Response Assessment of Aggressive Non-Hodgkin’s Lymphoma by Integrated International Workshop Criteria and Fluorine-18–Fluorodeoxyglucose Po
http://www.100md.com
《临床肿瘤学》
the Departments of Radiology, Internal Medicine, and Biostatistics, and the Holden Comprehensive Cancer Center, University of Iowa, Iowa City, IA
Department of Radiology, Mayo Clinic, Rochester, MN
Division of Nuclear Medicine, Ulm University Hospital, Ulm, Germany
Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE
Department of Internal Medicine, Georgetown University Hospital, Washington, DC
ABSTRACT
PURPOSE: To determine whether a response classification based on integration of fluorine-18–fluorodeoxyglucose positron emission tomography (FDG-PET) into the International Workshop Criteria (IWC) provides a more accurate response assessment than IWC alone in patients with non-Hodgkin's lymphoma (NHL).
PATIENTS AND METHODS: Fifty-four patients with aggressive NHL who underwent FDG-PET and computed tomography 1 to 16 weeks after four to eight cycles of chemotherapy with cyclophosphamide, doxorubicin, vincristine, and prednisone were assessed for complete response (CR), unconfirmed CR (CRu), partial response (PR), stable disease (SD), and progressive disease (PD) by the IWC and by integrated IWC and FDG-PET (IWC+PET). Progression-free survival (PFS) was also compared between IWC- and IWC+PET-assigned response designations.
RESULTS: By IWC, 17 patients had a CR, seven had a CRu, 19 had a PR, nine had SD, and two had PD. In comparison, by IWC+PET, 35 patients had a CR, 12 had a PR, six had SD, one had PD, and zero had a CRu. In separate multivariate models, PFS was significantly shorter in patients with PR than in those with a CR using IWC (hazard ratio [HR], 8.9; P = .021) or IWC+PET (HR, 29.7; P = .0003). However, when the two classifications were included in the same multivariate model, only IWC+PET was a statistically significant independent predictor for PFS (P = .008 v P = .72 for IWC). In addition, when patients with a PR by IWC and a CR by IWC+PET were compared with those with a CR by IWC and a CR by IWC+PET, there was no significant difference in PFS (HR, 1.6; P = .72), indicating that IWC+PET identified a subset of IWC-PR patients with a more favorable prognosis.
CONCLUSION: Compared with IWC, the IWC+PET-based assessment provides a more accurate response classification in patients with aggressive NHL.
INTRODUCTION
The International Workshop Criteria (IWC) are widely used and accepted for response assessment of non-Hodgkin’s lymphoma (NHL).1 These criteria are primarily based on computed tomography (CT), although bone marrow biopsy (BMB) and clinical and biochemical information are also taken into account when assigning a final response designation.1 The dominant role of CT in response assessment of NHLs warrants critical assessment considering that in perhaps no other cancer type, with the exception of Hodgkin’s lymphoma, has there been such a compelling documentation of the limitations of CT in assessment of response to therapy.1,3-16 These limitations are primarily due to the inability of CT to differentiate between viable tumor, necrosis, or fibrosis in residual mass(es) in patients with otherwise clinical complete response (CR), which occurs in approximately 40% of NHL patients treated with chemotherapy and/or radiation.1-15 Previous studies showing that only 10% to 20% of such patients have evidence of disease in these residual masses,1-6 considerably limits the value of CT for prediction of clinical outcome of NHL7,8,10,11,15 In contrast, metabolic imaging, particularly using fluorine-18 (18F) –fluorodeoxyglucose positron emission tomography (FDG-PET) has been shown to predict tumor viability (or lack thereof) in residual masses, with an accuracy approaching 80% to 90%.7,8,10,11,15 These results clearly suggest that compared with CT, FDG-PET more accurately reflects the actual response to treatment in these patients.
Despite the limited spatial resolution of current PET imaging systems, typically in the range of 8 to 12 mm in the clinical setting, FDG-PET has been shown to be more sensitive than CT in the detection of even small-volume tumors, such as in normal-sized lymph nodes (ie, < 1 cm in diameter), because of the greater contrast provided by the substantially enhanced tumoral FDG uptake compared with surrounding normal structures.17,18 Nevertheless, some studies have reported that FDG-PET may fail to detect tumor lesions less than 1 cm in diameter, which are often detectable by CT, when these exhibit no or only modestly increased FDG uptake compared with surrounding structures, and because of the effects of partial volume averaging.19,20 This scenario, albeit relatively rare, represents a limitation for PET that is generally addressed by CT despite its recognized lower specificity. Another potential limitation of PET that is related to partial volume averaging is the difficulty in deriving reliable estimates of tumor size or change in tumor size for pre- and/or post-therapy lesions that are smaller than twice the spatial resolution of the PET camera (ie, < approximately 1.5 to 2.5 cm in diameter). This issue becomes relevant when a determination of the exact percentage change (decrease or increase) in the size of such lesions is required for a correct assignment of a response designation after treatment.1
Because advantages of these two imaging modalities appear to be complementary, and because bone marrow sampling, clinical information, and biochemical analysis are also important for evaluation of response, we assessed whether integration of FDG-PET into the primarily CT-based IWC (IWC+PET) is a better predictor of progression-free survival (PFS) and, thereby, potentially results in more accurate response classification compared with IWC alone in patients with aggressive NHL treated with anthracycline-based chemotherapy. In contrast to previous investigations in this area, this study specifically defined IWC+PET-based response designations, derived by integrating the PET findings into the IWC designations, and then examined the association of the IWC- and IWC+PET-assigned response designations with patient outcome.
PATIENTS AND METHODS
Patient Selection and Investigation
All previously untreated NHL patients with a histologic diagnosis of an aggressive NHL treated at the University of Iowa (Iowa City, IA), the Mayo Clinic Rochester, MN), the University of Nebraska (Omaha, NE), and the University of Ulm (Ulm, Germany) between 1994 and early 2002, who met the following criteria were included in this retrospective evaluation: received four to eight cycles of anthracycline or anthracenedione-based chemotherapy; underwent post-therapy FDG-PET within 4 weeks of CT, with both scans performed within 1 to 16 weeks after chemotherapy; and had adequate clinical, laboratory, and radiologic follow-up for at least 18 months after beginning of therapy or until radiologic and/or histopathologic evidence of disease progression. Patients who received radiotherapy after their FDG-PET without a follow-up PET thereafter were not eligible for this study. Patients with incomplete resolution of radiographic abnormalities on CT (ie, those with a partial response [PR] or unconfirmed complete response [CRu] by IWC) were managed expectantly unless tissue biopsy was available to confirm residual active disease. Salvage chemotherapy was otherwise only given if there was evidence of disease progression by follow-up radiologic examinations (usually CT). In a few patients, a clinical decision was made to add consolidative radiation therapy to residual masses of uncertain significance. Only one patient in the current study who underwent PET scanning after all primary chemotherapy and radiation therapy was included in the current analysis.
Imaging Protocol
For PET imaging, patients were injected with 10 to 15 mCi of 18FDG after fasting for 4 hours before 18FDG injection and if they had a glucose level of 200 mg/dL at the time of injection.
Sixty minutes after the administration of 18FDG, two-dimensional whole-body PET scans encompassing the region between the earlobes and the proximal thighs were obtained using a model 4096 PET tomograph (GE Medical Systems, Waukesha, WI) at the University of Iowa, an Advance model (GE) at the Mayo Clinic, an ECAT EXACT (CTI/Siemens, Knoxville, TN) at the University of Nebraska, and CTI ECAT EXACT or HR+ (Siemens Medical Systems, Hoffman Estates, IL) at the University of Ulm. The emission scans (5 minutes per bed position) were followed by germanium-68 transmission scans (3 minutes per bed position) and the PET projection data were corrected for random coincidences, scatter, and attenuation. Transaxial images were reconstructed into 128 x 128 pixel images with a pixel size of 4.5 mm. The reconstructed PET images were then displayed on a computer workstation for viewing.
Baseline and follow-up helical CT scans of neck, chest, abdomen, and pelvis were obtained using intravenous contrast.
FDG-PET and CT Scan Interpretation
FDG-PET scans were interpreted by an experienced nuclear physician blinded to patient outcome independent of the clinical reading performed at the time of the actual study. However, the PET scans were interpreted together with pertinent clinical and CT scan findings to mimic the usual clinical situation, in which this information is incorporated into the assessment. This review was especially important if equivocal PET findings were subsequently explained by CT scan findings of an anatomic variation or clinical information (eg, increased uptake at a site of recent surgery or biopsy). Furthermore, the CT scan helped in deciding whether a particular area of increased uptake on the PET scan localized to a residual mass in the mediastinum or abdomen/pelvis, or was more associated with a vascular structure or physiologic bowel activity.
