a single compartment; and (e) the contribution of arterial blood activity, i.e., the signal arising from arterial activity within a ROI is negligible.

The last assumption states that tissue and (measured) ROI concentration are the same, i.e.:

$C_{\mathrm{ROI}}\ {=}\ \mathrm{C}_{\mathrm{t}}$

where Ct is given by Eq. B.

If the signal arising from the arterial blood is not negligible, this can be accounted for by incorporating an additional arterial blood volume (Va) term resulting in:

$C_{\mathrm{ROI}}\ {=}\ (1\ {-}\ \mathrm{V}_{\mathrm{a}})\ \mathrm{C}_{\mathrm{t}}\ {+}\ \mathrm{V}_{\mathrm{a}}\mathrm{C}_{\mathrm{a}}$

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

2

The abbreviations used are: PET, positron emission tomography; 18FDG, [18F]2-fluoro-2-deoxy-d-glucose; NSCLC, non-small cell lung cancer; BF, blood flow; SUV, standard uptake value; FBP, filtered back projection; ROI, regions of interest; MLN, mediastinal lymph node; T/N, tumor:normal tissue.

Fig. 1.

This figure shows H215O (A and B) and 18FDG (C and D) images for a patient with a left-sided lung tumor and a subcarinal lymph node metastasis. The H215O images are integral images over the period from 1 to 5 min after injection. The 18FDG images are integrated from 45 to 60 min after injection. Both axial slices at the level of primary tumor (A and C) and MLN metastasis (B and D) are shown. The tumor can clearly be seen in the H215O image; however, the MLN metastasis is difficult to detect based on the H215O image alone.

Fig. 1.

This figure shows H215O (A and B) and 18FDG (C and D) images for a patient with a left-sided lung tumor and a subcarinal lymph node metastasis. The H215O images are integral images over the period from 1 to 5 min after injection. The 18FDG images are integrated from 45 to 60 min after injection. Both axial slices at the level of primary tumor (A and C) and MLN metastasis (B and D) are shown. The tumor can clearly be seen in the H215O image; however, the MLN metastasis is difficult to detect based on the H215O image alone.

Close modal
Fig. 2.

A, example of an average blood time-activity curve (image derived input function). B, example of fits of a tumor ROI with (solid line) and without (dashed line) an arterial blood volume included in the fitting equation. The corresponding input function is given in A.

Fig. 2.

A, example of an average blood time-activity curve (image derived input function). B, example of fits of a tumor ROI with (solid line) and without (dashed line) an arterial blood volume included in the fitting equation. The corresponding input function is given in A.

Close modal
Fig. 3.

Scatter diagram of MRgluversus BF, illustrating poor correlation (r = 0.28, P = 0.31).

Fig. 3.

Scatter diagram of MRgluversus BF, illustrating poor correlation (r = 0.28, P = 0.31).

Close modal
Fig. 4.

Schematic diagram of the single tissue compartment model. For a full description, see the main text.

Fig. 4.

Schematic diagram of the single tissue compartment model. For a full description, see the main text.

Close modal
Table 1

Patient characteristics

n
Gender
Male
Female
Age
Mean 66.5 years
Range 45–76 years
Histology
Squamous cell carcinoma
Large cell carcinoma
Planocellular carcinoma
Time scan
Prior to start chemotherapy
After 1 cycle of chemotherapy
After 3 cycles of chemotherapy
n
Gender
Male
Female
Age
Mean 66.5 years
Range 45–76 years
Histology
Squamous cell carcinoma
Large cell carcinoma
Planocellular carcinoma
Time scan
Prior to start chemotherapy
After 1 cycle of chemotherapy
After 3 cycles of chemotherapy
Table 2

