| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Massachusetts College of Pharmacy and Health Sciences, 4 Brook Rd., Unit 11,
Salem, NH 03079.
2 Dana-Farber Cancer Institute, Boston, MA.
3 University of California Davis School of Medicine, Sacramento, CA.
4 Brigham and Women's Hospital, Boston, MA.
Received April 30, 2007;
accepted after revision June 19, 2007.
Address correspondence to C. H. Holdsworth
(clayholdsworth{at}yahoo.com).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Sixty-three patients enrolled in a multicenter trial evaluating imatinib mesylate therapy for advanced GIST underwent FDG PET at baseline and 1 month after initiation of treatment. Of these 63 patients, 58 underwent concomitant CT. Time-to-treatment failure (TTF) was used as the outcome measure. Patients were followed up over a range of 23.7 to 37 months (median, 31.7 months). The predictive power of change in CT bidimensional measurements, change in PET SUVmax, and PET SUVmax at 1 month after initiation of treatment were determined, optimized, and compared. The effectiveness of combining metrics was also evaluated.
RESULTS. Both a threshold PET SUVmax value of 2.5 at 1 month (p = 0.04) and the European Organization for Research and Treatment of Cancer (EORTC) criteria for partial response on FDG PET (25% reduction in PET SUVmax) at 1 month (p = 0.004) were predictive of prolonged treatment success. The Southwest Oncology Group (SWOG) criteria for partial response (3 50% reduction in CT bidimensional measurements) at 1 month were not predictive (p = 0.55) of TTF. Optimizing metrics improved results performance. An optimized PET SUVmax threshold of 3.4 (p = 0.00002), a reduction in the SUVmax of 40% (p = 0.002), and an optimized CT bidimensional measurement threshold—that is, no growth from baseline to 1 month (p = 0.00005)—outperformed the existing standards (i.e., EORTC and SWOG criteria). Combinations of metrics did not improve performance.
CONCLUSION. The two best metrics were the optimized PET SUVmax threshold of 3.4 at 1 month (p = 0.00002) and the optimized CT bidimensional measurement threshold (no growth from baseline to 1 month, p = 0.00005) in this patient group.
Keywords: CT • gastrointestinal stromal tumor • imatinib mesylate • oncologic imaging • PET
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The difference in metabolic activity after even a single dose of imatinib mesylate can be shown on 18F-FDG PET scans and is often dramatic [8, 9]. Little or no change can be seen on corresponding CT images initially [10], and over time, lesions may become more noticeable due to lower attenuation [11, 12]. Fewer than 15% of patients exhibit primary resistance to imatinib mesylate or rapidly fail to respond to therapy [6-8]. Patients who initially respond to imatinib mesylate may, over an extended period, develop secondary resistance to imatinib mesylate and the disease subsequently progresses; however, their prognosis remains substantially better than it would have been without imatinib mesylate therapy.
Objective response to therapy in oncology has commonly been defined using Response Evaluation Criteria in Solid Tumors (RECIST) [13] for unidimensional tumor measurements using CT or Southwest Oncology Group (SWOG) criteria [14], which uses CT bidimensional measurements. RECIST requires a 30% reduction in unidimensional measurements to meet the condition for partial response. This definition of partial response corresponds directly to the SWOG criteria that specify a 3 50% reduction in CT bidimensional measurements to satisfy the same condition.
Criteria for therapeutic response assessment have also been developed in the context of functional imaging. The European Organization for Research and Treatment of Cancer (EORTC) has defined guidelines for the use of PET using FDG. These guidelines state that a 25% reduction in the maximum standardized uptake value (SUVmax) should be considered as the threshold for definition of partial response [15, 16].
All of these criteria were originally defined for PET and CT evaluations of prognosis in patients receiving cytotoxic drug therapy or undergoing radiation therapy. Whether the use of such criteria after the initiation of therapy using molecularly targeted kinase inhibitors is generally predictive of outcome is not clear. The EORTC guidelines for partial response have been shown to correlate with outcome in patients with advanced GIST when SUV is measured 2 months after initiation of imatinib mesylate therapy [10]. Similar results were found when PET scans were obtained between 3 and 16 weeks after therapy start [17].