Visual assessment was used to categorize the FDG-PET scan findings as positive or negative. Abnormal (positive) 18FDG uptake using visual assessment was defined as any focal or diffuse area of increased activity, in a location incompatible with normal anatomy (eg, bowel, myocardium).11,12 Only in the case of questionable lung lesions and no prior history of pulmonary lymphoma was a different interpretation criterion used: a lesion was considered compatible with lymphoma only if its uptake exceeded that of mediastinal blood pool structures.21
If equivocal interpretation of the PET findings by the experienced nuclear medicine physician occurred, the scan was reinterpreted for consensus as positive or negative by one additional experienced nuclear physician.
CT scans were interpreted by an experienced radiologist blinded to patient outcome and the results of the FDG-PET scan. Blinding the radiologist to the PET scan findings was performed to assess accurately the impact of adding PET to the conventional interpretation of CT without a concomitant PET scan, the primary determinant of the IWC-based designations. Here again, if equivocal CT scans were interpreted by the experienced radiologist, the scan was reinterpreted for consensus by one additional experienced radiologist, who was also blinded to patient outcome.
IWC+PET-Based Response Designations
To determine whether the IWC+PET-based response assessment resulted in the assignment of a more accurate response classification compared with the IWC-based assessment alone in the studied patient population, a simplified approach was first used to define IWC+PET-based response designations, derived by integrating the PET findings into the predominantly CT-based IWC designations.1,22,23 This new IWC+PET-based classification continues to use standard response designations, such as CR, CRu, PR, and stable and progressive disease (SD and PD, respectively), albeit with different definitions for these designations compared with the IWC.
The IWC-based designations, as reported by Cheson et al1 are as follows. CR is complete disappearance of all detectable clinical and radiographic evidence of disease and disappearance of all disease-related symptoms if present before therapy, and normalization of biochemical abnormalities. BMB must be negative if it was positive before therapy. Previously involved nodes or nodal masses on CT more than 1.5 cm in their greatest diameter must regress to 1.5 cm and previously involved nodes/nodal masses of 1.1 to 1.5 cm must regress to 1.0 cm post-therapy. CRu is complete disappearance of all detectable clinical evidence of disease and disappearance of all disease-related symptoms if present before therapy, and normalization of biochemical abnormalities, but with a residual lymph mass more than 1.5 cm in greatest transverse diameter that has regressed by more than 75% or indeterminate bone marrow (increased number or size of aggregates without cytologic or architectural atypia). PR is at least 50% reduction in the sum of the product of the greatest diameters (SPD) of the six largest dominant nodes or nodal masses with no increase in the size of other nodes, liver, or spleen, and with no new sites of disease. BMB is irrelevant for determination of PR. SD is less than a PR but is not progressive disease. PD (for patients with PR or no response) is 50% increase in the SPD from nadir of any previously identified abnormal node for PRs and nonresponders, or appearance of any new lesion during or at the end of therapy. Relapsed disease (RD; for patients with a CR or CRu) is the appearance of any new lesion or increase by 50% in the size of previously involved sites, or 50% increase in the greatest diameter of any previously identified node greater than 1 cm in short axis or in the SPD of more than one node.
Table 1 shows the various IWC+PET-based designations as determined based on the IWC designations and PET findings, with definitions for CR, CRu, PR, SD, and PD. These definitions were established before examination of PFS data. Given that the emphasis in the current investigation is on the assessment of response immediately (ie, within 16 weeks) after completion of chemotherapy and not RD in patients who achieved a CR or CRu post-therapy, no attempt was made to define RD on the basis of IWC+PET.
A CR designation by IWC+PET may be assigned to patients with a CR, CRu, PR, SD, or even PD by IWC if the PET scan is completely negative. However, for PD indicated by IWC, the CT abnormality that led to the PD designation but that is negative by PET must be 1.5 cm in diameter ( 1.0 cm for lung lesions) to qualify for CR status by IWC+PET because of the known limited spatial resolution of the PET system, with possible nonvisualization of potentially malignant lesions less than 1.5 cm. The lower threshold of 1.0 cm for lung lesions is justified by the relatively low background activity in the lungs compared with other normal organs/tissues, thereby usually allowing the PET visualization of malignant lung lesions 1.0 cm. Furthermore, in patients with a CRu, PR, SD, or PD by IWC who are negative by PET, BMB must also be negative if it was positive before therapy. The low negative predictive value of FDG-PET for detecting limited marrow NHL and the commonly observed nonspecific increase in FDG marrow uptake post-therapy makes PET alone unreliable for assessment of lymphomatous marrow involvement.24
The CRu designation by IWC+PET only applies to patients classified as CRu by IWC based on an indeterminate BMB, again because of the unreliability of PET alone for assessment of marrow NHL.24 In contrast, patients classified as CRu by IWC on the basis of only a residual mass that regressed by 75% from its original size are classified as CR by IWC+PET if PET is negative, or as PR if PET is positive at this site.
The rationale for using the PR rather than the PD designation in patients with a CR, CRu, PR, or SD by IWC who have a positive PET finding outside a previously involved site lies in the fact that, lacking a baseline PET scan, it cannot be ascertained if the abnormality seen by PET but not on CT is new or pre-existent, but only detectable by PET. The reason for using the PR rather than the SD designation in patients with SD by IWC who have a positive PET finding only at the site of a previously involved node that regressed to 1.5 cm by CT (or 1.0 cm if baseline node was 1.1 to 1.5 cm) is because the anatomically more significant residual mass(es) by CT dictating the SD designation by IWC is PET negative. Hence, the SD designation by IWC+PET is reserved for patients with SD by IWC in whom the positive PET findings are demonstrated at the site of the previously involved mass(es) that did not show a significant change in size.
Finally, the PD designation by IWC+PET is assigned in two different scenarios: PD by IWC with positive PET findings corresponding to the new/progressing CT abnormalities, or PD by IWC based on new CT abnormalities less than 1.5 cm (< 1.0 cm in the lung), regardless of the PET findings.
Statistical Analysis
PFS was measured in months from the beginning of treatment until disease progression, death, or until the patient was last known to be alive. PFS was estimated by the Kaplan-Meier method.25 The SE estimates for the 2- and 3-year PFS rates were computed using Greenwood’s formula.26 Differences in PFS were determined between the various response designations by IWC- and IWC+PET-based classifications (Table 1), with a priori interest in the CR versus PR comparisons. Multivariate Cox regression analysis27 was performed to adjust for the potential effect of pretherapy prognostic factors and other potentially important factors related to administered treatment on PFS when comparing the IWC or IWC+PET classifications. More specifically, the effects of well-established baseline prognostic factors that were clinically available in this retrospective evaluation were examined. These included age, histologic type, and disease stage.
The other factors related to the treatment administered included chemotherapy type and number of chemotherapy cycles. Separate multivariate models were evaluated for IWC and IWC+PET, but these classifications also were included in the same model to be evaluated simultaneously. P values from Cox regression were based on the likelihood ratio test. Estimates of differences in PFS between particular response designations using IWC or IWC+PET (eg, CR v PR), after adjustment for these prognostic factors, are expressed as hazard ratios (HRs) for progression. Ninety-five percent CIs for HRs and 2- or 3-year PFS rates were based on normal approximations. The proportional hazards assumption was assessed as discussed by Allison.28 All P values were two-tailed and P values less than .05 were considered statistically significant. All statistical analyses were performed using SAS version 8.2 (SAS Institute, Cary, NC, 2001). Survival curves were created using SPlus 2000 (Professional Release 1; Math Soft Inc, Seattle, WA, 1988 to 1999).
RESULTS
Patient Characteristics
Table 2 shows the characteristics of the 54 patients studied. Fifty-one patients (94%) received cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) –based chemotherapy. Ninety-four percent had their PET scans within approximately 2 to 12 weeks; only three patients (6%) had their PET scans performed outside this range: at 8, 95, and 108 days post-therapy. The median number of days between the PET and CT scans was 3.5; 72% of patients had their PET scans within 8 days of CT.