Patient Data

PatientTime of scanningAgeHistology tumorT/N (0–5 min)SUVbsa (4–5 min)SUVbsa (0–5 min)BF (ml/ml/min)V                  dMRglu (micromol/ml/min)
Prior to start chemotherapy 67 Large cell carcinoma 95 112 0.53 0.62 0.119
Prior to start second cycle chemotherapy 76 Planocellular carcinoma 2.7 82 111 1.16 0.78 0.169
3a Prior to start chemotherapy 70 Adenocarcinoma 106 127 0.95 0.76 0.210
3b Prior to start second cycle chemotherapy   2.5 91 104 0.77 0.69 0.141
Prior to start chemotherapy 66 Adenocarcinoma 1.6 80 69 0.19 0.71 0.165
Prior to start second cycle chemotherapy 76 Large cell carcinoma 2.2 70 76 0.25 0.65 0.122
Prior to start second cycle chemotherapy 62 Adenocarcinoma 3.9 85 100 0.90 0.74 0.097
7a Prior to start chemotherapy 62 Adenocarcinoma 2.4 71 89 0.53 0.68 0.137
7b Prior to start second cycle chemotherapy   2.2 55 75 1.29 0.52 0.090
7c Prior to start third (last) cycle chemotherapy   1.8 63 85 0.50 0.55 0.054
8a Prior to start chemotherapy 74 Squamous cell carcinoma 1.7 80 74 0.43 0.63 0.235
8b Prior to start second cycle chemotherapy   1.9 59 65 0.15 0.61 0.070
8c Prior to start third (last) cycle chemotherapy   1.3 53 53 0.22 0.45 0.071
Prior to start third (last) cycle chemotherapy 68 Squamous cell carcinoma 1.4 55 59 0.21 0.48 0.035
10 Prior to start chemotherapy  Adenocarcinoma 1.8 144 141 0.73 0.56 0.170
PatientTime of scanningAgeHistology tumorT/N (0–5 min)SUVbsa (4–5 min)SUVbsa (0–5 min)BF (ml/ml/min)V                  dMRglu (micromol/ml/min)
Prior to start chemotherapy 67 Large cell carcinoma 95 112 0.53 0.62 0.119
Prior to start second cycle chemotherapy 76 Planocellular carcinoma 2.7 82 111 1.16 0.78 0.169
3a Prior to start chemotherapy 70 Adenocarcinoma 106 127 0.95 0.76 0.210
3b Prior to start second cycle chemotherapy   2.5 91 104 0.77 0.69 0.141
Prior to start chemotherapy 66 Adenocarcinoma 1.6 80 69 0.19 0.71 0.165
Prior to start second cycle chemotherapy 76 Large cell carcinoma 2.2 70 76 0.25 0.65 0.122
Prior to start second cycle chemotherapy 62 Adenocarcinoma 3.9 85 100 0.90 0.74 0.097
7a Prior to start chemotherapy 62 Adenocarcinoma 2.4 71 89 0.53 0.68 0.137
7b Prior to start second cycle chemotherapy   2.2 55 75 1.29 0.52 0.090
7c Prior to start third (last) cycle chemotherapy   1.8 63 85 0.50 0.55 0.054
8a Prior to start chemotherapy 74 Squamous cell carcinoma 1.7 80 74 0.43 0.63 0.235
8b Prior to start second cycle chemotherapy   1.9 59 65 0.15 0.61 0.070
8c Prior to start third (last) cycle chemotherapy   1.3 53 53 0.22 0.45 0.071
Prior to start third (last) cycle chemotherapy 68 Squamous cell carcinoma 1.4 55 59 0.21 0.48 0.035
10 Prior to start chemotherapy  Adenocarcinoma 1.8 144 141 0.73 0.56 0.170
1
Ito M., Lammertsma A. A., Wise R. J., Bernardi S., Frackowiak R. S., Heather J. D., McKenzie C. G., Thomas D. G., Jones T. Measurement of regional cerebral blood flow and oxygen utilisation in patients with cerebral tumours using 15O and positron emission tomography: analytical techniques and preliminary results.
,
23
:
63
-74,
1982
.
2
Lammertsma A. A., Wise R. J., Jones T. In vivo measurements of regional cerebral blood flow and blood volume in patients with brain tumours using positron emission tomography.
Acta Neurochir.
,
69
:
5
-13,
1983
.
3
Beaney R. P., Brooks D. J., Leenders K. L., Thomas D. G., Jones T., Halnan K. E. Blood flow and oxygen utilisation in the contralateral cerebral cortex of patients with untreated intracranial tumours as studied by positron emission tomography, with observations on the effect of decompressive surgery.
J. Neurol. Neurosurg. Psychiatry
,
48
:
310
-319,
1985
.
4
Lammertsma A. A., Wise R. J., Cox T. C., Thomas D. G., Jones T. Measurement of blood flow, oxygen utilisation, oxygen extraction ratio, and fractional blood volume in human brain tumours and surrounding oedematous tissue.
