| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Commentary |
1 Synarc Inc. and University of California, San Francisco, CA, 2 University of Iowa Hospitals and Clinics, Iowa City, IA, USA
Correspondence: Ali Guermazi, MD, Synarc Inc., 575 Market Street, 17th Floor, San Francisco, CA 94105, USA.
Non-Hodgkin's lymphoma (NHL) is a group of heterogeneous tumours originating in lymphoid tissue. The outcome of patients with lymphoma is, in general, better than non-lymphoid malignancies. Indeed, aggressive NHLs, which make about two-thirds of the NHL subtypes, are potentially curable (about 60% of patients are cured) with conventional and high-dose therapy. More than 70% of patients with newly diagnosed lymphoma respond well to chemotherapy alone or to chemoimmunotherapy with or without radiotherapy. In addition to the availability of new treatments, accurate staging and risk stratification with appropriate selection of treatments, as well as improved assessment of response to therapy, have resulted in a higher success rate in the treatment of lymphoma. The increasing number of treatment options available for patients with lymphoma currently being tested in clinical trials clearly requires the availability of standardized guidelines for assessment of response to therapy, which should ensure comparability of success rates among these trials.
The International Workshop Criteria (IWC) for response assessment of NHL proved to be quite useful for standardized assessment of response to various lymphoma treatments tested in a significant number of clinical trials [1]. These 1999 criteria rely on clinical, biochemical, histopathological and imaging studies, especially CT. This first attempt at standardization of response criteria in NHL proved to be a significant step forward; yet with the advent of functional imaging with positron emission tomography (PET), particularly using fluorodeoxyglucose (FDG), the limitations of these criteria have become increasingly apparent.
Six years after their publication in the Journal of Clinical Oncology, it was clear that the IWC were in need of revision. Dr Bruce Cheson from Georgetown University in Washington DC, who organized the first International Workshop to Standardize Response Criteria for NHL in 1998, organized a meeting together with Dr Volker Diehl from the University of Cologne, Germany, called the International Harmonization Project (IHP), at which an invited international group of participants with clinical, histopathological and imaging expertise in NHL deliberated on the need for harmonization of clinical trial parameters and revision of the currently utilized response criteria in lymphoma. The meeting took place in June 2005 at the 9th International Conference on Malignant Lymphoma in Lugano, Switzerland. Six subcommittees were formed — response criteria, clinical features, endpoints, pathology/immunohistology, paediatrics and imaging. Each subcommittee presented the results of their deliberations.
The imaging subcommittee, including the authors of this commentary among others, stressed the importance of the introduction of new technologies, such as PET and PET/CT. Indeed, one of the most vexing clinical dilemmas in response assessment and management of NHL is the detection of active disease in a residual mass by conventional imaging (usually CT), following treatment despite resolution of clinical symptoms and normalization of laboratory tests (i.e. in patients with an otherwise clinical complete response). Such masses, usually seen in the mediastinum and abdomen/pelvis, may be seen in up to 40% of patients with aggressive NHL and an even higher percentage of patients with Hodgkin's disease. Yet the vast majority of these masses represent necrosis and/or fibrosis and only a small percentage represents or contains residual active lymphoma. It is also noteworthy that the frequency of occurrence of these residual masses correlates with the size of tumour mass(es) at diagnosis, with a much more common occurrence of residual masses in patients with initially bulky compared with non-bulky lymphoma. Obviously, characterization of these residual masses is important: residual lymphoma portends a poor prognosis and requires further salvage therapy, while an inactive residual mass usually confers an excellent prognosis without further treatment.
Residual masses can be assessed with biopsy/fine needle aspiration or even surgery. However, these approaches, while generally accurate, are quite invasive and associated with some morbidity. Non-invasive and yet accurate approaches for assessment of these masses are clearly warranted.