PFS in All Patients
The 2- and 3-year PFS rates for the 54 patients were 70.4% and 63.1%, respectively. Among patients still progression free at last follow-up, the median follow-up was 35 months (range, 25.1 to 60.2 months). On the basis of multivariate analysis, chemotherapy type and number of chemotherapy cycles were significant prognostic factors (P = .008 and P < .0001, respectively). In contrast, age, disease stage, and histologic type were not significant prognostic factors in the present study (P = .34, P = .69, and P = .56, respectively). However, these factors were included in the adjustment because some of these variables are used in the calculations for the International Prognostic Index.29
Response Designations According to the IWC and IWC+PET Classifications and Association With PFS
On the basis of the IWC, 17 patients had a CR, seven had a CRu, 19 had a PR, nine had SD, and two had PD (Table 3). In comparison, by IWC+PET, 35 patients had a CR, 12 had a PR, six had SD, one had PD, and zero had a CRu. Table 3 lists the concordance/discordance between the IWC and IWC+PET response designations in the 54 patients studied. Overall, 33 patients (61.1%; 95% CI, 48.1% to 74.1%) had concordant designations and 21 patients had discordant designations. Figures 1 and 2 show two representative examples of discordance between the IWC and IWC+PET classifications.
The most pronounced discordance was observed in the CRu by IWC designation, in which all seven CRu designations by IWC patients were reclassified as CR (n = 5) or PR (n = 2) by IWC+PET. The CRu designation by IWC in all seven patients was dictated by the occurrence of a residual mass that regressed by 75% of its original size and not by an indeterminate bone marrow. Given that five of the seven patients were completely PET negative and none had bone marrow involvement with NHL at baseline (a post-therapy BMB was therefore not required according to the recommendations of the International Workshop1), the patients were appropriately reclassified as CR by IWC+PET. All five patients remain progression free at a median of more than 32 months. The other two CRu designations by IWC patients were PET positive at the site of the residual mass and, hence were reclassified as PR by IWC+PET. One of the two patients experienced disease progression at 17 months post-therapy, whereas the other remains progression free at 36 months of follow-up.
The other major discordance was found in the PR by IWC designation, in which 10 of the 19 patients (52.6%) in this category were reclassified as CR by IWC+PET. Because all of the PR by IWC patients reclassified as CR by IWC+PET had radiologic abnormalities post-therapy, post-therapy BMB was not required based on the International Workshop recommendations. However, based on the IWC+PET classification, such patients can only be considered true CRs if there was no clinical or laboratory evidence of disease and the BMB was negative, if initially positive. None of the 10 patients had clinical or laboratory evidence of disease. Seven of 10 patients did not have evidence of bone marrow involvement at baseline. Two of three patients who had previous involvement had a negative post-therapy BMB, but one with residual hepatic CT abnormalities did not undergo a post-therapy BMB. This patient experienced disease relapse, however, 28.6 months post-therapy at the CT-positive post-therapy site (liver) and not in the marrow; hence, it is presumed that the patient’s marrow was negative post-therapy. Thus, all 10 patients with PR by IWC reclassified as CR by IWC+PET were appropriately classified as such according to the International Workshop guidelines for establishing a CR designation. All but the one patient who experienced relapse remain without evidence of disease progression at 26.4 to 38 months (median, 33.8 months) of follow-up; the estimated 3-year PFS rate in this group is 88.9%. In contrast, only two of the nine patients with concordant PR designation by IWC and IWC+PET are progression free at 26.8 and 53.1 months of follow-up, respectively. The other seven patients experienced disease progression at a median of 7 months post-therapy.
Table 4 lists the 2- and 3-year PFS rates in the various response designations by IWC and IWC+PET.
Of the 17 patients with a CR by IWC, four experienced disease progression at a median of 17.1 months (range, 8 to 31 months), and one of the seven patients with CRu experienced disease progression at 17 months. Eight of the 19 PR patients progressed at a median of 10.7 months, and six of the nine patients with SD progressed at a median of 4.2 months. Only one of the two PD patients progressed at 3.4 months.
Of the 35 patients with a CR by IWC+PET, six experienced disease progression at a median of 17.2 months (range, 8 to 31 months). Eight of the 12 patients with a PR experienced disease progression at a median of 8.5 months, and five of the six patients with SD experienced disease progression at a median of 4 months. The only patient with PD by IWC+PET (also PD by IWC) experienced disease progression at 3.4 months.
Median PFS of the 17 patients with a CR by IWC was identical to that of the 35 patients with a CR by IWC+PET (31.5+ months); the median PFS of the seven patients with CRu by IWC was similar (32.0+ months).
Figure 3 shows the Kaplan-Meier plots of PFS times for the patients with a CR and a PR by IWC and IWC+PET. In separate multivariate analyses with adjustment for age, disease stage, histologic type, chemotherapy type, and number of chemotherapy cycles, PFS time was significantly shorter in patients with a PR compared with those with a CR both based on the IWC and IWC+PET (P = .021 and P = .0003, respectively). The hazard of progression was 29.7-fold higher (95% CI, 4.7 to 189.5) for the PR than for the CR patients by IWC+PET compared with an HR of 8.9 (95% CI, 1.4 to 56.6) between the PR and CR patients by IWC. Moreover, when IWC (PR v CR) and IWC+PET (PR v CR) were included in the same multivariate model, the HR (PR v CR) for IWC was 1.6 (95% CI, 0.1 to 23.1; P = .72) compared with 35.9 (95% CI, 2.6 to 504.9; P = .008) for IWC+PET.
Because it is common to include the CRu values together with CR values when reporting response data in NHL using the IWC, and because our own data, albeit in a small number of patients, showed similar PFS times (Table 4) in patients with a CR by IWC and a CRu by IWC, we evaluated the differences between the PR and CR/CRu classifications. The Kaplan-Meier plots of PFS times for the PR and CR/CRu patients by IWC and IWC+PET are shown in Figure 4. The hazard of progression between PR by IWC and CR/CRu by IWC was 5.3 (95% CI, 1.2 to 24.7) compared with 29.7 (95% CI, 4.7 to 189.5) between PR by IWC+PET and CR/CRu by IWC+PET. Note that the latter ratio is identical to that between PR by IWC+PET and CR by IWC+PET because none of the patients had a CRu by IWC+PET. Here again, when IWC (PR v CR/CRu) and IWC+PET (PR v CR/CRu) were included in the same multivariate model, the HR for IWC was 1.0 (95% CI, 0.1 to 7.6; P = .97) compared with 48.8 (95% CI, 3.9 to 606.9; P = .003) for IWC+PET.
Finally, there was no significant difference in PFS between the 10 patients with a PR by IWC and a CR by IWC+PET versus the 17 patients with a CR by IWC and a CR by IWC+PET (HR, 1.6; 95% CI, 0.1 to 23.1; P = .72). In comparison, the hazard of progression for the nine patients with PR by IWC and PR by IWC+PET was 58.6 times that for the 17 patients with CR by IWC and CR by IWC+PET (P = .001; Fig 5) . This suggests that PET scanning identified two distinct subgroups of patients within the PR by IWC designation in terms of prognosis, one with a favorable and another with an unfavorable prognosis. The PR by IWC designation appears to be a mixture of these two subgroups with an intermediate prognosis (estimated 3-year PFS, 62.2% v 88.9% and 33.3% in the two subgroups, respectively).
DISCUSSION
The primary goal of this retrospective evaluation was to assess whether supplementing the IWC with PET information (IWC+PET) results in better discrimination in outcome between the various conventionally used response designations. As noted above, the decision to integrate PET with the IWC rather than to test its feasibility as a single modality for response assessment was based on considerations related to its recognized limitations, including the potential inability to characterize the nature of new lesions less than 1 to 1.5 cm in diameter and to reliably evaluate the status of bone marrow.
Perhaps the two most striking findings in our study are the substantially higher proportion of patients in the CR designation by IWC+PET compared with the IWC alone (35 of 54 v 17 of 54, respectively) and the higher HR between PR and CR or between PR and CR/CRu by IWC+PET compared with IWC. More importantly, when the two classifications were included in the same multivariate model, only IWC+PET was a statistically significant independent predictor for PFS (P = .008). The lower HR between the PR and CR designations by IWC is most likely related to the apparent existence of two distinct subgroups within the PR by IWC category with respect to outcome: a subgroup of PET-positive patients with poor outcome and another of PET-negative patients with excellent outcome (Fig 5). The biologic explanation of this finding is likely related to the fact that that the residual radiographic abnormalities represented necrosis and/or fibrosis in the vast majority of the PET-negative patients, whereas the abnormalities represented viable tumor in the majority of PET-positive patients.
