INTRODUCTION

In recent years, 18F-fluoro-deoxy-D-glucose positron emission tomography (FDG-PET) has been incorporated in the management of malignant lymphomas. FDG-PET provides valuable information to clinicians in disease staging, evaluation of response to chemotherapy, planning of radiotherapy, and finally, re-staging and follow-up after therapy.¹^–⁴ Therapeutic implications for patients with Hodgkin’s disease (HD) and non-Hodgkin’s lymphoma (NHL) emphasize the importance of initial accurate disease staging. Until recently, gallium scintigraphy played an important role in lymphomas in monitoring the response to treatment and clarifying the nature of residual masses detected on computed tomography (CT).^5,⁶ Its limited resolution failed to detect smallvolume residual disease and sometimes non-specific uptake in benign lesions. In addition, it is less sensitive in intra-abdominal disease.⁷ For these reasons, FDG-PET has largely replaced the gallium scan in many centers.^8,⁹ Other than being a valuable staging tool, it provides accurate prognostic information for both HD and NHL since a reduction of FDG uptake after initial chemotherapy is a highly predictive factor of favorable outcome, while residual uptake is indicative of unfavorable prognosis.^10,¹¹ PET helps to distinguish complete disease remission from partial response,¹²^–¹⁴ which can lead to a change in management plan.

FDG-PET can be very valuable for radiotherapy purpose, and it is currently used successfully in several centers worldwide. The most important application of PET is for radiotherapy field delineation. By accurate assessment of initially involved anatomic sites and response to chemotherapy, radiation oncologists can reduce the volume of irradiation, spare the surrounding healthy tissues, and adjust the prescribed dose. FDG-PET has been reported as a useful tool for early identification of patients who may benefit from salvage therapy.^14,¹⁵

There are several unique features in the biology and presentation of HD and NHL that must be considered when using PET scans to diagnose disease, monitor response to treatment, and plan radiation treatment fields and doses. Compared to other tumors, lymphoma can have a wide range of metabolic activity, due in part to the heterogeneity of these neoplasms, both in terms of location and grade. In addition, the large age spectrum in which lymphoma presents requires the interpreter to be able to delineate between a neoplastic process vs. an age-dependent metabolic activity, particularly in younger populations.

Rationale for Use of PET and PET-CT for Lymphoma

The Value of PET in Determining the Indication of Radiation Therapy

The PET-CT can be particularly helpful in distinguishing between residual disease and complete response after initial chemotherapy, and can help identify the favorable group of patients with either HD or NHL who can be treated with reduced intensity chemotherapy or radiotherapy or with a single modality after achieving PET negativity. In conjunction with CT and magnetic resonance imaging (MRI) criteria, a metabolic image can add substantial validity to the interpretation of a chemotherapy response (complete response [CR] vs. partial response [PR]). For instance, functional imaging may play a more critical role in the tumor response evaluation particularly in the case of nodular sclerosing Hodgkin’s lymphoma, which is characterized by the presence of a thickened lymph node capsule and is commonly associated with extensive tumor bulk. Because of the abundance of collagen, the radiographic appearance of these lesions (particularly in the mediastinum) may only slowly return to normal, even when the patient is responding to therapy.¹⁶ Regression or resolution of metabolic tumor activity by PET serves as a surrogate for tumor response and precludes the need for invasive procedures to document the disease status.

However, because of the novelty of this tool, one should cautiously correlate PET interpretation with clinical data before changing the patient’s management plans. There is an emerging body of literature on various causes of false-positive or false-negative results, which will be discussed later in this chapter. Ongoing studies, hopefully, will help determine whether or not PET might be a suitable stratification tool for those patients who may benefit from radiotherapy and those who may not.

Integration of PET-CT in the Process of Radiation Treatment for Lymphomas

During the past two decades, CT and MRI have become an integral part of threedimensional conformal radiation therapy (3D-CRT) by providing detailed anatomic information on target volumes and critical structures; however, these modalities have a significant limitation. CT-based radiation treatment planning criteria are somewhat subjective and depend on anatomic features of disease (size of lymph nodes, mass, or infiltration in the parenchyma of the organs). CT does not identify lymphoma in normal-size lymph nodes nor does it distinguish non-lymphoma nodal enlargement from involved nodal masses. Therefore, involved nodes smaller than the threshold size may be excluded from treatment fields; likewise, enlarged but uninvolved nodes, which may arise as a result of prior treatment, may be unnecessarily irradiated, thereby increasing morbidity. It is in these difficult cases that functional imaging may add substantial information to improve radiotherapy target accuracy.