In the literature, a cut point of 2.5 in the SUVmax for FDG PET has also been used to assess response to imatinib mesylate treatment of GIST and has been shown to correlate with outcome when measured 1 month after initiation of imatinib mesylate therapy [9, 18-20]. In contrast, standard CT criteria, such as the RECIST and SWOG criteria, have not been found to be predictive of outcome [21, 22]; however, thresholds showing little or no reduction in CT bidimensional measurements at early time points after initiation of treatment were found to correlate well with outcome [10, 18].
The purpose of study was to evaluate and optimize the use of CT and FDG PET at baseline and 1 month after therapy initiation as predictors of time-to-treatment failure (TTF) in the context of imatinib mesylate treatment of patients with advanced GIST. TTF is defined as the time from the first dose of imatinib mesylate to the earliest occurrence of disease progression, death, or discontinuation from the trial for any medical reason.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This study was approved by the institutional review board and performed in an ambulatory setting in two tertiary care oncology academic centers (Dana-Farber Cancer Institute and Brigham and Women's Hospital). Written informed consent was obtained from all patients. The cohort for this study was a subset of the patients enrolled in a wider multicenter trial supported by Novartis Oncology.
Imaging
All 63 patients underwent FDG PET before treatment and 21-40 days after
initiation of therapy. Fifty-eight of these patients also underwent CT at
baseline and 21-40 days after initiation of therapy. All FDG PET data were
acquired on a tomograph (ECAT Exact HR+, Siemens/CTI). The mean dose given to
patients was 20 mCi (740 MBq) (range, 7-22 mCi [259-814 MBq]) for the baseline
scans and 20 mCi (740 MBq) (range, 16-27 mCi [592-999 MBq]) for the 1-month
scans. The mean time between FDG injection and the start of scanning was 51
minutes (range, 27-82 minutes) for baseline scans and 52 minutes (range, 27-87
minutes) for the 1-month scans.
The PET SUVmax value was calculated for the lesion with the most intense uptake at baseline and was subsequently calculated for the same lesion on follow-up scans. No corrections for partial volume effects, lean body mass, or blood glucose levels were applied. Because of the substantial changes in uptake, lesions were not always visible on follow-up images, and in these cases, guidance for location of the lesion for PET SUVmax calculations was obtained from the concomitant CT scan. All scans were obtained after the patient had fasted for 6 hours.
CT data were acquired on an MDCT scanner (Somatom Volume Zoom, Siemens Medical Solutions). Patients were asked to fast for 6 hours before scanning. Oral contrast material (diatrizoate meglumine, 5.0 g, and diatrizoate sodium, 0.75 g, in a 500-mL aqueous solution) was administered approximately 150 minutes and then again 90 minutes before scanning. Immediately before scanning, 100 mL of ioxilan (300 mg/mL) iodinated contrast material was administered IV at a rate of 2 mL/s or slower in patients with limited IV access. Scanning was performed from the thoracic inlet through the pelvis, with the patient breath-holding during scanning of the chest and quietly breathing during scanning of the abdomen and pelvis.
The nominal scanning parameters were as follows, with adjustments as necessary for body habitus: beam current, 165 mAs; beam energy, 120 kVp; beam collimation, 2.5 cm; and table speed, 5.4 cm/s. CT images were reconstructed at 7.0-mm intervals. Images were examined using standard clinical window and level settings, and measurable lesions, at least one lesion and up to 12 lesions per patient, were chosen by an experienced radiologist. Axial plane measurements of the maximum diameter of the lesions and their maximal perpendicular dimension were made in the standard fashion.