A relatively small fraction of aggressive NHL treated at our institutions during the 8-year period could be included in this study. One obvious explanation for this fact is that not all patients treated underwent PET scanning after therapy. Unrestricted PET scanning was not approved for reimbursement in the United States until July 2001; PET was approved in July 1999 only when used as an alternative to gallium scanning. Many referring oncologists continued to request gallium scans until 2001 or even beyond because of their greater familiarity with gallium and abundant literature on its usefulness in the restaging of patients with lymphoma. Only in the last 3 years or so did PET begin to be more widely used in the restaging of lymphoma patients. All or some of these factors contributed to the fact that only a fraction of the treated patients underwent PET scanning at our institutions. Another explanation for the relatively small number of patients in our study is the relatively strict criteria for inclusion in the retrospective evaluation, resulting in exclusion of some patients who underwent PET scanning. For example, a substantial fraction of patients were excluded because of the requirement that both the PET and CT scans be performed within 4 weeks of each other; this exclusion was needed for a more rigorous comparison of PET and CT. However, it is important to note that all aggressive NHL patients treated at our institutions in the specified period who underwent PET scanning and met the entry criteria for the study were included.
One concern with a retrospective evaluation of IWC versus IWC+PET is that the performed post-therapy PET studies may be skewed toward patients demonstrating residual radiographic abnormalities by CT or other evidence of disease, with no or only few patients referred who had no clinical or radiographic evidence of disease. Because of the current lack of established guidelines for a standardized or routine use of PET, the decision for PET scanning was individualized for each patient. Yet, this practice did not seem to substantially affect the expected composition of CT-based response categories that is usually seen after CHOP-based chemotherapy in newly diagnosed aggressive NHL: although 69% of the 54 patients included in the analysis based on the entry criteria had residual radiographic abnormalities, 31% had no clinical or radiographic evidence of disease (ie, CRs by the IWC). Although somewhat low, this fraction is similar or close to the lower limits for CR rates expected for CHOP or rituximab+CHOP chemotherapy, for which the 95% CIs for CR rates range between 31% to 45% and 46% to 60%, respectively.30 In general, patients with CR by IWC underwent PET scanning because of residual lymph nodes of 1.1 to 1.5 cm in diameter at sites of previous nodes or nodal masses more than 1.5 cm in diameter. Although such patients are considered to have a CR by IWC (without clinical or biochemical evidence of disease and a negative BMB), the referring oncologist sought to confirm that these patients also had a metabolic/functional CR by PET. PET was also occasionally performed in patients with 1.0 cm residual nodes simply to establish this metabolic CR status. It is therefore important to note that none of these CR by IWC patients had a false-positive PET, indicating that the addition of PET to CT-based assessment does not result in misclassification of any appreciable fraction of such patients.
Although our study was performed without the use of the PET/CT systems, which combine a PET and a CT scanner in a single instrument, our study has obvious implications for PET/CT studies in response assessment of NHL. With PET/CT, the functional PET images are coregistered with the anatomic CT images obtained by the almost simultaneously acquired CT scan.31-34 This approach can result in a significant improvement in the diagnostic accuracy of PET, principally because the more accurate anatomic localization of the PET findings by coregistered CT leads to fewer false-positive PET interpretations due to interpatient variability in physiologic FDG uptake.33 Given that our data suggest that PET combined with visual correlation with a contrast-enhanced CT better predicts PFS compared with CT alone in this setting, it is likely that a truly integrated approach of fused PET and CT images will yield similar or superior results. In fact, a recently reported study in the restaging of 27 patients with lymphoma showed an identical accuracy of 96% for PET/CT and PET interpreted in visual correlation with contrast-enhanced CT, whereas contrast-enhanced CT alone had an accuracy of only 67%.32
In conclusion, response assessment using anatomic and functional imaging with CT and PET according to the IWC+PET classification appears to better discriminate between patients with a CR/CRu and a PR, with a larger difference in outcome between the response groups. Patients with PR or SD by IWC+PET should be evaluated for persistent disease and considered for additional therapy.
These results challenge the current paradigm in response assessment of patients with NHL, and require validation in a prospective trial with a larger number of patients.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the Lymphoma Specialized Programs of Research Excellence (SPORE) grant from the National Cancer Institute (CA972784) at the University of Iowa and the Mayo Clinic.
Presented in part at the American Society of Hematology Annual Meeting in San Diego, CA, December 6-9, 2003, and the 39th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Cheson BD, Horning SJ, Coiffier B, et al: Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas. J Clin Oncol 17:1244-1253, 1999
Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the International Criteria for Neuroblastoma Diagnosis, Staging, and Response to Treatment. J Clin Oncol 11:1466-1477, 1993
Fuks JZ, Aisner J, Wiernik PH: Restaging laparotomy in the management of the non-Hodgkin's lymphomas. Med Pediatr Oncol 10:429-438, 1982
Stewart FM, Williamson BR, Innes DJ, et al: Residual tumor masses following treatment for advanced histiocytic lymphoma. Cancer 55:620-623, 1985
Surbone A, Longo DL, DeVita VT Jr, et al: Residual abdominal masses in aggressive non-Hodgkin's lymphoma after combination chemotherapy: Significance and management. J Clin Oncol 6:1832-1837, 1988
Coiffier B, Gisselbrecht C, Herbrecht R, et al: LNH-84 regimen: A multicenter study of intensive chemotherapy in 737 patients with aggressive malignant lymphoma. J Clin Oncol 7:1018-1026, 1989
Canellos GP: Residual mass in lymphoma may not be residual disease. J Clin Oncol 6:931-933, 1988
Kaplan WD, Jochelson MS, Herman TS, et al: Gallium-67 imaging: A predictor of residual tumor viability and clinical outcome in patients with diffuse large-cell lymphoma. J Clin Oncol 8:1966-1970, 1990
Front D, Bar-Shalom R, Mor M, et al: Aggressive non-Hodgkin lymphoma: Early prediction of outcome with 67Ga scintigraphy. Radiology 214:253-257, 2000
Vose JM, Bierman PJ, Anderson JR, et al: Single-photon emission computed tomography gallium imaging versus computed tomography: Predictive value in patients undergoing high-dose chemotherapy and autologous stem-cell transplantation for non-Hodgkin's lymphoma. J Clin Oncol 14:2473-2479, 1996
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 94:429-433, 1999
Spaepen K, Stroobants S, Dupont P, et al: Prognostic value of positron emission tomography (PET) with fluorine 18 fluorodeoxyglucose ([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: Is [18F]FDG-PET a valid alternative to conventional diagnostic methods J Clin Oncol 19:414-419, 2001
Cremerius U, Fabry U, Neuerburg J, et al: Positron emission tomography with 18F-FDG to detect residual disease after therapy for malignant lymphoma. Nucl Med Commun 19:1055-1063, 1998
Rmer W, Hanauske A-R, Ziegler S, et al: Positron emission tomography in non-Hodgkin's lymphoma: Assessment of chemotherapy with Fluorodeoxyglucose. Blood 91:4464-4471, 1998
Kostakoglu L, Goldsmith SJ: Fluorine-18 fluorodeoxyglucose positron emission tomography in the staging and follow-up of lymphoma: Is it time to shift gears Eur J Nucl Med 27:1564-1578, 2000
Juweid M, Wiseman G, Menda Y, et al: FDG-PET predicts with high accuracy the 1-year progression-free survival (1-year PFS) of patients with aggressive non-Hodgkin's lymphoma (NHL) following anthracycline-based first-line chemotherapy. J Nucl Med 43:124P, 2002 (abstr 283)
Pieterman R, van Putten JWG, Meuzelaar JJ, et al: Preoperative staging of non-small cell lung cancer with positron-emission tomography. N Engl J Med 343:254-261, 2000
Newman JS, Francis IR, Kaminski MS, et al: Imaging of lymphoma with PET with 2-[18F] fluoro-2-deoxy-D-glucose: Correlation with CT. Radiology 190:111-116, 1994
Avril N, Rose CA, Schelling M, et al: Breast imaging with positron emission tomography and fluorine-18-fluorodeoxyglucose: Use and limitations. J Clin Oncol 18:3495-3502, 2000
Tyler DS, Onaitis M, Kherani A, et al: Positron emission tomography scanning in malignant melanoma. Cancer 89:1019-1025, 2000
Lowe VJ, Hoffman JM, Delong DM, et al: Semi-quantitative and visual analysis of FDG-PET images in pulmonary abnormalities. J Nucl Med 35:1771-1776, 1994
Juweid ME, Wiseman GA, Menda Y, et al: Integrated PET based response classification for non-Hodgkin's lymphoma (NH). J Clin Oncol 22:565S, 2004 (abstr 6533)
Juweid ME, Wiseman G, Menda Y, et al: Integrated PET plus CT-based response assessment of aggressive NHL. Blood 102:629a, 2003 (abstr 2321)
Kostakoglu L, Leonard JP, Coleman M, et al: Evaluation of FDG-ET in the detection of bone marrow involvement in lymphoma. J Nucl Med 41:71P, 2000 (abstr 279)
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, NY, John Wiley & Sons, 1980
Cox DR: Regression models and life tables (with discussion). J R Stat Soc B 34:187-220, 1972
Allison PD: Survival Analysis Using the SAS System: A Practical Guide. Cary, NC, SAS Institute Inc, 1995
Shipp MA, Harrington DP, Anderson JR, et al: A predictive model for aggressive non-Hodgkin's lymphoma. N Engl J Med 329:987-994, 1993
Coiffier B, Lepage E, Briere J, et al: CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med 346:235-242, 2002
Lardinois D, Weder W, Hany F, et al: Staging of non-small-cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 348:2500-2507, 2003
Freudenberg LS, Antoch G, Schutt P, et al: FDG-PET/CT in re-staging of patients with lymphoma. Eur J Nucl Med Mol Imaging 31:325-329, 2004
Allen-Auerbach M, Quon A, Weber WA, et al: Comparison between 2-deoxy-2-[(18)F]fluoro-D-glucose positron-emission tomography and positron-emission tomography/computed tomography hardware fusion for staging of patients with lymphoma. Mol Imaging Biol 6:411-416, 2004
Goerres GW, von Schulthes GK, Steinert HC: Why most of lung and head and neck cancer will be PET/CT. J Nucl Med 45:66S-71S, 2004(Malik E. Juweid, Gregory )
Department of Radiology, Mayo Clinic, Rochester, MN
Division of Nuclear Medicine, Ulm University Hospital, Ulm, Germany
Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE
Department of Internal Medicine, Georgetown University Hospital, Washington, DC
ABSTRACT
PURPOSE: To determine whether a response classification based on integration of fluorine-18–fluorodeoxyglucose positron emission tomography (FDG-PET) into the International Workshop Criteria (IWC) provides a more accurate response assessment than IWC alone in patients with non-Hodgkin's lymphoma (NHL).
