Early 18F-labeled fluoro-2-deoxy-D-glucose positron emission tomography scanning in the lymphomas

Changing the paradigms of treatments?

Abstract

The identification of refractory or nonresponding tumors at an early period during therapy may lead to abbreviation of the current therapy regimen or timely institution of an alternative therapy protocol. Nevertheless, evaluation of treatment response is consequential clinically if the tumor potentially is curable and if effective treatment alternatives exist, so that change in treatment ultimately may increase the probability of response and survival. Hodgkin disease (HD) and diffuse large cell lymphoma (LCL) fit in this model well. However, a subset of patients is either refractory to first-line treatment or develops recurrent disease after an initial remission. Improvements in the treatment of these diseases rely not only on new therapy modalities and accurate assessment of disease extent but also on the assessment of disease extent and on timely and accurate therapy response to enable a more effective management plan. Although the International Prognostic Score for HD and International Prognostic Index for LCL have proved valuable for the stratification of patients in clinical trials, there is variability in outcome within the individual risk groups. Recently, positron emission tomography using ¹⁸F-labeled fluoro-2-deoxy-D-glucose (FDG-PET) imaging has been suggested as a sensitive and relatively more specific means to reflect tumor biologic changes after therapy. With increasingly compelling evidence, early FDG-PET provides a reliable means to assess tumor response accurately that may lead to better management with an effective therapeutic approach. Cancer 2006. © 2006 American Cancer Society.

A continuing dilemma for modern cancer therapy is how much treatment is sufficient when it is succeeding and when to switch therapy if it is failing. For greater than 3 decades, 6 to 8 cycles of chemotherapy have been the standard of treatment for disseminated Hodgkin disease (HD) and large cell lymphoma (LCL), which are considered potentially curable lymphomas.1 Although the rational for these arbitrary treatment schedules was not based on firm scientific footing, there have been relatively few organized attempts to alter the cycles or duration of therapy.

The cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) regimen for LCL, as reported originally, employed 8 cycles of chemotherapy every 3 weeks for 24 weeks or 6 months.2 However, 6 months erroneously were interpreted as 6 cycles, which became the most widely used “standard” of treatment. Only recently, randomized studies confirmed that a 6-cycle course was as effective as 8 cycles and was somewhat less toxic.3, 4 Armitage and associates, using a modification of the cyclophosphamide, vincristine, prednisone, bleomycin, doxorubicin, and procarbazine (COPBLAM) regimen, adjusted the duration of treatment to the rapidity of response, giving 2 cycles of therapy beyond complete remission, which was determined at 3 months, 5 months, and 7 months.5, 6 Rapid response, requiring less therapy, was the key determinant in prognosis. Likewise, Engelhard et al. using response-adapted COPBLAM treatment in a prospective, multicenter, randomized trial to corroborate those observations.7 Haq et al. also noted the poor prognosis of partial or slow responding patients with aggressive non-Hodgkin lymphoma (NHL).8

Data generated in HD are similar to those observed with LCL and/or aggressive NHL. Coleman et al., reporting for Cancer and Leukemia Group B on a randomized study between 6 cycles and 12 cycles (months) of treatment for HD, observed that 6 cycles were superior and provided a similar response but with less toxicity.9 Those data confirmed that patients who are apt to respond will do so earlier.10

In patients with LCL and HD, attempts to avoid radiation, even in limited application, have gained adherents who have concerns about the long-term consequences to the thyroid, heart, breast, bone marrow, and lung, among other sites.11 Adjuvant radiation, however, remains a “gold standard” for bulky and/or lower stage disease, since freedom from recurrence and disease-free survival are enhanced significantly by its application. Recent data generated for patients with limited LCL and for patients with HD have brought into question the utility of radiation.12, 13 Although there is no doubt that radiation provides excellent local control, it is uncertain whether it has an impact on overall long-term survival of patients with LCL and HD, particularly for those with disseminated disease.9, 14 One vexing problem in the past has been the seeming inability to determine whether residual masses, which sometimes are seen after treatment of bulky disease, represent active disease or scar.15, 16 Because of this uncertainty, radiation therapy is usually employed. However, only a minority of these residual masses represent active disease.16, 17

Thus, the accurate assessment of response to therapy is of vital importance in the management of HD, LCL, and aggressive NHL. The primary thrust of treatment for these diseases is the achievement of a complete remission, which is the paramount necessary first step for cure. Previous evaluations for response relied conventionally on computed tomography (CT) scanning for tumor volume reduction or gallium scanning for residual metabolic activity. Positron emission tomography (PET) measuring the uptake of glucose analog, ¹⁸F-labeled fluoro-2-deoxy-D-glucose (FDG) currently is employed widely in the evaluation of HD and LCL.18–30 It is now well established that FDG-PET is superior to both CT and gallium for determining the extent of disease, response to therapy, and whether a residual mass after treatment represents scar or active disease.

