PET & PET/CT Earn Role in Oncologic Therapeutic Response

Slowly but steadily, radiation oncologists are adopting PET and PET/CT to measure the early response of cancers to radiotherapy and other treatments. And progress has been significant.

For nearly 30 years of oncological medicine, patients and their physicians waited months to determine if prescribed chemo- or radiation therapies were successful. Assessments were made slowly because they relied on changes to tumor size that appeared long after treatment ends.   

Now, clinicians are beginning to apply functional PET/CT assessments as well, based on the realization that the benefits of findings by noninvasive molecular imaging performed during or soon after therapy outweigh the risks.

FDG-PET/CT plays an essential role in practice guidelines for the early response of Hodgkin’s lymphoma to chemotherapy from the National Comprehensive Cancer Network (NCCN), a nonprofit global consortium of 21 cancer research centers. The NCCN supports similar guidelines that rely on FDG-PET/CT for gauging early response of diffuse large B cell lymphoma and Non-Hodgkin’s lymphoma to therapies. Progress on the development of early response protocols using PET/CT for solid tumors has begun for cancers, such as non-small cell lung, esophageal and head and neck.

Positive early findings certify the appropriateness of treatment and assure patients about a probable cure. Negative results can lead the clinician to modify the type, timing and intensity of treatment to spare the patient unnecessary pain and morbidity, while possibly steering therapy toward a more effective course.

The acceptance of PET among oncologists has grown rapidly since Medicare authorized reimbursement for the staging and restaging of a few selected cancers with FDG and a dedicated PET camera 10 years ago. It is now generally accepted as an essential instrument for nodal and metastatic staging of most cancers. PET differentiates between locoregional cancers that may respond to various therapies and systemic disease that offers few curative options.

The capability of FDG-PET to identify location and extent of metastatic disease establishes a diagnostic foundation for treatment planning, says Hani L. Ashamalla, MD, chairman of radiation oncology at New York Methodist Hospital in New York City. Its ability to detect metastases missed by CT or MR spares patients from unnecessary radiation therapy (RT), or it can lead to life-preserving therapy of primary cancers that other modalities misclassified.

The introduction of conformal RT and intensity-modulated RT greatly improved the radiation oncologists’ ability to control the shape and intensity of the x-ray beam to focus on a specific volume of tissue.

Ashamalla et al demonstrated the 25 percent interobserver variability still encountered by using anatomic CT to plan RT could be cut to less than 10 percent by adding FDG-PET to the process.

Two studies led by Jeffrey D. Bradley, MD, chief of the thoracic service at Washington University in St. Louis, has strengthened the scientific rationale favoring the use of FDG-PET during RT planning.

One 2004 single-center prospective study involving 26 patients with confirmed non-small cell lung cancer (NSCLC) found that FDG-PET altered the TNM (tumor lymph nodes metastasis) stage in 31 percent of patients, led to changes in the RT volume for 58 percent and differentiated between tumor and atelectasis in all three patients with partially collapsed lungs (Int J Radiat Oncol Biol Phys 2004;59(1): 78-86).

Radiation Therapy Oncology Group (RTOG) 0515, a phase II prospective trial involving 52 patients with Stage II-III NSCLC demonstrated the clinical advantages of FDG PET/CT over CT alone for defining the gross tumor volume (GTV) treated by radiation therapy. The results indicated tumor volumes defined by PET/CT were significantly smaller (86.2 mL vs. 98.7 mL) than those derived with CT. PET/CT helped identify nodal lymph node targets for inclusion within the radiation field that would have been excluded by CT alone. Nodal contours were altered in 51 percent of the cases, and FDG-PET/CT delineated tumor from atelectasis in three cases (Int J Radiat Oncol Biol Phys 2010; online).

“We had only one patient who failed in lymph node target outside the PET/CT defined volume, so we feel fairly comfortable going forward with using PET/CT defined tumor in other trials,” Bradley says.

Therapy monitoring challenges

Though the value of PET/CT in RT planning now appears clear, its role for assessing the response of cancers to chemo, chemoradiation, or  RT alone is not well-established. Challenges posed by the inflammatory response of tissues to RT and the arduous task of developing specific protocols for various types of cancers stand in the way of general clinical acceptance.

Radiation-induced inflammation poses a formidable barrier to using FDG to measure early tumor response to RT. Many clinicians will not perform FDG-PET during or soon after treatment for fear of confusing regions of avid FDG uptake from inflammation with hyperintensities thought to arise from metabolically active tumor.