Because tumor control rates are excellent with current approaches, there is an opportunity for the reduction of treatment intensity in an attempt to minimize late effects, especially in younger patients. In the era of combined chemoradiotherapy for lymphoma, there has been a substantial paradigm shift from extended field to involved field radiotherapy, but the concept of contiguous lymphatic tumor spread in HD is still applied in designing radiotherapy fields.¹⁷ Radiation oncologists usually attempt to cover lymphatic regions rather than individual involved nodes. PET can help define the target volume and tailor radiation doses according to the disease burden. The treatment volume irradiated following the delivery of systemic therapy is often modified relative to the initially defined volume based on changes in normal tissue anatomy (defined on CT and MRI) as the primary mass shrinks following chemotherapy. The incorporation of PET imaging into radiation therapy treatment planning could allow further reduction of the treatment dose and/or volume, based on the response following systemic chemotherapy. This strategy of dose reduction was undertaken by several investigators specifically for pediatric HD.^17,¹⁸

As part of radiotherapy planning, the first important goal for the radiation oncologist is to provide a careful evaluation of the patient for initially involved sites and the response to chemotherapy. This is better accomplished with collaboration from radiology colleagues. Previously, when only CT or MRI was used to assess the response to initial chemotherapy, a decrease in the size of a lymphomatous mass was considered a response to treatment. This led to the development of formal criteria based on the measurement of the cross-sectional area.^19,²⁰ Current radiotherapy design guidelines are still mainly based on size criteria as determined from CT or MRI scans, and PET serves only as a correlative tool. The evaluation of residual masses on anatomic imaging following treatment is a major challenge in a significant number of patients with lymphoma. It may take a long time before the decrease in size of the tumor mass becomes evident on anatomic imaging scans. Other processes such as fibrosis, necrosis, or inflammation may cause only a partial reduction of the mass, and this may be influenced by its location, histology, original size, and treatment.^16,^19,²⁰

New software programs allow more precise imaging acquisition and fusion, and newer PET-CT-simulator integrated programs allow PET and CT (or MRI) parallel volume contouring. In radiotherapy planning, localization of the tumor is essential and therefore it is critical to have high-accuracy PET-CT co-registration with radiotherapy planning CT. One needs to keep in mind that these techniques are heavily operator-dependent and require a certain level of training.

PET for Post-Treatment Evaluation

FDG-PET scanning can serve as an important prognostic factor for post-treatment evaluation of patients with either HD or NHL. As in upfront staging, it is essential to avoid any misinterpretation of functional images after therapy, and there may be multiple factors that may influence results.

Fusion Technique and Target Delineation

Understanding of fusion technique and target delineation begins with the knowledge of ideal scheduling of the exam, patient preparation, administration of FDG, and acquisition and processing of images. Optimally, discussion in a multidisciplinary setting can lead to appropriate positioning of the patient in a radiation treatment position in case radiation therapy is anticipated at the time of diagnosis. This, however, is often difficult to achieve in the absence of a discussion with all potentially involved clinicians prior to the initiation of chemotherapy and because it is impossible to predict all sites of the body that may be involved a priori.

In addition, patients may have received a series of PET scans that reflect the initial staging and the response to chemotherapy, all of which may ultimately be incorporated in radiation treatment planning. Indeed, the ideal timing of PET scans after the completion of chemotherapy has not been firmly established for patients with lymphoma. Some have advocated that scans be obtained 2 weeks after the completion of chemotherapy to avoid fluctuations in FDG metabolism.²¹^–²³ There is firm evidence that for radiotherapy planning purposes, the pre-chemotherapy PET scan is the most reliable and will help to avoid mis- or under-dosing of the areas at high risk of relapse.¹⁴ Accepting the variability in body positioning from the initial PET scan to the planning CT or PET-CT scan, it is still desirable to fuse these image sets together to incorporate all the information on disease involvement in radiotherapy planning. This step will help to most accurately delineate all target volumes and also assess disease volume reduction and resulting change in normal structure anatomy.

Acquisition of FDG-PET and CT Images for Radiation Planning Purposes

Prior to administration of FDG, patients are requested to abstain from physical exertion for 24 hours to limit muscle uptake and to fast for at least 4 hours. Serum glucose must be < 200 units. Patients then ingest oral contrast containing 2% to 3% barium for bowel opacification to assist in detecting potentially involved mesenteric lymphadenopathy. Typically 10 to 20 mCi (0.14–0.2 mCi/kg or 370–740 MBq) is injected intravenously. The uptake phase usually lasts from 60 to 90 minutes, during which patients are kept in a quiet location and asked to limit movement and talking, again to minimize uptake in normal tissues.

The acquisition of images begins with the CT. Flat tabletop is preferred for images that are used in radiotherapy planning. Ideally the patient is placed in the radiation treatment position at this time, assuming the site to receive future treatment can be successfully anticipated and altered positioning does not impede travel through the gantry. More commonly, though, arms are placed above the head to limit streak artifact during the PET portion of the procedure due to motion. Images are obtained with a 5-mm slice thickness with low mA of 80 and 140 kEV. The auto mA protocol is aimed at getting a CT scan that is mainly used for attenuation correction and anatomic localization. PET images are reconstructed using iterative algorithm. CT reconstruction is done with filtered back projection. Metal and barium can create streak artifacts due to filtered back projection and polychromatic nature of CT.