PATIENTS AND METHODS: Fifty-four patients with aggressive NHL who underwent FDG-PET and computed tomography 1 to 16 weeks after four to eight cycles of chemotherapy with cyclophosphamide, doxorubicin, vincristine, and prednisone were assessed for complete response (CR), unconfirmed CR (CRu), partial response (PR), stable disease (SD), and progressive disease (PD) by the IWC and by integrated IWC and FDG-PET (IWC+PET). Progression-free survival (PFS) was also compared between IWC- and IWC+PET-assigned response designations.
RESULTS: By IWC, 17 patients had a CR, seven had a CRu, 19 had a PR, nine had SD, and two had PD. In comparison, by IWC+PET, 35 patients had a CR, 12 had a PR, six had SD, one had PD, and zero had a CRu. In separate multivariate models, PFS was significantly shorter in patients with PR than in those with a CR using IWC (hazard ratio [HR], 8.9; P = .021) or IWC+PET (HR, 29.7; P = .0003). However, when the two classifications were included in the same multivariate model, only IWC+PET was a statistically significant independent predictor for PFS (P = .008 v P = .72 for IWC). In addition, when patients with a PR by IWC and a CR by IWC+PET were compared with those with a CR by IWC and a CR by IWC+PET, there was no significant difference in PFS (HR, 1.6; P = .72), indicating that IWC+PET identified a subset of IWC-PR patients with a more favorable prognosis.
CONCLUSION: Compared with IWC, the IWC+PET-based assessment provides a more accurate response classification in patients with aggressive NHL.
INTRODUCTION
The International Workshop Criteria (IWC) are widely used and accepted for response assessment of non-Hodgkin’s lymphoma (NHL).1 These criteria are primarily based on computed tomography (CT), although bone marrow biopsy (BMB) and clinical and biochemical information are also taken into account when assigning a final response designation.1 The dominant role of CT in response assessment of NHLs warrants critical assessment considering that in perhaps no other cancer type, with the exception of Hodgkin’s lymphoma, has there been such a compelling documentation of the limitations of CT in assessment of response to therapy.1,3-16 These limitations are primarily due to the inability of CT to differentiate between viable tumor, necrosis, or fibrosis in residual mass(es) in patients with otherwise clinical complete response (CR), which occurs in approximately 40% of NHL patients treated with chemotherapy and/or radiation.1-15 Previous studies showing that only 10% to 20% of such patients have evidence of disease in these residual masses,1-6 considerably limits the value of CT for prediction of clinical outcome of NHL7,8,10,11,15 In contrast, metabolic imaging, particularly using fluorine-18 (18F) –fluorodeoxyglucose positron emission tomography (FDG-PET) has been shown to predict tumor viability (or lack thereof) in residual masses, with an accuracy approaching 80% to 90%.7,8,10,11,15 These results clearly suggest that compared with CT, FDG-PET more accurately reflects the actual response to treatment in these patients.
Despite the limited spatial resolution of current PET imaging systems, typically in the range of 8 to 12 mm in the clinical setting, FDG-PET has been shown to be more sensitive than CT in the detection of even small-volume tumors, such as in normal-sized lymph nodes (ie, < 1 cm in diameter), because of the greater contrast provided by the substantially enhanced tumoral FDG uptake compared with surrounding normal structures.17,18 Nevertheless, some studies have reported that FDG-PET may fail to detect tumor lesions less than 1 cm in diameter, which are often detectable by CT, when these exhibit no or only modestly increased FDG uptake compared with surrounding structures, and because of the effects of partial volume averaging.19,20 This scenario, albeit relatively rare, represents a limitation for PET that is generally addressed by CT despite its recognized lower specificity. Another potential limitation of PET that is related to partial volume averaging is the difficulty in deriving reliable estimates of tumor size or change in tumor size for pre- and/or post-therapy lesions that are smaller than twice the spatial resolution of the PET camera (ie, < approximately 1.5 to 2.5 cm in diameter). This issue becomes relevant when a determination of the exact percentage change (decrease or increase) in the size of such lesions is required for a correct assignment of a response designation after treatment.1
Because advantages of these two imaging modalities appear to be complementary, and because bone marrow sampling, clinical information, and biochemical analysis are also important for evaluation of response, we assessed whether integration of FDG-PET into the primarily CT-based IWC (IWC+PET) is a better predictor of progression-free survival (PFS) and, thereby, potentially results in more accurate response classification compared with IWC alone in patients with aggressive NHL treated with anthracycline-based chemotherapy. In contrast to previous investigations in this area, this study specifically defined IWC+PET-based response designations, derived by integrating the PET findings into the IWC designations, and then examined the association of the IWC- and IWC+PET-assigned response designations with patient outcome.
PATIENTS AND METHODS
Patient Selection and Investigation
All previously untreated NHL patients with a histologic diagnosis of an aggressive NHL treated at the University of Iowa (Iowa City, IA), the Mayo Clinic Rochester, MN), the University of Nebraska (Omaha, NE), and the University of Ulm (Ulm, Germany) between 1994 and early 2002, who met the following criteria were included in this retrospective evaluation: received four to eight cycles of anthracycline or anthracenedione-based chemotherapy; underwent post-therapy FDG-PET within 4 weeks of CT, with both scans performed within 1 to 16 weeks after chemotherapy; and had adequate clinical, laboratory, and radiologic follow-up for at least 18 months after beginning of therapy or until radiologic and/or histopathologic evidence of disease progression. Patients who received radiotherapy after their FDG-PET without a follow-up PET thereafter were not eligible for this study. Patients with incomplete resolution of radiographic abnormalities on CT (ie, those with a partial response [PR] or unconfirmed complete response [CRu] by IWC) were managed expectantly unless tissue biopsy was available to confirm residual active disease. Salvage chemotherapy was otherwise only given if there was evidence of disease progression by follow-up radiologic examinations (usually CT). In a few patients, a clinical decision was made to add consolidative radiation therapy to residual masses of uncertain significance. Only one patient in the current study who underwent PET scanning after all primary chemotherapy and radiation therapy was included in the current analysis.
Imaging Protocol
For PET imaging, patients were injected with 10 to 15 mCi of 18FDG after fasting for 4 hours before 18FDG injection and if they had a glucose level of 200 mg/dL at the time of injection.
Sixty minutes after the administration of 18FDG, two-dimensional whole-body PET scans encompassing the region between the earlobes and the proximal thighs were obtained using a model 4096 PET tomograph (GE Medical Systems, Waukesha, WI) at the University of Iowa, an Advance model (GE) at the Mayo Clinic, an ECAT EXACT (CTI/Siemens, Knoxville, TN) at the University of Nebraska, and CTI ECAT EXACT or HR+ (Siemens Medical Systems, Hoffman Estates, IL) at the University of Ulm. The emission scans (5 minutes per bed position) were followed by germanium-68 transmission scans (3 minutes per bed position) and the PET projection data were corrected for random coincidences, scatter, and attenuation. Transaxial images were reconstructed into 128 x 128 pixel images with a pixel size of 4.5 mm. The reconstructed PET images were then displayed on a computer workstation for viewing.