Conventional morphologic imaging (e.g. CT and conventional MRI) alone do not allow discrimination between active tumour and fibrosis and/or necrosis in residual masses. Most CT studies report poor specificity in the characterization of residual masses (40–50%). Conventional MRI has been shown to be slightly more reliable in this setting, but its performance is still unsatisfactory. Gallium scintigraphy (GS) has been shown to be substantially more accurate than CT and MRI for characterization of these masses, particularly in patients with aggressive lymphoma, with less reliable results in low-grade or follicular lymphoma that are substantially less gallium-avid. Importantly, GS is not well standardized and is also not a convenient imaging modality for both the patients and referring physicians because of the frequent requirement for imaging at multiple time points over several days to more than a week. The limited availability of the gallium-67 isotope, because of limited production and distribution, poses another significant problem. All of these issues in addition to the only moderate accuracy (about 60–70%) resulted in a relatively limited utilization of GS in both the USA and Europe.
The advent of PET promises to solve many of the problems associated with the use of GS, being much more convenient to use with less than 2 h required for a PET study to be completed. More importantly, PET clearly has higher accuracy than GS and morphologic imaging for assessment of most subtypes of lymphoma, including follicular histology. PET alone or coupled with CT (PET/CT) has demonstrated a high accuracy (i.e. > 80%) for assessment of response to therapy in patients with aggressive NHL, particularly with respect to characterization of residual masses, and it is now generally recognized that PET is the most accurate non-invasive method for differentiating necrosis and/or scar from viable tumour in residual masses [2–6]. Largely because of the recognized high accuracy of PET in characterizing the nature of residual masses following treatment of lymphoma, the IHP workshop has unequivocally agreed that PET or PET/CT should be recommended in assessment of response or restaging of aggressive NHL and Hodgkin's lymphoma (HL) at the completion of therapy.
This decision is supported by several persuasive studies by various groups of investigators who evaluated the predictive value of PET following first-line therapy of patients with aggressive NHL and/or HL compared with CT or conventional staging methods [3–5]. More recently, Juweid et al [6] went one step further and compared a proposed response classification based on integration of FDG-PET into the IWC (IWC+PET) with the IWC alone in patients with aggressive NHL. This retrospective study included 54 patients with aggressive NHL (87% diffuse large cell) who underwent PET and CT 1–16 weeks (median 4 weeks; 83% within 
3–11 weeks) after 4–8 cycles of chemotherapy (84% received 6–8 cycles) mostly with cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP) with or without rituximab. These patients were assessed for complete response (CR), unconfirmed complete response (CRu), partial response (PR), stable disease (SD) and progressive disease (PD) by the IWC and by specifically defined IWC+PET-based response designations, derived by integrating the PET findings into the previously defined IWC designations. Progression-free survival (PFS) was used as the primary outcome measure to compare the IWC-assigned and IWC+PET-assigned response designations. The investigators found that there was a strikingly higher proportion of patients with CR by IWC+PET compared with IWC alone (35/54 versus 17/54, respectively), with no patients designated as CRu by IWC+PET compared with 7 with IWC alone, because such patients were either designated as CR if PET was negative (fibrosis/necrosis) or PR if PET was positive (residual tumour). This latter finding is significant since, in fact, several studies have shown a substantial rate of CRu designations by the predominantly CT-based IWC classification due to the findings of residual masses by CT. Thus, using the IWC+PET classification, the vast majority of these CRus would be re-classified as CR or, less frequently, as PR. The most critical finding of the study by Juweid et al, however, was that based on PFS as an outcome measure, the IWC+PET-based response classification was more accurate than IWC-based classification in predicting the "expected" outcome of patients in the various response categories. For example, whereas two-thirds of the patients designated as PR by IWC+PET progressed within a median of only 8.5 months, almost 60% of patients with a PR designation by IWC were progression-free at a median of nearly 3 years post-therapy. Interestingly, in the 35 patients with CR by IWC+PET, PFS was identical to that in the only 17 patients with CR by IWC (31.5+ months), although slightly more than half of the IWC+PET CRs were actually thought to be in PR or even only having SD. This indicates that the "quality" of a CR by IWC+PET, in terms of conferring a much more favourable prognosis compared with PR or SD, was very similar if not equal to the quality of a CR by IWC. The potential implications for the management of patients with aggressive NHL are apparent: whereas patients with PR or SD by IWC+PET should be evaluated for persistent disease and, if confirmed, considered for additional therapy, patients with CR by IWC+PET may be safely observed. In contrast, at least the PR designation by IWC, assigned to slightly more than one-third of the patients in this retrospective study, provides incorrect characterization of the actual response to treatment in an unacceptably high fraction (60%) of patients with such designation, casting serious doubts about its usefulness.