Upon completion of the CT scan, the tabletop is then advanced into the PET scanner. Typically, a full set of CT images can be acquired in 1 minute, while a full body PET scan can take up to 15 to 25 minutes (each 16-cm bed position takes 3 minutes). The clinical significance of this time discrepancy lies in the in the possibility of misregistration of lung lesions, especially near the base, and bladder filling during the PET portion of the scan that can obscure pelvic adenopathy. One way to correct for this limitation is to acquire pelvis-first PET images if this is known in advance to be an area of interest. The emission data are then iteratively reconstructed, processed, and exported via Digital Imaging and Communication in Medicine (DICOM) to the selected radiation treatment planning workstation.

Patients typically present for CT simulation on a different date from the PET-CT scan. CT simulation begins with patients placed in the treatment position for 3D planning. Images are usually acquired with a 2.5-mm slice thickness and are obtained for the treatment area of interest as opposed to a full body scan. The image set is then transferred to a radiation treatment planning system with fusion capabilities. If the treatment planning system is networked to radiology, then the PET-CT can be directly imported. Outside or non-networked studies can be imported into the treatment planning DICOM.

Co-registration, Fusion, and Target Delineation

Co-registration and fusion occurs between the CT component of the PET-CT scan and the CT planning image set. Software fusion can be accomplished via a variety of methods including the placement of fiducial markers or anatomic localization of similar points. At our institution, manual co-registration is undertaken with the selection of easily identifiable anatomic landmarks. We typically choose a minimum of three non-coplanar anatomic points on both image sets, and fusion is considered acceptable when a mean error of 2 mm by root-mean-square is achieved.

Contours typically include both a CT-gross tumor volume (GTV) and a PET-GTV. Contours on the PET portion of the PET-CT automatically transfer to the planning CT. When delineating the PET tumor volume, it is important to understand the threshold intensity that was selected in interpreting PET images and determining what shows increased FDG uptake. A commonly employed mechanism is to interpret images at the 50% threshold intensity or isocontour.^24,²⁵ A potential limitation of this approach is that if a tumor shows high avidity for FDG, selection of the 50% intensity may underestimate the true tumor volume. Conversely, the true volume for lesions with low avidity may be overestimated.

The impact of PET-CT fusion on the radiation field setup has not been firmly established. It is helpful to remember that involvement of nodal sites in lymphoma is more likely to represent a “pushing volume” rather than an infiltrative lesion with respect to the surrounding tissues. This knowledge can be useful in regions such as the mediastinum where protection of the adjacent lung is important. In these circumstances, it is common to use the pre-chemotherapy volume to design the inferior and superior extent of a radiation field and the post-chemotherapy volume for the lateral extent.

Pitfalls Associated with PET-CT

Some of the pitfalls of PET in regard to lymphoma management are described in the following paragraphs. We specifically focus on some of the limitations during its application for radiotherapy planning.

Interpretation of Pre-treatment PET-CT

While PET imaging generally provides information about tumor activity that complements anatomic findings on CT and facilitates assessment of treatment response, FDG avidity related to other nonmalignant conditions may, on occasion, confound interpretation of tumor response. Multiple factors can account for changes in PET avidity within a CT-defined mass. According to the literature, one of the most frequent causes of non-tumoral FDG uptake is non-specific inflammation; therefore, active inflammation/infection, granulomas, and abscesses can be associated with increased standardized uptake value (SUV) and can be falsely interpreted as a malignant process.^26,²⁷ Also, thymic hyperplasia can cause FDG uptake, especially following chemotherapy (so-called “thymic rebound”).²⁸ FDG uptake may also occur in up to 30% of normal thymus glands in patients younger than age 30.²⁹ More recently, the presence of brown adipose tissue has been described as related to focal FDG accumulation, especially in the neck and upper posterior thorax and causing some difficulty in PET scan interpretation.³⁰^–³² Previous lymphangio-graphy and radiation pneumonitis are other causes responsible for false-positive FDG uptake.^33,³⁴ Other, less frequent causes of potential pitfalls due to FDG uptake in treated malignant lymphoma patients have been described as esophagitis, gastritis, and colitis.³⁵ Moreover, the possibility of an infectious disease concomitant to malignant lymphoma, even if rarely described, should be taken into account by the clinician.^36,³⁷ There are reports on observed transient intense FDG-avidity of the skeleton and spleen in patients treated with growth-factor support following myelosuppressive therapy.³⁸ Subsequent resolution of these findings without therapeutic intervention establishes the relationship of FDG avidity and granulocyte-colony stimulating factor (G-CSF)-induced stimulation of myelopoiesis. If folow-up imaging shows improvement or resolution without a change in antineoplastic therapy, a non-malignant source of FDG avidity is suggested.