Over the past decade, in an effort to tailor therapy more effectively, prognostic indicators have been advanced for both LCL and HD. Although both the International Prognostic Index (IPI) (for LCL) and the International Prognostic Score (IPS) (for HD) have proven useful, they have limitations.31, 32 The classic IPI does not discriminate well among patients with an intermediate prognosis and has proven less useful with the advent of rituximab.33 The IPS for HD only selects a small subset of patients with a poorer prognosis who may benefit from an alternative approach.32 Laboratory evaluation for cell proliferation or tumor suppression have provided only rough or variable prognostications.34 An inherent problem associated with these determinations has been their static quality, rather than reflecting the dynamics of a tumor's metabolic response to therapy, as with FDG-PET. FDG-PET appears to be a stronger predictor than the IPI.20, 23

The high accuracy of FDG-PET for determining the status of response in patients with LCL and HD after the completion of therapy has been well characterized.24–29 It has been incorporated into the International Workshop Criteria to provide more accurate response assessment.35 After the completion of therapy, the positive and negative predictive values for refractory disease or freedom from disease hover around 90%, generally ranging from 80% to 100% accuracy.24–29 Positive FDG-PET results occasionally may be problematic with false-positive readings because of thymic rebound, infection, secondary inflammation, sarcoid-like granulomatosis or brown adipose tissue; however, in general, the results remain highly accurate.36 Nevertheless, biopsy is always recommended prior to any substantial treatment alteration.

Recently, studies have established that an earlier, interim treatment FDG-PET scan performed after 1 to 4 cycles of therapy is equal or superior to scanning performed at the completion of therapy.18, 23 In a series of 47 patients who were studied after 1 cycle of therapy, all patients who had results that changed to negative remained free of disease at 24 months.37, 38 These data, in addition to similar results after 2, 3, and 4 cycles, offer many new possibilities and raise many questions. For instance, if early interim FDG-PET proves accurate, then are 6-cycle or 8-cycle regimens necessary if the study is negative after only 1 cycle of treatment? If the data reported previously by Armitage et al. using earlier, less sophisticated techniques prove true, then may the treatment be shortened to 3 or 4 cycles? Is radiation for bulky disease needed with a negative interim scan? What is to be done with an early, positive interim scan? Should therapy be altered if a follow-up scan after the next cycle of treatment remains strongly positive? Is a significant percentage drop of the metabolic activity (standardized uptake value [SUV]) to a mildly positive level, particularly for patients with limited disease, still consistent with a potential cure by radiotherapy, or does any early interim positivity connote ultimate failure? Even with the ability to identify early on that patients with a positive FDG-PET are destined to fair poorly with first-line therapy, is it possible that early alternative treatments, such as transplantation, may not result in any survival benefit? Indeed, an early, positive interim FDG-PET may be a sad harbinger of a dismal prognosis that reflects inherent drug resistance which may not be overcome by high dose therapy with stem cell transplantation.39

There remain many other important questions to be addressed. What is the optimal time to use FDG-PET for early response determination: after 1, 2, or 3 cycles? Should serial FDG-PET be performed for accurate prognostication? For instance, if the results are positive after 1 cycle, should FDG-PET be performed after the 2nd cycle? What is the most robust PET technique for assessing response? Does a qualitative “eyeball” approach, or a semiquantitative SUV score, or a combination of both yield the most accurate results? Will treatment paradigms change? If so, will survival be improved because of the therapy modifications based on FDG-PET results? Well planned, controlled, randomized trials are currently underway that are attempting to answer these crucial questions. Regardless, the use of FDG-PET for early, dynamic assessment of response represents an exciting new development in the management of lymphomas.