Baseline and follow-up helical CT scans of neck, chest, abdomen, and pelvis were obtained using intravenous contrast.
FDG-PET and CT Scan Interpretation
FDG-PET scans were interpreted by an experienced nuclear physician blinded to patient outcome independent of the clinical reading performed at the time of the actual study. However, the PET scans were interpreted together with pertinent clinical and CT scan findings to mimic the usual clinical situation, in which this information is incorporated into the assessment. This review was especially important if equivocal PET findings were subsequently explained by CT scan findings of an anatomic variation or clinical information (eg, increased uptake at a site of recent surgery or biopsy). Furthermore, the CT scan helped in deciding whether a particular area of increased uptake on the PET scan localized to a residual mass in the mediastinum or abdomen/pelvis, or was more associated with a vascular structure or physiologic bowel activity.
Visual assessment was used to categorize the FDG-PET scan findings as positive or negative. Abnormal (positive) 18FDG uptake using visual assessment was defined as any focal or diffuse area of increased activity, in a location incompatible with normal anatomy (eg, bowel, myocardium).11,12 Only in the case of questionable lung lesions and no prior history of pulmonary lymphoma was a different interpretation criterion used: a lesion was considered compatible with lymphoma only if its uptake exceeded that of mediastinal blood pool structures.21
If equivocal interpretation of the PET findings by the experienced nuclear medicine physician occurred, the scan was reinterpreted for consensus as positive or negative by one additional experienced nuclear physician.
CT scans were interpreted by an experienced radiologist blinded to patient outcome and the results of the FDG-PET scan. Blinding the radiologist to the PET scan findings was performed to assess accurately the impact of adding PET to the conventional interpretation of CT without a concomitant PET scan, the primary determinant of the IWC-based designations. Here again, if equivocal CT scans were interpreted by the experienced radiologist, the scan was reinterpreted for consensus by one additional experienced radiologist, who was also blinded to patient outcome.
IWC+PET-Based Response Designations
To determine whether the IWC+PET-based response assessment resulted in the assignment of a more accurate response classification compared with the IWC-based assessment alone in the studied patient population, a simplified approach was first used to define IWC+PET-based response designations, derived by integrating the PET findings into the predominantly CT-based IWC designations.1,22,23 This new IWC+PET-based classification continues to use standard response designations, such as CR, CRu, PR, and stable and progressive disease (SD and PD, respectively), albeit with different definitions for these designations compared with the IWC.
The IWC-based designations, as reported by Cheson et al1 are as follows. CR is complete disappearance of all detectable clinical and radiographic evidence of disease and disappearance of all disease-related symptoms if present before therapy, and normalization of biochemical abnormalities. BMB must be negative if it was positive before therapy. Previously involved nodes or nodal masses on CT more than 1.5 cm in their greatest diameter must regress to 1.5 cm and previously involved nodes/nodal masses of 1.1 to 1.5 cm must regress to 1.0 cm post-therapy. CRu is complete disappearance of all detectable clinical evidence of disease and disappearance of all disease-related symptoms if present before therapy, and normalization of biochemical abnormalities, but with a residual lymph mass more than 1.5 cm in greatest transverse diameter that has regressed by more than 75% or indeterminate bone marrow (increased number or size of aggregates without cytologic or architectural atypia). PR is at least 50% reduction in the sum of the product of the greatest diameters (SPD) of the six largest dominant nodes or nodal masses with no increase in the size of other nodes, liver, or spleen, and with no new sites of disease. BMB is irrelevant for determination of PR. SD is less than a PR but is not progressive disease. PD (for patients with PR or no response) is 50% increase in the SPD from nadir of any previously identified abnormal node for PRs and nonresponders, or appearance of any new lesion during or at the end of therapy. Relapsed disease (RD; for patients with a CR or CRu) is the appearance of any new lesion or increase by 50% in the size of previously involved sites, or 50% increase in the greatest diameter of any previously identified node greater than 1 cm in short axis or in the SPD of more than one node.
Table 1 shows the various IWC+PET-based designations as determined based on the IWC designations and PET findings, with definitions for CR, CRu, PR, SD, and PD. These definitions were established before examination of PFS data. Given that the emphasis in the current investigation is on the assessment of response immediately (ie, within 16 weeks) after completion of chemotherapy and not RD in patients who achieved a CR or CRu post-therapy, no attempt was made to define RD on the basis of IWC+PET.
A CR designation by IWC+PET may be assigned to patients with a CR, CRu, PR, SD, or even PD by IWC if the PET scan is completely negative. However, for PD indicated by IWC, the CT abnormality that led to the PD designation but that is negative by PET must be 1.5 cm in diameter ( 1.0 cm for lung lesions) to qualify for CR status by IWC+PET because of the known limited spatial resolution of the PET system, with possible nonvisualization of potentially malignant lesions less than 1.5 cm. The lower threshold of 1.0 cm for lung lesions is justified by the relatively low background activity in the lungs compared with other normal organs/tissues, thereby usually allowing the PET visualization of malignant lung lesions 1.0 cm. Furthermore, in patients with a CRu, PR, SD, or PD by IWC who are negative by PET, BMB must also be negative if it was positive before therapy. The low negative predictive value of FDG-PET for detecting limited marrow NHL and the commonly observed nonspecific increase in FDG marrow uptake post-therapy makes PET alone unreliable for assessment of lymphomatous marrow involvement.24
The CRu designation by IWC+PET only applies to patients classified as CRu by IWC based on an indeterminate BMB, again because of the unreliability of PET alone for assessment of marrow NHL.24 In contrast, patients classified as CRu by IWC on the basis of only a residual mass that regressed by 75% from its original size are classified as CR by IWC+PET if PET is negative, or as PR if PET is positive at this site.
The rationale for using the PR rather than the PD designation in patients with a CR, CRu, PR, or SD by IWC who have a positive PET finding outside a previously involved site lies in the fact that, lacking a baseline PET scan, it cannot be ascertained if the abnormality seen by PET but not on CT is new or pre-existent, but only detectable by PET. The reason for using the PR rather than the SD designation in patients with SD by IWC who have a positive PET finding only at the site of a previously involved node that regressed to 1.5 cm by CT (or 1.0 cm if baseline node was 1.1 to 1.5 cm) is because the anatomically more significant residual mass(es) by CT dictating the SD designation by IWC is PET negative. Hence, the SD designation by IWC+PET is reserved for patients with SD by IWC in whom the positive PET findings are demonstrated at the site of the previously involved mass(es) that did not show a significant change in size.
Finally, the PD designation by IWC+PET is assigned in two different scenarios: PD by IWC with positive PET findings corresponding to the new/progressing CT abnormalities, or PD by IWC based on new CT abnormalities less than 1.5 cm (< 1.0 cm in the lung), regardless of the PET findings.
Statistical Analysis
PFS was measured in months from the beginning of treatment until disease progression, death, or until the patient was last known to be alive. PFS was estimated by the Kaplan-Meier method.25 The SE estimates for the 2- and 3-year PFS rates were computed using Greenwood’s formula.26 Differences in PFS were determined between the various response designations by IWC- and IWC+PET-based classifications (Table 1), with a priori interest in the CR versus PR comparisons. Multivariate Cox regression analysis27 was performed to adjust for the potential effect of pretherapy prognostic factors and other potentially important factors related to administered treatment on PFS when comparing the IWC or IWC+PET classifications. More specifically, the effects of well-established baseline prognostic factors that were clinically available in this retrospective evaluation were examined. These included age, histologic type, and disease stage.
The other factors related to the treatment administered included chemotherapy type and number of chemotherapy cycles. Separate multivariate models were evaluated for IWC and IWC+PET, but these classifications also were included in the same model to be evaluated simultaneously. P values from Cox regression were based on the likelihood ratio test. Estimates of differences in PFS between particular response designations using IWC or IWC+PET (eg, CR v PR), after adjustment for these prognostic factors, are expressed as hazard ratios (HRs) for progression. Ninety-five percent CIs for HRs and 2- or 3-year PFS rates were based on normal approximations. The proportional hazards assumption was assessed as discussed by Allison.28 All P values were two-tailed and P values less than .05 were considered statistically significant. All statistical analyses were performed using SAS version 8.2 (SAS Institute, Cary, NC, 2001). Survival curves were created using SPlus 2000 (Professional Release 1; Math Soft Inc, Seattle, WA, 1988 to 1999).
