Cancer does not always respond to treatment even an agent that generally is efficacious against that type of cancerand oncologists have long sought a method of determining what is happening to permit early discontinuation of ineffective measures. Can imaging help? In some circumstances, the answer clearly is yes. This article looks at a few of the many techniques radiologists are employing in the monitoring of cancer treatment.

IMAGING ANATOMY

The traditional method of assessing cancer response is measuring changes in lesion size, but unless the target lesions disappear, there are numerous difficulties.  First, a "lesion" seen on an image may not be all cancer, so a change in size does not necessarily mean regression or progression of disease. In a series of pulmonary neoplasms, a reactive zone ranging from 2% to 48% of the lesion diameter was documented.¹ Second, a tumor can respond without an obvious change in its size if cancer is replaced by fibrosis or necrosis. Third, measurement of change is not always reliable. This problem came to public attention during Congressional hearings this summer. Raymond B. Weiss, MD, long-time chair of the Data Audit Committee of Cancer and Leukemia Group B, described to a House subcommittee the conclusions of two groups of reviewers who assessed the response to cetuximab, the growth factor receptor antagonist being developed by ImClone Systems for colorectal cancer. Whereas the criteria for treatment success had been satisfied according to one set of interpretations, Weiss pointed out that the other group questioned whether the study endpoint had been met. "A comparison of these two sets of evaluations indicates the subjectivity that can occur in making assessments of the same CT scans," Weiss commented.²

Figure 1. Non-small cell lung cancer in a 68-year-old woman manifesting as a spiculated mass. CT (lung and mediastinal windows) shows a poorly marginated mass in the left lower lobe (arrows). Because of the irregular margin, measurement differences between different readers (inter-observer variability) were large with 109% variability in maximum diameter and 245% in product size. Errors in measurement falsely indicate progression of malignancy using RECIST and WHO criteria, respectively. H=hiatal hernia, E=pleural effusion.

Further evidence of the error potential was obtained in an as yet-unpublished study at The University of TexasMD Anderson Cancer Center, in which 41 pulmonary nodules were evaluated on sequential scans by five readers. Jeremy J. Erasmus, Jr, MD, associate professor of diagnostic radiology at the center, characterized the findings.

"There was enough variability between readers looking at the same nodule that the results could have been classified as a partial response by one and progressive disease by another. By the World Health Organization criteria, a 50% change in lesion diameter is required to say there has been a response, and by the RECIST (Response Evaluation Criteria in Solid Tumors) criteria, the requirement is 30%. Those levels are easily reached with interobserver variability." The research team also recommended greater standardization of the RECIST criteria to ensure that imaging studies will be obtained and reviewed in the same way.

Although Erasmus and others are convinced that there often is no objective way to determine the response of a lesion accurately by anatomic measurements alone, several clinical trials at MD Anderson require such measurements. To minimize variability, "we usually have one person assigned to do all of the measurements for each protocol," Erasmus reports. "So at 6-week intervals, the same person will measure the same target lesion(s)."

Viability after treatment is a metabolic, not a geometric, characteristic of a lesion. As imaging becomes more able to depict biochemical and molecular biology features, the options for cancer diagnosis and monitoring are expanding rapidly, particularly as knowledge accumulates about the ways in which malignant and nonmalignant tissue differ chemically.

IMAGING METABOLISM

The first great success in metabolic imaging can be traced to the discovery by the great biological chemist Otto Warburg in the 1920s that cancer cells generate most of their energy by glycolytic metabolism of glucose, whereas benign cells employ aerobic metabolism. Glycolysis is inefficient: cancer cells need to increase their metabolic rate dramatically to obtain enough energy to support themselves and their rapid replication.

Like glucose, fluorodeoxyglucose (FDG) is taken up in abundance by most cancer cells and phosphorylated to FDG-6-phosphate. Unlike glucose-6-phosphate, however, FDG-6-phosphate cannot be processed further by glycolysis, nor can cancer cells dephosphorylate it quickly, so it is trapped in the cells for detection by positron emission tomography (PET).