Unambiguous FDG-PET/CT can be performed after post-RT inflammation subsidies, but the delay would defeat the purpose of early assessment, according to Alan D. Waxman, MD, chief of nuclear medicine at Cedars Sinai Medical Center in Los Angeles.

“If you wait for three, four or six months, you are not always doing the patient a great deal of service. We need answers earlier,” he says.

Some studies suggest that a precise quantitative analysis of FDG uptake using standard uptake values (SUVmax) can enable the clinician to focus on imaging data relevant to therapeutic response during and soon after RT despite inflammatory response.

A study of the maximal FDG uptake among 23 patients treated with RT for advanced NSCLC illustrated the confounding effect of induced inflammation. Baardwijk and colleagues uncovered striking intra-individual heterogeneity in the evolution of SUVmax during therapy (Rad Onc 2007;82:145-152). Distinct trends for responders and non-responders also were identified. SUVmax increased 48 percent for non-responders during the first week of therapy. It then fell 15 percent during the second week. Higher SUVmax was measured for non-responders than responders at all times points. No significant changes were seen for responders during RT.  

The results suggest to Roland Hustinx, MD, PhD, nuclear medicine chair at the University Hospital of Liege in Belgium, that the effects of inflammation can be resolved in weeks rather than months after RT.

“The metabolic response is actually a stability within the metabolic activity,” he says, “Patients with no significant changes in FDG uptake after seven days of irradiation will ultimately respond to treatment. The non-responders are those with an increase in FDG uptake after seven days.”

Enter FLT-PET

FLT-PET/CT could play a complementary role with F18 FDG. FLT-PET serves as a noninvasive marker for cell proliferation. As with metabolism, the process measured with FDG, cell proliferation also is affected by radiation-induced inflammation. Clinicians must take care not to confuse regions of high FLT uptake associated with inflammation in non-malignant areas and high uptake indicating an incomplete response or progressive disease after therapy. However, there may still be advantages for FLT to assess response to XRT compared with FDG.  One hypothesis is that the repair response to XRT may be energy consuming but not proliferative, suggesting FLT might provide a better measure of response.

However, the healing process, which is accompanied by inflammation, is transient and subsides over a short period of time, notes Amin J. Mirhadi, MD, a radiation oncologist at Cedars Sinai.

“The clinical utility of these markers is detecting a difference between proliferating cells that are malignant and proliferating cells that are part of the healing process. This is reflected by clinical outcomes rather than what occurs at the molecular level,” he says.

FLT proved superior to FDG for assessing responses to RT for six patients in a recent pilot study conducted by Mirhadi et al. All patients received curative RT between baseline and follow-up FDG and FLT scans performed about four weeks after the start of therapy.

FLT proved to be a better biomarker than FDG for early response. SUV and tumor-to-background measures with the investigational probe correlated with the clinical outcomes of RT for all patients, while the same measurement scheme for FDG correlated poorly with outcomes for three of six patients. A patient with the highest percentage rise in FDG uptake, suggesting a poor treatment response, actually had no biopsy confirmed residual tumor.

“You can’t rely on FDG a short time after the completion of RT,” Waxman says, “but immediately after, FLT will be outstanding.”

Many bridges to cross

The variable nature of cancer, in general, requires researchers to customize therapeutic response protocols for each one, according to Hustinx. A decrease of more than 35 percent in FDG SUV indicates a positive response for esophageal cancer to RT, for example, while no change in FDG uptake signifies a positive response for NSCLC.

“We will have to figure out what the metabolic response is for every type of cancer and treatment,” he says.

The understanding that particular classes of cancers are subdivided in complex phenotypes adds to the challenge. However, no international guidelines have been published for FDG-PET and PET/CT for assessing the response of breast cancer treatment, despite intense research interest, he notes.

Success is limited, Hustinx says, because of the heterogeneous nature of breast cancers. Various phenotypes of breast cancer, reflected in presence and absence of specific hormone receptors, respond variably to neoadjuvant chemotherapy and chemoradiation, making the development of protocols based on a metabolic measure of response with FDG-PET extremely difficult, he adds.

Researchers have produced encouraging results using dynamic contrast-enhanced MRI measures of tumor blood flow to assess the response of locally aggressive breast cancers to neoadjuvant therapy before surgery. Diffusion-weighted MRI, proton MR spectroscopy, F18 FLT-PET and F18 alpha-fluorestradiol also show promise.

Developing practice guidelines

It is the relatively homogeneous nature of lymphomas that placed them at the front of the line for response-to-therapy guidelines published by NCCN, according to Hustinx.