Outcome Measure
The outcome measure was TTF. It was defined as the time from the first dose
of imatinib mesylate to the earliest occurrence of disease progression, death,
or discontinuation from the trial for any medical reason. Patients for whom
the treatment had not failed according to this definition were censored at the
date of their last disease assessment. Four patients had to stop taking
imatinib mesylate because of medical complications independent of the
effectiveness of this drug on their disease. Because predicting these side
effects using imaging is not possible, these patients skew our results, but
only very slightly due to the small number of patients involved. Patients were
followed for up to 37 months after treatment initiation.
Prognostic Metrics
The following specific metrics about the single largest lesion for each
image were examined: first, percent change in CT bidimensional measurements
from baseline to 1 month after initiation of treatment; second, PET
SUVmax at 1 month after initiation of treatment; and, third,
percent change in PET SUVmax from baseline to 1 month after
treatment initiation. Thresholds that divided the patient population into
separate poor- and good-prognosis groups were used. These thresholds were
varied over the entire range of measured values to find the threshold that
yielded optimum performance for each metric.
The optimized thresholds were then compared with the following standards:
first, 
50% reduction in CT bidimensional measurements from baseline to 1
month after initiation of treatment (SWOG criteria for partial response);
second, PET SUVmax of less than 2.5 at 1 month after initiation of
treatment (physiologic FDG uptake in tissue); and, third, 25% reduction in PET
SUVmax from baseline to 1 month after initiation of treatment
(EORTC guidelines for partial response). The performance of combinations of
metrics (i.e., PET SUVmax and CT bidimensional measurements) was
also investigated.
Statistical Analysis
The goal of this effort was to find an early prognostic imaging indicator
of treatment efficacy (i.e., TTF). To evaluate existing thresholds, including
SWOG and EORTC cutoffs for partial response, the p value from the
log-rank test was used. To identify thresholds for PET SUVmax and
CT bidimensional measurements at 1 month, recursive partitioning was used
[23]. In this method, the
maximally selected chi-square statistic is used to identify an optimal cut
point.
|
|
|
|
|
|
Imaging data were also analyzed by receiver operating characteristic (ROC) techniques [25]. Two binary classification tasks were defined by specifying survival thresholds of 180 and 365 days. Fifty-four patients with all imaging variables were included. Imaging metrics were used as input to a publicly available continuous ROC analysis program (ROCKIT, University of Chicago) [24, 25]. Classification performance was quantified using the area under the ROC curve.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
Kaplan-Meier plots for percent reduction in PET SUVmax between baseline and 21-40 days are shown in Figure 4A. EORTC guidelines (i.e., 25% reduction in PET SUVmax) were predictive of outcome (p = 0.004). The performance of the percent reduction in PET SUVmax test was improved using the optimal threshold of 40% reduction (p =0.002); however, the improvement in performance was smaller than the improvement for other imaging measures.
|
|
|
A summary of performance of the evaluated metrics is shown in Table 1. Median TTFs for the predicted treatment success groups and predicted treatment failure groups for the various indicators are included.
|
ROC Analysis
ROC analysis was also performed on the data. Percent change in PET
SUVmax at 1 month after initiation of treatment, PET
SUVmax at 1 month, and percent change in CT bidimensional
measurements at 1 month were used to discriminate patients for whom treatment
would fail less than 180 days after the initiation of treatment from those who
would survive more than 180 days after the initiation of treatment.
Performance was quantified by the area under the ROC curve and found to be
0.850 ± 0.056 (mean ± standard error [SE]) for percent change in
PET SUVmax from baseline to 1 month, 0.874 ± 0.047 for PET
SUVmax at 1 month, and 0.857 ± 0.080 for percent change in
CT bidimensional measurements from baseline to 1 month.
This analysis was repeated for 1-year survival. The area under the ROC curve was 0.676 ± 0.084 for percent change in PET SUVmax from baseline to 1 month, 0.711 ± 0.075 for PET SUVmax at 1 month, and 0.770 ± 0.080 for percent change in CT bidimensional measurements from baseline to 1 month.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
No growth or reduction in CT bidimensional measurement determined at 1
month after therapy initiation was an effective indicator of prolonged
treatment success in this patient group, with no significant difference in
predictive power compared with the best PET metric (PET SUVmax 
3.4 at 1 month). No combination of metrics could outperform the threshold of
3.4 in PET SUVmax at 1 month.