The utility of FDG PET in detecting and staging cancer is well established,³ and increasingly, it is being used to monitor treatment. In 22 patients with advanced breast cancer, FDG uptake by drug-responsive tumors had fallen below 55% of the baseline value as early as the first course of chemotherapy. The eventual histopathologic response of a given cancer could be predicted with an accuracy of 88% after the first course of drugs and 91% after the second course.⁴ In another series of 30 women with large or locally advanced breast cancers, the mean pretreatment dose uptake ratio (DUR) was significantly higher in the tumors that responded completely to chemotherapy, and there was a correlation between the amount of decline in the DUR and the extent of the tumor response. In this series, PET scans after the first course were able to predict the eventual response with a sensitivity of 90% and a specificity of 74%.⁵ Also, the FDG influx constant (K) was significantly higher before any treatment in those lymph node metastases destined to respond, and the changes in the DUR and K were significantly greater after the first course than in those lesions that did not respond. The value of FDG PET in monitoring chemohormonal therapy for large or locally advanced breast cancers has also been demonstrated.⁶ These authors noted that among eight patients having partial or complete responses (all of which were detectable by PET within a week), the sizes of the tumors had not changed radiographically by day 63, when surgery was performed.

Similar value for PET has been demonstrated in other cancers. For example, Lowe and associates of St Louis University found FDG PET useful for identifying persistent head and neck cancer after chemotherapy, the sensitivity and specificity in this application being 90% and 83%, respectively.⁷ In two patients, follow-up biopsies were negative, but PET led to second biopsies that confirmed persistent disease. These same investigators found PET valuable for post-treatment surveillance: PET was the only study that detected all recurrences, and in almost a third of these patients, recurrence was detected only by PET.⁸ Farber and colleagues of the Hospital of the University of Pennsylvania used FDG PET to identify persistent head and neck cancer after radiotherapy,⁹ and Sakamoto and associates of Osaka City University Medical School noted that FDG PET was more accurate in identifying responses of this cancer to radiotherapy with or without chemotherapy than were CT and MRI.¹⁰ Similar utility has been described in cancers of the lung,¹¹ esophagus,¹² and colon,¹³ as well as Hodgkin's disease^14,15 and non-Hodgkin's lymphoma^15,16 among other cancers. In evaluating patients with lymphoma, FDG PET may also have prognostic value in that parameters such as the tumor:normal-tissue contrast ratio correlate with the proliferative index.¹⁷ In prostate cancer, in which FDG-PET generally is not useful because of the cancer's location and slow metabolic rate, the study is nevertheless able to distinguish active bony metastases from other types of bone lesions.¹⁸

Although glucose has been the primary contrast agent for PET, it is far from the only metabolite that can be imaged in this way. Oxygen metabolism, blood flow, and amino acid, protein, and nucleic acid metabolism have all been depicted.¹⁹ Physicians at Wayne State University used FDG in combination with ¹¹C-thymidine to monitor both tumor metabolism and DNA synthesis (ie, cell division) in small-cell lung cancer and high-grade sarcoma. In all patients whose disease responded, uptake of both FDG and thymidine declined significantly within a week, whereas there was only a slight change in thymidine uptake in the patients whose disease progressed.²⁰

Certain cancers, such as those of the brain and prostate, have high choline content, a feature that is being exploited in various ways. At Duke University Medical Center, ¹⁸F-choline has been synthesized. According to R. Edward Coleman, MD, division head of nuclear medicine and vice chair of radiology, prostate cancer accumulates far more fluorocholine than FDG. The team has submitted an Investigational New Drug application.

"One of the questions is whether we can diagnose or rule out prostate cancer in a man with an elevated serum prostate-specific antigen concentration. An equally important question is whether the cancer has spread to the lymph nodes and elsewhere. In our pilot study, we were able to see nodal or distant metastases much more clearly with fluorocholine than we could with FDG. With the combined PET/CT scanners, it should also be possible to determine whether the primary tumor has penetrated the capsule, which would alter the treatment." Among these patients, lesions responding to androgen deprivation had striking falls in fluorocholine uptake.²¹

Choline is also a target for magnetic resonance spectroscopy (MRS).