,
58
:
725
-734,
1985
.
5
Mineura K., Yasuda T., Kowada M., Shishido F., Ogawa T., Uemura K. .
Quantitative analysis of regional cerebral hemodynamics and metabolism in gliomas using O-15 and F-18-fluorodeoxyglucose (FDG) PET. Proceedings of the International Symposium on Current and Future Aspects of Cancer Diagnosis with Positron Emission Tomography
,
82
-85, Senda Japan
1985
.
6
Taki W., Handa H., Ishikawa M., Kobayashi A., Yamashita J., Yonekawa Y., Tanada S., Senda M., Yonekura Y., Torizuka K., Fujita T., Fukuyama H., Fukimota N., Harada K., Kameyama M. .
Blood flow and oxygen utilization in brain tumors. Proceedings of the International Symposium on Current and Future Aspects of Cancer Diagnosis with Positron Emission Tomography
,
91
-95, Tohoku University
1985
.
7
Mineura K., Yasuda T., Kowada M., Shishido F., Ogawa T., Uemura K. Positron emission tomographic evaluation of histological malignancy in gliomas using oxygen-15 and fluorine-18-fluorodeoxyglucose.
Neurol. Res.
,
8
:
164
-168,
1986
.
8
Beaney R. P., Lammertsma A. A., Jones T., McKenzie C. G., Halnan K. E. Positron emission tomography for in-vivo measurement of regional blood flow, oxygen utilisation, and blood volume in patients with breast carcinoma.
Lancet
,
1
:
131
-134,
1984
.
9
Strauss L. G., Clorius J. H., Schlag P., Lehner B., Kimmig B., Engenhart R., Marin-Grez M., Helus F., Oberdorfer F., Schmidlin P. Recurrence of colorectal tumors: PET evaluation.
,
170
:
329
-332,
1989
.
10
Kubo S., Yamamoto K., Magata Y., Iwasaki Y., Tamaki N., Yonekura Y., Konishi J. Assessment of pancreatic blood flow with positron emission tomography and oxygen-15 water.
Ann. Nucl. Med.
,
5
:
133
-138,
1991
.
11
Wilson C. B., Snook D. E., Dhokia B., Taylor C. V., Watson I. A., Lammertsma A. A., Lambrecht R., Waxman J., Jones T., Epenetos A. A. Quantitative measurement of monoclonal antibody distribution and blood flow using positron emission tomography and 124iodine in patients with breast cancer.
Int. J. Cancer
,
47
:
344
-347,
1991
.
12
Wilson C. B., Lammertsma A. A., McKenzie C. G., Sikora K., Jones T. Measurements of blood flow and exchanging water space in breast tumors using positron emission tomography: a rapid and noninvasive dynamic method.
Cancer Res.
,
52
:
1592
-1597,
1992
.
13
Dimitrakopoulou-Strauss A., Strauss L. S., Goldschmidt H., Lorenz W. J., Maier-Borst W., van Kaick G. Evaluation of tumour metabolism and multidrug resistance in patients with treated malignant lymphomas.
Eur. J. Nucl. Med.
,
22
:
434
-442,
1995
.
14
Ziegler S. I., Haberkorn U., Byrne H., Tong C., Kaja S., Richolt J. A., Schosser R., Krieter H., Lammertsma A. A., Price P. Measurement of liver blood flow using oxygen-15 labelled water and dynamic positron emission tomography: limitations of model description.
Eur. J. Nucl. Med.
,
23
:
169
-177,
1996
.
15
Kuhn G., Reiser C., Dimitrakopoulou-Strauss A., Oberdorfer F., Strauss L. G. PET studies of perfusion and glucose metabolism in patients with untreated head and neck tumours.
Onkologie
,
20
:
226
-230,
1997
.
16
Dimitrakopoulou-Strauss A., Strauss L. G., Schlag P., Hohenberger P., Irngartinger G., Oberdorfer F., Doll J., van Kaick G. Intravenous and intra-arterial oxygen-15-labeled water and fluorine-18- labeled fluorouracil in patients with liver metastases from colorectal carcinoma.
J. Nucl. Med.
,
39
:
465
-473,
1998
.
17
Ponto L. L. B., Madsen M. T., Hichwa R. D., Mayr N., Yuh W. T. C., Magnotta V. A., Watkins G. L., Ehrhardt J. C. Assessment of blood flow in solid tumors using PET.
Clin. Positron Imaging
,
1
:
117
-121,
1998
.
18
Schwarzbach M., Willeke F., Dimitrakopoulou-Strauss A., Strauss L. G., Zhang Y. M., Mechtersheimer G., Hinz U., Lehnert T., Herfarth C. Functional imaging and detection of local recurrence in soft tissue sarcomas by positron emission tomography.