FDG-PET/CT is the cornerstone of patient management in NCCN’s consortium detailing practice guidelines uses, Hustinx says. The protocol for Hodgkin’s lymphoma uses FDG-PET/CT to differentiate between complete responders, partial responders and patients with progressive disease after two cycles of chemotherapy. Separate treatment regimes are recommended for each subset of patients including three additional PET/CT restaging studies following courses of chemotherapy for partial responders and patients with stable disease.

“The whole treatment algorithm shows PET is more than a promising tool. It is directing the treatment of these patients,” Hustinx says.

NCCN has published similar guidelines for diffuse large B cell lymphoma and Non-Hodgkin’s lymphoma.

The professional consensus necessary before such guidelines can be published has yet to be achieved for the treatment and monitoring of solid tumors. Cancers of the esophagus, head and neck and NCSLC top the list for consideration, he says.

The capability of FDG-PET to measure the response of esophageal carcinoma to neoadjuvant therapy was examined in a cohort study by Ott et al. Metabolic responders identified with PET performed 14 days after the start of treatment had a 70 percent survival rate after three years, compared to 35 percent for nonresponders. A metabolic response was defined as a 35 percent decreased in SUV from the baseline and intra-therapy scans (J Clin Onc 2006;24:4692-4698).

Similar findings arose in a study of 119 patients with locally advanced adenocarcinoma of the esogastric junction evaluated in the MUNICON II trial (Lancet Onc 2007;8(9):797-805), which used the same 35 percent SUV reduction to differentiate between responders and non-responders. After three years, the median event-free survival for responders was 29.7 months, compared with 14.1 months. Withholding additional neoadjuvant chemotherapy to PET non-responders did not adversely affect their outcomes.

Head and neck cancers

Preliminary research from Klaus Zöphel, MD, of the University of Dresden University of Technology in Germany, suggests a response measure by FDG-PET is possible in the first or second week after RT for advanced stage head and neck squamous cell carcinoma.  

Hustinx described the 37-patient study during SNM 2011. Among those with a greater than 50 percent decline in SUVmax in the primary tumor from scans performed before and one to two weeks after the initiation of RT, 88 percent were still living 26 months after treatment. Locoregional control was maintained in 55 percent of responders after two years. For patients with less than a 50 percent decline in SUVmax, only 38 percent survived at 26 months, and locoregional control was maintained for 40 percent.

Head and neck cancers offer a quintessential example of how PET/CT aids RT treatment planning and assessment, says Dwight E. Heron, MD, vice chairman of radiation oncology at the University of Pittsburgh Medical Center (UPMC). Primary tumors may have necrotic cores or regions of aggressive, treatment-resistant cell clusters. The distribution of such cells can be identified in the tumor volume with FDG-PET/CT during RT planning. Appropriate radiation intensities can be applied throughout the tumor volume to assure eradication during intensity-modulated RT.

The procedure has become so reliable that UPMC oncologists monitor patients with a complete response by FDG-PET/CT with serial scans to spare them from surgical dissection, Heron says. One study found the two-year, progression-free status for patients with and without complete responses was 93 percent and 48 percent, respectively (Ann Oncol 2010;21(11):2278-2283).  

Questioning assumptions

While qualitative assessments are routinely performed at UPMC, quantitation is emphasized at Washington University in St. Louis to assess tumor heterogeneities in the FDG PET/CT of cervical cancer.

Perry W. Grigsby, MD, a radiation oncologist, works on both the experimental and theoretical levels, a combination that has led to interesting findings. In a recent paper, Grisby and Brooks conclude that the current measure of metabolic heterogeneity within cervical cancer do not predict disease outcome (Rad Onc 2011; online).

They argue that cancerous tumors are intrinsically heterogeneous. The lesions feature variations in gene expression, cell proliferation, vascularization and hypoxia. In a typical PET/CT image, such variations can be seen in bright regions juxtaposed with relatively dark areas.

From a detailed analysis of PET image formation and presentation, Grigsby and Brooks conclude that the delineation of successfully and unsuccessfully treated cervical cancer patients is based exclusively on tumor volume, not from measures of tumor heterogeneity. They note that a more precise, geometric quantification of metabolic variations would lead to more effective treatment plans. Tumors having a particular shape or asymmetry may indicate the likely clinical outcome.

Such information could be used to determine the appropriateness of selected therapies and to more effectively target radiation to less responsive tumor regions, Grigsby says. As their research moves forward, Grigsby and Brooks are juggling variables relating to both tumor structure and physiology to identify more predictive combinations.