These conclusions would be strengthened by validation in an independent data set; however, the plausibility of the zero growth threshold for CT bidimensional measurements is improved by the broadly similar findings from other groups [10, 26]—namely, the use of a 5% reduction threshold in CT bidimensional measurements at 2 months in a similar patient group [10] was also predictive of outcome. The difference between the thresholds found may be due to statistical sampling error or to the difference in imaging time points. However, we note that a 5% threshold at 1 month was also predictive of outcome in our data set.
Of note, the median reduction in CT bidimensional measurements of 34% at 1 month is large, suggesting that it may be possible to use CT as a prognostic metric at even earlier time points after therapy initiation. Care would be needed because there is anecdotal evidence that GIST lesions may swell as they begin to respond to imatinib mesylate, but the use of measurements of cystic changes in combination with size criteria may possibly ameliorate the risk of misclassification [12]. It is, of course, highly unlikely that CT could be used as early as FDG PET, which has shown strong responses as early as 24 hours after therapy initiation [18].
With the advent of targeted therapies, there is a need to develop prognostic metrics that can stratify a population into groups who may benefit from therapy and those who may not. Significant efforts have been undertaken to apply molecular or genomic markers in this context: The use of epidermal growth factor receptor (EGFR) overexpression, mutations, and polysomy as outcome predictors for therapy with the EGFR tyrosine kinase inhibitor erlotinib is a prime example [27, 28]. However, the specificity and practicality of such prognostic tests are not perfect, and it is reasonable to look for additional imaging biomarkers that can also be used either prospectively or shortly after therapy initiation. The results presented here are applicable only to the use of imatinib mesylate therapy in GIST, but they are encouraging and strongly suggest the need for further work on imaging biomarkers for use with other targeted therapies.
Our results show that conventional objective response criteria are not generally applicable to prognosis in therapies involving the new generation of molecularly targeted agents such as imatinib mesylate. This finding suggests that similar studies should be performed for each new therapeutic agent, at least until it can be determined whether the results found here are generally applicable to this class of therapeutic agents.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. B. Contractor and E. O. Aboagye Monitoring Predominantly Cytostatic Treatment Response with 18F-FDG PET J. Nucl. Med., May 1, 2009; 50(Suppl_1): 97S - 105S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Stacchiotti, P. Collini, A. Messina, C. Morosi, M. Barisella, R. Bertulli, C. Piovesan, P. Dileo, V. Torri, A. Gronchi, et al. High-Grade Soft-Tissue Sarcomas: Tumor Response Assessment--Pilot Study to Assess the Correlation between Radiologic and Pathologic Response by Using RECIST and Choi Criteria Radiology, May 1, 2009; 251(2): 447 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K. Shankar, A. Van den Abbeele, J. Yap, R. Benjamin, S. Scheutze, and T.J. FitzGerald Considerations for the Use of Imaging Tools for Phase II Treatment Trials in Oncology Clin. Cancer Res., March 15, 2009; 15(6): 1891 - 1897. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A. Adjei, M. Christian, and P. Ivy Novel Designs and End Points for Phase II Clinical Trials Clin. Cancer Res., March 15, 2009; 15(6): 1866 - 1872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. O. Prior, M. Montemurro, M.-V. Orcurto, O. Michielin, F. Luthi, J. Benhattar, L. Guillou, V. Elsig, R. Stupp, A. B. Delaloye, et al. Early Prediction of Response to Sunitinib After Imatinib Failure by 18F-Fluorodeoxyglucose Positron Emission Tomography in Patients With Gastrointestinal Stromal Tumor J. Clin. Oncol., January 20, 2009; 27(3): 439 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Spratlin, N. J. Serkova, and S. G. Eckhardt Clinical Applications of Metabolomics in Oncology: A Review Clin. Cancer Res., January 15, 2009; 15(2): 431 - 440. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