RESULTS
Patient Characteristics
Table 2 shows the characteristics of the 54 patients studied. Fifty-one patients (94%) received cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) –based chemotherapy. Ninety-four percent had their PET scans within approximately 2 to 12 weeks; only three patients (6%) had their PET scans performed outside this range: at 8, 95, and 108 days post-therapy. The median number of days between the PET and CT scans was 3.5; 72% of patients had their PET scans within 8 days of CT.
PFS in All Patients
The 2- and 3-year PFS rates for the 54 patients were 70.4% and 63.1%, respectively. Among patients still progression free at last follow-up, the median follow-up was 35 months (range, 25.1 to 60.2 months). On the basis of multivariate analysis, chemotherapy type and number of chemotherapy cycles were significant prognostic factors (P = .008 and P < .0001, respectively). In contrast, age, disease stage, and histologic type were not significant prognostic factors in the present study (P = .34, P = .69, and P = .56, respectively). However, these factors were included in the adjustment because some of these variables are used in the calculations for the International Prognostic Index.29
Response Designations According to the IWC and IWC+PET Classifications and Association With PFS
On the basis of the IWC, 17 patients had a CR, seven had a CRu, 19 had a PR, nine had SD, and two had PD (Table 3). In comparison, by IWC+PET, 35 patients had a CR, 12 had a PR, six had SD, one had PD, and zero had a CRu. Table 3 lists the concordance/discordance between the IWC and IWC+PET response designations in the 54 patients studied. Overall, 33 patients (61.1%; 95% CI, 48.1% to 74.1%) had concordant designations and 21 patients had discordant designations. Figures 1 and 2 show two representative examples of discordance between the IWC and IWC+PET classifications.
The most pronounced discordance was observed in the CRu by IWC designation, in which all seven CRu designations by IWC patients were reclassified as CR (n = 5) or PR (n = 2) by IWC+PET. The CRu designation by IWC in all seven patients was dictated by the occurrence of a residual mass that regressed by 75% of its original size and not by an indeterminate bone marrow. Given that five of the seven patients were completely PET negative and none had bone marrow involvement with NHL at baseline (a post-therapy BMB was therefore not required according to the recommendations of the International Workshop1), the patients were appropriately reclassified as CR by IWC+PET. All five patients remain progression free at a median of more than 32 months. The other two CRu designations by IWC patients were PET positive at the site of the residual mass and, hence were reclassified as PR by IWC+PET. One of the two patients experienced disease progression at 17 months post-therapy, whereas the other remains progression free at 36 months of follow-up.
The other major discordance was found in the PR by IWC designation, in which 10 of the 19 patients (52.6%) in this category were reclassified as CR by IWC+PET. Because all of the PR by IWC patients reclassified as CR by IWC+PET had radiologic abnormalities post-therapy, post-therapy BMB was not required based on the International Workshop recommendations. However, based on the IWC+PET classification, such patients can only be considered true CRs if there was no clinical or laboratory evidence of disease and the BMB was negative, if initially positive. None of the 10 patients had clinical or laboratory evidence of disease. Seven of 10 patients did not have evidence of bone marrow involvement at baseline. Two of three patients who had previous involvement had a negative post-therapy BMB, but one with residual hepatic CT abnormalities did not undergo a post-therapy BMB. This patient experienced disease relapse, however, 28.6 months post-therapy at the CT-positive post-therapy site (liver) and not in the marrow; hence, it is presumed that the patient’s marrow was negative post-therapy. Thus, all 10 patients with PR by IWC reclassified as CR by IWC+PET were appropriately classified as such according to the International Workshop guidelines for establishing a CR designation. All but the one patient who experienced relapse remain without evidence of disease progression at 26.4 to 38 months (median, 33.8 months) of follow-up; the estimated 3-year PFS rate in this group is 88.9%. In contrast, only two of the nine patients with concordant PR designation by IWC and IWC+PET are progression free at 26.8 and 53.1 months of follow-up, respectively. The other seven patients experienced disease progression at a median of 7 months post-therapy.
Table 4 lists the 2- and 3-year PFS rates in the various response designations by IWC and IWC+PET.
Of the 17 patients with a CR by IWC, four experienced disease progression at a median of 17.1 months (range, 8 to 31 months), and one of the seven patients with CRu experienced disease progression at 17 months. Eight of the 19 PR patients progressed at a median of 10.7 months, and six of the nine patients with SD progressed at a median of 4.2 months. Only one of the two PD patients progressed at 3.4 months.
Of the 35 patients with a CR by IWC+PET, six experienced disease progression at a median of 17.2 months (range, 8 to 31 months). Eight of the 12 patients with a PR experienced disease progression at a median of 8.5 months, and five of the six patients with SD experienced disease progression at a median of 4 months. The only patient with PD by IWC+PET (also PD by IWC) experienced disease progression at 3.4 months.
Median PFS of the 17 patients with a CR by IWC was identical to that of the 35 patients with a CR by IWC+PET (31.5+ months); the median PFS of the seven patients with CRu by IWC was similar (32.0+ months).
Figure 3 shows the Kaplan-Meier plots of PFS times for the patients with a CR and a PR by IWC and IWC+PET. In separate multivariate analyses with adjustment for age, disease stage, histologic type, chemotherapy type, and number of chemotherapy cycles, PFS time was significantly shorter in patients with a PR compared with those with a CR both based on the IWC and IWC+PET (P = .021 and P = .0003, respectively). The hazard of progression was 29.7-fold higher (95% CI, 4.7 to 189.5) for the PR than for the CR patients by IWC+PET compared with an HR of 8.9 (95% CI, 1.4 to 56.6) between the PR and CR patients by IWC. Moreover, when IWC (PR v CR) and IWC+PET (PR v CR) were included in the same multivariate model, the HR (PR v CR) for IWC was 1.6 (95% CI, 0.1 to 23.1; P = .72) compared with 35.9 (95% CI, 2.6 to 504.9; P = .008) for IWC+PET.
Because it is common to include the CRu values together with CR values when reporting response data in NHL using the IWC, and because our own data, albeit in a small number of patients, showed similar PFS times (Table 4) in patients with a CR by IWC and a CRu by IWC, we evaluated the differences between the PR and CR/CRu classifications. The Kaplan-Meier plots of PFS times for the PR and CR/CRu patients by IWC and IWC+PET are shown in Figure 4. The hazard of progression between PR by IWC and CR/CRu by IWC was 5.3 (95% CI, 1.2 to 24.7) compared with 29.7 (95% CI, 4.7 to 189.5) between PR by IWC+PET and CR/CRu by IWC+PET. Note that the latter ratio is identical to that between PR by IWC+PET and CR by IWC+PET because none of the patients had a CRu by IWC+PET. Here again, when IWC (PR v CR/CRu) and IWC+PET (PR v CR/CRu) were included in the same multivariate model, the HR for IWC was 1.0 (95% CI, 0.1 to 7.6; P = .97) compared with 48.8 (95% CI, 3.9 to 606.9; P = .003) for IWC+PET.
Finally, there was no significant difference in PFS between the 10 patients with a PR by IWC and a CR by IWC+PET versus the 17 patients with a CR by IWC and a CR by IWC+PET (HR, 1.6; 95% CI, 0.1 to 23.1; P = .72). In comparison, the hazard of progression for the nine patients with PR by IWC and PR by IWC+PET was 58.6 times that for the 17 patients with CR by IWC and CR by IWC+PET (P = .001; Fig 5) . This suggests that PET scanning identified two distinct subgroups of patients within the PR by IWC designation in terms of prognosis, one with a favorable and another with an unfavorable prognosis. The PR by IWC designation appears to be a mixture of these two subgroups with an intermediate prognosis (estimated 3-year PFS, 62.2% v 88.9% and 33.3% in the two subgroups, respectively).
DISCUSSION
The primary goal of this retrospective evaluation was to assess whether supplementing the IWC with PET information (IWC+PET) results in better discrimination in outcome between the various conventionally used response designations. As noted above, the decision to integrate PET with the IWC rather than to test its feasibility as a single modality for response assessment was based on considerations related to its recognized limitations, including the potential inability to characterize the nature of new lesions less than 1 to 1.5 cm in diameter and to reliably evaluate the status of bone marrow.
Perhaps the two most striking findings in our study are the substantially higher proportion of patients in the CR designation by IWC+PET compared with the IWC alone (35 of 54 v 17 of 54, respectively) and the higher HR between PR and CR or between PR and CR/CRu by IWC+PET compared with IWC. More importantly, when the two classifications were included in the same multivariate model, only IWC+PET was a statistically significant independent predictor for PFS (P = .008). The lower HR between the PR and CR designations by IWC is most likely related to the apparent existence of two distinct subgroups within the PR by IWC category with respect to outcome: a subgroup of PET-positive patients with poor outcome and another of PET-negative patients with excellent outcome (Fig 5). The biologic explanation of this finding is likely related to the fact that that the residual radiographic abnormalities represented necrosis and/or fibrosis in the vast majority of the PET-negative patients, whereas the abnormalities represented viable tumor in the majority of PET-positive patients.