"Elevation of the choline levels is a relatively consistent indicator of the presence of tumor," says Suresh K. Mukherji, MD, chief of neuroradiology and head and neck radiology at the University of Michigan. "It is not 100% reliable as other etiologies can cause increased choline levels such as acute radiation necrosis and active multiple sclerosis plaques. There have been reports of changes in choline in head and neck cancers within 3 to 4 weeks after the start of treatment including work at our institution. However, these findings need to be examined systematically. We were recently funded to study the effects of radiation treatment on the amount of choline in human tumor xenografts using MRS to confirm that changes reflect tumor response and to determine the timeline."

IMAGING VASCULARITY

As Judah Folkman demonstrated 30 years ago, cancers cannot thrive unless they induce the formation of new blood vessels to supply their metabolic needs. Angiography that identifies these generally abnormal vessels has been an important part of cancer diagnosis for many years. Increasingly, reductions in tumor vascularity are being used to monitor treatment.

Uptake of thallium 201, as measured by single photon emission computed tomography (SPECT), reflects tissue perfusion and may also indicate the rate of cell proliferation.^{22, 23} Numerous articles attest to the utility of SPECT in treatment monitoring. For example, in nine patients with head and neck cancers, there was no thallium uptake on post-treatment scans in the four whose disease proved to have been controlled, whereas abnormal uptake persisted in the five patients in whom treatment failed.²⁴ Also, in 16 patients with highly aggressive astrocytomas, thallium SPECT was more reliable than CT in identifying tumor responsiveness to chemotherapy.²⁵ The utility of thallium scanning in bone cancer also has been described.²⁶

Other imaging modalities likewise depict changes in perfusion resulting from treatment. A team at the Regina Elena Cancer Institute in Rome, examined 30 women with locally advanced breast cancer using ⁹⁹mTc-sestamibi washout measurements during neoadjuvant chemotherapy. The study predicted drug resistance with a sensitivity of 100%, a specificity of 80%, and positive and negative predictive values of 83% and 100%, respectively.²⁷

Magnetic resonance imaging has long been used to study tissue diffusion and perfusion. Radiation therapists at the Mount Vernon Centre for Cancer Treatment in Northwood, UK, found that the response of advanced head and neck tumors to radiotherapy could be predicted by dynamic contrast-enhanced MRI, with a maximum enhancement of less than 8 and a mean decline in the time to maximum enhancement of 27.3 seconds soon after completion of radiation being predictive of enduring local control.²⁸ Reasoning that because the apparent diffusion coefficient (ADC) of a cancer can indicate the amount of necrosis, a multicenter study in England measured the ADC of locally advanced rectal cancers before and after chemotherapy or chemoradiation. There was a strong negative correlation between the mean pretreatment ADC and the therapy response.²⁹ Mukherji and his colleagues at the University of Michigan are in the process of prospectively evaluating the ability of serial MR diffusion and perfusion performed during treatment to predict early response in patients with intracranial astrocytomas.

MRI perfusion studies employing naturally occurring deoxyhemoglobin as a contrast agent (blood oxygen level-dependent or BOLD) have been used primarily for functional studies of the brain, but the technique may also be applicable to tumor imaging.³⁰ However, considerable work remains to be done to refine the imaging parameters and test the clinical utility. Employment of 3 T scanners may improve the contrast available from BOLD studies and increase their utility for surgical planning and treatment monitoring.³¹

Further developments in MRI for the vascular assessment of tumors are likely in the near future. Ultrasmall superparamagnetic iron oxide particles,³² rapid-clearance blood-pool contrast agents,³³ and macromolecular contrast media for characterizing changes in the microvasculature during treatment³⁴ are under investigation.

THE FUTURE

The US National Cancer Institute (NCI) has made imaging for treatment monitoring a scientific priority. The agency already has funded the establishment of multidisciplinary in vivo cellular and molecular imaging centers at Massachusetts General Hospital, Memorial Sloan-Kettering Cancer Center, University of California, Los Angeles, Washington University, and, most recently, the University of Michigan and has provided planning grants intended to lead to 16 more such centers. The NCI also has started the Development of Clinical Imaging Drugs and Enhancers (DCIDE) to help with preclinical development of promising agents that do not have strong corporate sponsors because of a perceived absence of a large market. Among the potential new agents are gamma-emitting metallopharmaceuticals recognized as substrates by the multidrug resistance P-glycoprotein, radioligands for sigma-2 receptors that may measure cell proliferation, and biochemically activated contrast agents for MRI.³⁵ Clearly, the importance of imaging in the treatment of cancer is only beginning.