Anticancer Res.
,
19
:
1343
-1349,
1999
.
19
Taniguchi H., Yamaguchi A., Kunishima S., Koh T., Masuyama M., Koyama H., Oguro A., Yamagishi H. Using the spleen for time-delay correction of the input function in measuring hepatic blood flow with oxygen-15 water by dynamic PET.
Ann. Nucl. Med.
,
13
:
215
-221,
1999
.
20
Kety S. The theory and application of the exchange of inert gas at the lungs and tissues.
Pharmacol. Rev.
,
3
:
1
-41,
1951
.
21
Jones T., Chesler D. A., Ter Pogossian M. M. The continuous inhalation of oxygen-15 for assessing regional oxygen extraction in the brain of man.
,
49
:
339
-343,
1976
.
22
Frackowiak R. S., Lenzi G. L., Jones T., Heather J. D. Quantitative measurement of regional cerebral blood flow and oxygen metabolism in man using 15O and positron emission tomography: theory, procedure, and normal values.
J. Comput. Assist. Tomogr.
,
4
:
727
-736,
1980
.
23
West J. B., Dollery C. T. Uptake of oxygen-15 labeled CO2 compared with carbon-11 labeled CO2 in the lung.
J. Appl. Physiol.
,
17
:
9
-13,
1962
.
24
Brooks D. J., Beaney R. P., Thomas D. G., Marshall J., Jones T. Studies on regional cerebral pH in patients with cerebral tumours using continuous inhalation of 11CO2 and positron emission tomography.
J. Cereb. Blood Flow Metab.
,
6
:
529
-535,
1986
.
25
Lammertsma A. A., Jones T. Low oxygen extraction fraction in tumours measured with the oxygen-15 steady state technique: effect of tissue heterogeneity.
,
65
:
697
-700,
1992
.
26
Blomqvist G., Lammertsma A. A., Mazoyer B., Wienhard K. Effect of tissue heterogeneity on quantification in positron emission tomography.
Eur. J. Nucl. Med.
,
22
:
652
-663,
1995
.
27
Ginsberg M. D., Lockwood A. H., Busto R., Finn R. D., Butler C. M., Cendan I. E., Goddard J. A simplified in vivo autoradiographic strategy for the determination of regional cerebral blood flow by positron emission tomography: theoretical considerations and validation studies in the rat.
J. Cereb. Blood Flow Metab.
,
2
:
89
-98,
1982
.
28
Herscovitch P., Markham J., Raichle M. E. Brain blood flow measured with intravenous H215O. I. Theory and error analysis.
J. Nucl. Med.
,
24
:
782
-789,
1983
.
29
Raichle M. E., Martin W. R., Herscovitch P., Mintun M. A., Markham J. Brain blood flow measured with intravenous H215O. II. Implementation and validation.
J. Nucl. Med.
,
24
:
790
-798,
1983
.
30
Lammertsma A. A., Frackowiak R. S., Hoffman J. M., Huang S. C., Weinberg I. N., Dahlbom M., MacDonald N. S., Hoffman E. J., Mazziotta J. C., Heather J. D. The C15O2 build-up technique to measure regional cerebral blood flow and volume of distribution of water.
J. Cereb. Blood Flow Metab.
,
9
:
461
-470,
1989
.
31
Iida H., Kanno I., Takahashi A., Miura S., Murakami M., Takahashi K., Ono Y., Shishido F., Inugami A., Tomura N. Measurement of absolute myocardial blood flow with H215O and dynamic positron-emission tomography. Strategy for quantification in relation to the partial-volume effect.
Circulation
,
78
:
104
-115,
1988
.
32
Bergmann S. R., Herrero P., Markham J., Weinheimer C. J., Walsh M. N. Noninvasive quantitation of myocardial blood flow in human subjects with oxygen-15-labeled water and positron emission tomography.
J. Am. Coll. Cardiol.
,
14
:
639
-652,
1989
.
33
Lammertsma A. A., Cunningham V. J., Deiber M. P., Heather J. D., Bloomfield P. M., Nutt J., Frackowiak R. S., Jones T. Combination of dynamic and integral methods for generating reproducible functional CBF images.
J. Cereb. Blood Flow Metab.
,
10
:
675
-686,
1990
.
34
Araujo L. I., Lammertsma A. A., Rhodes C. G., McFalls E. O., Iida H., Rechavia E., Galassi A., De Silva R., Jones T., Maseri A. Noninvasive quantification of regional myocardial blood flow in coronary artery disease with oxygen-15-labeled carbon dioxide inhalation and positron emission tomography.