Obviously, the study by Juweid et al [6] had its limitations: the retrospective nature of the evaluation performed; the relatively small number of patients evaluated; and the somewhat variable timing of PET following treatment and variable number and type of chemotherapy cycles administered. Yet this "proof-of-principle" study, along with similar evidence of superiority of PET compared with CT, were convincing enough that the IHP adopted IWC+PET-based definitions of CR, PR, SD and PD similar to those presented in the study by Juweid et al [6], with a proposal to completely eliminate the CRu designation. These new definitions were recently presented by Dr Bruce Cheson at the 2005 Annual meeting of the American Society of Hematology in Atlanta, Georgia [7]. In addition to the proposed elimination of the CRu designation, a noteworthy change compared with the criteria proposed by Juweid et al [6] is that the criteria will now take into consideration that some subtypes of NHL, such as extranodal marginal zone lymphoma (MZL) of mucosa-associated lymphoid tissue (MALT) are frequently not FDG-avid, hence assessment will still be based on conventional imaging. The criteria will now also be applicable to patients with typically FDG-avid indolent NHLs, such as follicular NHL, and also to patients with HL. The details of these revised criteria will be reported in an upcoming manuscript, to be submitted for publication shortly. A more imaging-focused report of the imaging committee of the IHP will follow subsequently.
Despite the clearly favourable contribution of PET or PET/CT on response assessment of NHL, it should be noted that, like any other modality, PET or PET/CT are not perfect, as shown by several studies including the study of Juweid et al [6]. In the latter study, one-third of the patients (4/12) with a PR designation by IWC+PET based on positive PET findings suggesting residual disease remained progression-free for more than 3 years. False-positive PET findings at the site of residual masses may occur due to thymic rebound in the vicinity of a residual mass in the superior mediastinum, sometimes mistaken as uptake in the mass, or post-therapy inflammatory changes, the latter substantially more frequent following radiation than after chemotherapy or chemoimmunotherapy [5, 8]. False-positive findings outside the site of residual masses also may be caused by rebound thymic hyperplasia, or infectious or inflammatory processes including sarcoidosis.
With PET/CT there is a significant improvement in the diagnostic accuracy of PET, principally because of the more accurate anatomic localization of the PET findings provided by the almost-simultaneously acquired CT. The generally precise registration of the PET and CT images largely eliminates the problem of false-positive physiologic FDG uptake, such as that in muscle or brown fat, which is highly variable between patients [8]. In fact, a recent study on the staging and re-staging of 73 lymphoma patients showed that PET/CT was superior to PET alone with reported accuracies of 93% and 83%, respectively [9]. To further improve the diagnostic accuracy of PET, it will also be critical to address the issue of standardization of PET technique (e.g. timing after last treatment, timing of scanning after the injection) in future studies.
Finally, it is important to note that PET and PET/CT are likely to be useful not only for response assessment after completion of therapy, but also for initial staging and early response assessment after only a few cycles (e.g. 1–4) of a 6–8-cycle chemotherapy or chemoimmunotherapy regimen. In fact, several studies have shown that an interim assessment with PET after 2–4 cycles of chemotherapy is at least as accurate an assessment of disease response in terms of prediction of progression free survival as an end of treatment assessment [8, 10–12].
In summary, PET or PET/CT is poised to alter the response assessment paradigm in patients with lymphoma which is currently largely based on conventional imaging and has, in fact, already resulted in a significant and important modification of the currently widely used and accepted IWC for response assessment of lymphoma. The full report of this revision is eagerly awaited.
Received for publication September 14, 2005. Revision received December 16, 2005. Accepted for publication January 16, 2006.
References
This article has been cited by other articles:
 |
|
 |
|
  K Strobel, N. Schaefer, C Renner, P Veit-Haibach, D Husarik, A. Koma, and T. Hany Cost-effective therapy remission assessment in lymphoma patients using 2-[fluorine-18]fluoro-2-deoxy-D-glucose-positron emission tomography/computed tomography: is an end of treatment exam necessary in all patients? Ann. Onc., April 1, 2007; 18(4): 658 - 664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 BJR review of the year - 2006 Br. J. Radiol., March 1, 2007; 80(951): 147 - 151. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