Judith Gunn Bronson, MS, is a contributing writer for Decisions in Imaging Economics.

References:

1. Layfield LJ, Liu K, Erasmus JJ. Radiologically determined diameter, pathologic diameter, and reactive zone surrounding pulmonary neoplasms: implications for transthoracic fine-needle aspiration of pulmonary neoplasms. Diagn Cytopathol. 1999;21:250252.
2. Weiss RB. Review of the clinical trial undertaken by ImClone Systems to assess the efficacy and toxicity of cetuximab (Erbitux). Report to the House Subcommittee on Oversight and Investigations. Washington, DC, June 13, 2002. Available at http://energycommerce.house.gov/107/hearings/06132002Hearing587/Papineau1014.htm.
3. Czernin J, Phelps ME. Positron emission tomography scanning: current and future applications. Annu Rev Med. 2002;53:89112.
4. Schelling M, Avril N, Nahrig J, et al. Positron emission tomography using [18(F)] fluorodeoxyglucose for monitoring primary chemotherapy in breast cancer. J Clin Oncol. 2000;18:16891695.
5. Smith IC, Welch AE, Hutcheon AW, et al. Positron emission tomography using [(18)F)-fluorodeoxy-D-glucose to predict the pathologic response of breast cancer to primary chemotherapy. J Clin Oncol. 2000;18:16761688.
6. Wahl RL, Zasadny K, Helvie M, Hutchins GD, Weber B, Cody R. Metabolic monitoring of breast cancer chemohormonotherapy using positron emission tomography: initial evaluation. J Clin Oncol. 1993;11:21012111.
7. Lowe VJ, Dunphy FR, Varvares M, et al. Evaluation of chemotherapy response in patients with advanced head and neck cancer using [F-18] fluorodeoxyglucose positron emission tomography. Head Neck. 1997;19:666674.
8. Lowe VJ, Boyd JH, Dunphy FR, et al. Surveillance for recurrent head and neck cancer using positron emission tomography. J Clin Oncol. 2000;18:651658.
9. Farber LA, Benard F, Machtay M, et al. Detection of recurrent head and neck squamous cell carcinomas after radiation therapy with 2-18F-fluoro-2-deoxy-D-glucose positron emission tomography. Laryngoscope. 1999;109:970975.
10. Sakamoto H, Nakai Y, Ohashi Y, et al. Monitoring of response to radiotherapy with fluorine-18 deoxyglucose PET of head and neck squamous cell carcinomas. Acta Otolaryngol Suppl. 1998;538:254260.
11. Thomas DM, Mitchell PLR, Berlangieri SU, et al. The role of positron emission tomography in assessing response to neoadjuvant chemotherapy for non-small cell lung cancer [letter] [erratum at Med J Austral 169:344][see comment]. Med J Aust. 1998;169:227.
12. Brucher BL, Weber W, Bauer M, et al. Neoadjuvant therapy of esophageal squamous cell carcinoma: response evaluation by positron emission tomography. Ann Surg. 2001;233:300309.
13. Yoshioka T, Fukuda H, Fujiwara T, et al. FDG PET evaluation of residual masses and regrowth of abdominal lymph node metastases from colon cancer compared with CT during chemotherapy. Clin Nucl Med. 1999;24:261263.
14. Hueltenschmidt B, Sautter-Bihl ML, Lang O, et al. Whole body positron emission tomography in the treatment of Hodgkin disease. Cancer. 2001;91:302310.
15. Jerusalem G, Beguin Y, Fassotte MF, et al. Whole-body positron emission tomography using 18F-fluorodeoxyglucose for posttreatment evaluation in Hodgkin's disease and non-Hodgkin's lymphoma has higher diagnostic and prognostic value than classical computed tomography scan imaging. Blood. 1999;94:429433.
16. Okada J, Yoshikawa K, Imazeki K, et al. The use of FDG-PET in the detection and management of malignant lymphoma: correlation of uptake with prognosis. J Nucl Med. 1991;32:686691.