Circulation
,
83
:
875
-885,
1991
.
35
Weinberg I. N., Huang S. C., Hoffman E. J., Araujo L., Nienaber C., Grover-McKay M., Dahlbom M., Schelbert H. Validation of PET-acquired input functions for cardiac studies.
J. Nucl. Med.
,
29
:
241
-247,
1988
.
36
Iida H., Rhodes C. G., De Silva R., Araujo L. I., Bloomfield P. M., Lammertsma A. A., Jones T. Use of the left ventricular time-activity curve as a noninvasive input function in dynamic oxygen-15-water positron emission tomography.
J. Nucl. Med.
,
33
:
1669
-1677,
1992
.
37
Keyes J. W. J. SUV: standard uptake or silly useless value?.
J Nucl. Med.
,
36
:
1836
-1839,
1995
.
38
Boellaard R., van Lingen A., Lammertsma A. A. Experimental and clinical evaluation of iterative reconstruction (OSEM) in dynamic PET: quantitative characteristics and effects on kinetic modeling.
J. Nucl. Med.
,
42
:
808
-817,
2001
.
39
Phelps M. E., Huang S. C., Hoffman E. J., Selin C., Sokoloff L., Kuhl D. E. Tomographic measurement of local cerebral glucose metabolic rate in humans with (F-18)2-fluoro-2-deoxy-d-glucose: validation of method.
Ann. Neurol.
,
6
:
371
-388,
1979
.
40
Hoekstra C., Paglianiti I., Hoekstra O., Smit E. F., Postmus P. E., Teule G. J. J., Lammertsma A. A. Monitoring response to therapy in cancer using [18F]-2-fluoro-2-deoxy-d-glucose and positron emission tomography: an overview of different analytical methods.
Eur. J. Nucl. Med.
,
27
:
731
-743,
2000
.
41
Hoekstra C., Hoekstra O., Lammertsma A. A. On the use of image derived input functions in oncological FDG PET studies.
Eur. J. Nucl. Med.
,
26
:
1489
-1492,
1999
.
42
Akaike H. A new look at the statistical identification.
IEEE Trans. Automat. Contr.
,
19
:
716
-723,
1978
.
43
Schwarz G. Estimating the dimension of a model.
Ann. Stat.
,
6
:
461
-464,
1978
.
44
Lodge M. A., Carson R. E., Carrasquillo J. A., Whatley M., Libutti S. K., Bacharach S. L. Parametric images of blood flow in oncology PET studies using [15O]water.
J. Nucl. Med.
,
41
:
1784
-1792,
2000
.
45
Huang S. C., Carson R. E., Hoffman E. J., Carson J., MacDonald N., Barrio J. R., Phelps M. E. Quantitative measurement of local cerebral blood flow in humans by positron computed tomography and 15O-water.
J. Cereb. Blood Flow Metab.
,
3
:
141
-153,
1983
.
46
Alpert N. M., Eriksson L., Chang J. Y., Bergstrom M., Litton J. E., Correia J. A., Bohm C., Ackerman R. H., Taveras J. M. Strategy for the measurement of regional cerebral blood flow using short-lived tracers and emission tomography.
J. Cereb. Blood Flow Metab.
,
4
:
28
-34,
1984
.
47
Koeppe R. A., Holden J. E., Ip W. R. Performance comparison of parameter estimation techniques for the quantitation of local cerebral blood flow by dynamic positron computed tomography.
J. Cereb. Blood Flow Metab.
,
5
:
224
-234,
1985
.
48
Carson R. E., Huang S. C., Green M. V. Weighted integration method for local cerebral blood flow measurements with positron emission tomography.
J. Cereb. Blood Flow Metab.
,
6
:
245
-258,
1986
.
49
Hermansen F., Ashburner J., Spinks T. J., Kooner J. S., Camici P. G., Lammertsma A. A. Generation of myocardial factor images directly from the dynamic oxygen-15-water scan without use of an oxygen-15-carbon monoxide blood-pool scan.
J. Nucl. Med.
,
39
:
1696
-1702,
1998
.
50
Lammertsma A. A., Martin A. J., Friston K. J., Jones T. In vivo measurement of the volume of distribution of water in cerebral grey matter: effects on the calculation of regional cerebral blood flow.
J. Cereb. Blood Flow Metab.
,
12
:
291
-295,
1992
.
51
Kety S. Measurement of local blood flow by the exchange of an inert, diffusable substance.
Methods Med. Res.
,
8
:
228
-236,
1960
.