17. Okada J, Yoshikawa K, Itami M, et al. Positron emission tomography using fluorine-18-fluorodeoxyglucose in malignant lymphoma: a comparison with proliferative activity. J Nucl Med. 1992;33:325329.
18. Morris MJ, Akhurst T, Osman I, et al. Fluorinated deoxyglucose positron emission tomography imaging in progressive metastatic prostate cancer. Urology. 2002;59:913918.
19. Scott AM. Current status of positron emission tomography in oncology. Australasian Radiol. 2002;46:154162.
20. Shields AF, Mankoff DA, Link JM, et al. Carbon-11-thymidine and FDG to measure therapy response. J Nucl Med. 1998;39:17571762.
21. DeGrado TR, Coleman RE, Wang S, et al. Synthesis and evaluation of 18F-labeled choline as an oncologic tracer for positron emission tomography: initial findings in prostate cancer. Cancer Res. 2001;61:110117.
22. Ishibashi M, Fujii T, Yamana H, et al. Relationship between cancer cell proliferation and thallium-201 uptake in lung cancer. Ann Nucl Med. 2000;14:255261.
23. Gungor F, Bezircioglu H, Guvenc G, et al. Correlation of thallium-201 uptake with proliferating cell nuclear antigen in brain tumors. Nucl Med Commun. 2000;21:803810.
24. Mukerji SK, Gapany M, Neelon B, McCartney W. Evaluation of 201Tl SPECT for predicting early treatment response in patients with squamous cell carcinoma of the extracranial head and neck treated with nonsurgical organ preservation therapy: initial results. J Comput Assist Tomogr. 2000;24:146151.
25. Kallen K, Geijer B, Malmstrom P, et al. Quantitative 201Tl SPET imaging in the follow-up of treatment for brain tumour: a sensitive tool for the early identification of response to chemotherapy. Nucl Med Commun. 2000;21:259267.
26. Sumiya H, Taki J, Higuchi T, Tonami N. Nuclear imaging of bone tumors: thallium-201 scintigraphy. Semin Musculoskel Radiol. 2001;5:177182.
27. Sciuto R, Pasqualoni R, Bergomi S, et al. Prognostic value of 99mTc-sestamibi washout in predicting response of locally advanced breast cancer to neoadjuvant chemotherapy. J Nucl Med. 2002;43:745751.
28. Hoskin PJ, Saunders MI, Goodchild K, Powell ME, Taylor NJ, Baddeley H. Dynamic contrast enhanced magnetic resonance scanning as a predictor of response to accelerated radiotherapy for advanced head and neck cancer. Br J Radiol. 1999;72:10931098.
29. Dzik-Jurasz A, Domenig C, George M, et al. Diffusion MRI for prediction of response of rectal cancer to chemoradiation. Lancet. 2002;360:307308.
30. Howe FA, Robinson SP, McIntyre DJ, Stubbs M, Griffiths JR. Issues in flow and oxygenation dependent contrast (FLOOD) imaging of tumours. NMR Biomed. 2001;14:497506.
31. Thulborn KR. Clinical rationale for very-high-field (3.0 Tesla) functional magnetic resonance imaging. Top Magn Reson Imaging. 1999;10:3750.
32. Turetschek K, Roberts TP, Floyd E, et al. Tumor microvascular characterization using ultrasmall superparamagnetic iron oxide particles (USPIO) in an experimental breast cancer model. J Magn Reson Imaging. 2001;13:882888.
33. Turetschek K, Floyd E, Shames DM, et al. Assessment of a rapid clearance blood pool MR contrast medium (P792) for assays of microvascular characteristics in histopathology. Magn Reson Med. 2001;45:880886.
34. Brasch R, Turetschek K. MRI characterization of tumors and grading angiogenesis using macromolecular contrast media: status report. Eur J Radiol. 2000;34:148155.
35. Luker GD. Special Conference of the American Association for Cancer Research on Molecular Imaging in Cancer: Linking biology, function, and clinical applications in vivo. Cancer Res. 2002;62:21952198.