A relatively small fraction of aggressive NHL treated at our institutions during the 8-year period could be included in this study. One obvious explanation for this fact is that not all patients treated underwent PET scanning after therapy. Unrestricted PET scanning was not approved for reimbursement in the United States until July 2001; PET was approved in July 1999 only when used as an alternative to gallium scanning. Many referring oncologists continued to request gallium scans until 2001 or even beyond because of their greater familiarity with gallium and abundant literature on its usefulness in the restaging of patients with lymphoma. Only in the last 3 years or so did PET begin to be more widely used in the restaging of lymphoma patients. All or some of these factors contributed to the fact that only a fraction of the treated patients underwent PET scanning at our institutions. Another explanation for the relatively small number of patients in our study is the relatively strict criteria for inclusion in the retrospective evaluation, resulting in exclusion of some patients who underwent PET scanning. For example, a substantial fraction of patients were excluded because of the requirement that both the PET and CT scans be performed within 4 weeks of each other; this exclusion was needed for a more rigorous comparison of PET and CT. However, it is important to note that all aggressive NHL patients treated at our institutions in the specified period who underwent PET scanning and met the entry criteria for the study were included.
One concern with a retrospective evaluation of IWC versus IWC+PET is that the performed post-therapy PET studies may be skewed toward patients demonstrating residual radiographic abnormalities by CT or other evidence of disease, with no or only few patients referred who had no clinical or radiographic evidence of disease. Because of the current lack of established guidelines for a standardized or routine use of PET, the decision for PET scanning was individualized for each patient. Yet, this practice did not seem to substantially affect the expected composition of CT-based response categories that is usually seen after CHOP-based chemotherapy in newly diagnosed aggressive NHL: although 69% of the 54 patients included in the analysis based on the entry criteria had residual radiographic abnormalities, 31% had no clinical or radiographic evidence of disease (ie, CRs by the IWC). Although somewhat low, this fraction is similar or close to the lower limits for CR rates expected for CHOP or rituximab+CHOP chemotherapy, for which the 95% CIs for CR rates range between 31% to 45% and 46% to 60%, respectively.30 In general, patients with CR by IWC underwent PET scanning because of residual lymph nodes of 1.1 to 1.5 cm in diameter at sites of previous nodes or nodal masses more than 1.5 cm in diameter. Although such patients are considered to have a CR by IWC (without clinical or biochemical evidence of disease and a negative BMB), the referring oncologist sought to confirm that these patients also had a metabolic/functional CR by PET. PET was also occasionally performed in patients with 1.0 cm residual nodes simply to establish this metabolic CR status. It is therefore important to note that none of these CR by IWC patients had a false-positive PET, indicating that the addition of PET to CT-based assessment does not result in misclassification of any appreciable fraction of such patients.
Although our study was performed without the use of the PET/CT systems, which combine a PET and a CT scanner in a single instrument, our study has obvious implications for PET/CT studies in response assessment of NHL. With PET/CT, the functional PET images are coregistered with the anatomic CT images obtained by the almost simultaneously acquired CT scan.31-34 This approach can result in a significant improvement in the diagnostic accuracy of PET, principally because the more accurate anatomic localization of the PET findings by coregistered CT leads to fewer false-positive PET interpretations due to interpatient variability in physiologic FDG uptake.33 Given that our data suggest that PET combined with visual correlation with a contrast-enhanced CT better predicts PFS compared with CT alone in this setting, it is likely that a truly integrated approach of fused PET and CT images will yield similar or superior results. In fact, a recently reported study in the restaging of 27 patients with lymphoma showed an identical accuracy of 96% for PET/CT and PET interpreted in visual correlation with contrast-enhanced CT, whereas contrast-enhanced CT alone had an accuracy of only 67%.32
In conclusion, response assessment using anatomic and functional imaging with CT and PET according to the IWC+PET classification appears to better discriminate between patients with a CR/CRu and a PR, with a larger difference in outcome between the response groups. Patients with PR or SD by IWC+PET should be evaluated for persistent disease and considered for additional therapy.
These results challenge the current paradigm in response assessment of patients with NHL, and require validation in a prospective trial with a larger number of patients.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the Lymphoma Specialized Programs of Research Excellence (SPORE) grant from the National Cancer Institute (CA972784) at the University of Iowa and the Mayo Clinic.
Presented in part at the American Society of Hematology Annual Meeting in San Diego, CA, December 6-9, 2003, and the 39th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Cheson BD, Horning SJ, Coiffier B, et al: Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas. J Clin Oncol 17:1244-1253, 1999
Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the International Criteria for Neuroblastoma Diagnosis, Staging, and Response to Treatment. J Clin Oncol 11:1466-1477, 1993
Fuks JZ, Aisner J, Wiernik PH: Restaging laparotomy in the management of the non-Hodgkin's lymphomas. Med Pediatr Oncol 10:429-438, 1982
Stewart FM, Williamson BR, Innes DJ, et al: Residual tumor masses following treatment for advanced histiocytic lymphoma. Cancer 55:620-623, 1985
Surbone A, Longo DL, DeVita VT Jr, et al: Residual abdominal masses in aggressive non-Hodgkin's lymphoma after combination chemotherapy: Significance and management. J Clin Oncol 6:1832-1837, 1988
Coiffier B, Gisselbrecht C, Herbrecht R, et al: LNH-84 regimen: A multicenter study of intensive chemotherapy in 737 patients with aggressive malignant lymphoma. J Clin Oncol 7:1018-1026, 1989
Canellos GP: Residual mass in lymphoma may not be residual disease. J Clin Oncol 6:931-933, 1988
Kaplan WD, Jochelson MS, Herman TS, et al: Gallium-67 imaging: A predictor of residual tumor viability and clinical outcome in patients with diffuse large-cell lymphoma. J Clin Oncol 8:1966-1970, 1990
Front D, Bar-Shalom R, Mor M, et al: Aggressive non-Hodgkin lymphoma: Early prediction of outcome with 67Ga scintigraphy. Radiology 214:253-257, 2000
Vose JM, Bierman PJ, Anderson JR, et al: Single-photon emission computed tomography gallium imaging versus computed tomography: Predictive value in patients undergoing high-dose chemotherapy and autologous stem-cell transplantation for non-Hodgkin's lymphoma. J Clin Oncol 14:2473-2479, 1996
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 94:429-433, 1999
Spaepen K, Stroobants S, Dupont P, et al: Prognostic value of positron emission tomography (PET) with fluorine 18 fluorodeoxyglucose ([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: Is [18F]FDG-PET a valid alternative to conventional diagnostic methods J Clin Oncol 19:414-419, 2001
Cremerius U, Fabry U, Neuerburg J, et al: Positron emission tomography with 18F-FDG to detect residual disease after therapy for malignant lymphoma. Nucl Med Commun 19:1055-1063, 1998
Rmer W, Hanauske A-R, Ziegler S, et al: Positron emission tomography in non-Hodgkin's lymphoma: Assessment of chemotherapy with Fluorodeoxyglucose. Blood 91:4464-4471, 1998
Kostakoglu L, Goldsmith SJ: Fluorine-18 fluorodeoxyglucose positron emission tomography in the staging and follow-up of lymphoma: Is it time to shift gears Eur J Nucl Med 27:1564-1578, 2000
Juweid M, Wiseman G, Menda Y, et al: FDG-PET predicts with high accuracy the 1-year progression-free survival (1-year PFS) of patients with aggressive non-Hodgkin's lymphoma (NHL) following anthracycline-based first-line chemotherapy. J Nucl Med 43:124P, 2002 (abstr 283)
Pieterman R, van Putten JWG, Meuzelaar JJ, et al: Preoperative staging of non-small cell lung cancer with positron-emission tomography. N Engl J Med 343:254-261, 2000
Newman JS, Francis IR, Kaminski MS, et al: Imaging of lymphoma with PET with 2-[18F] fluoro-2-deoxy-D-glucose: Correlation with CT. Radiology 190:111-116, 1994
Avril N, Rose CA, Schelling M, et al: Breast imaging with positron emission tomography and fluorine-18-fluorodeoxyglucose: Use and limitations. J Clin Oncol 18:3495-3502, 2000
Tyler DS, Onaitis M, Kherani A, et al: Positron emission tomography scanning in malignant melanoma. Cancer 89:1019-1025, 2000
Lowe VJ, Hoffman JM, Delong DM, et al: Semi-quantitative and visual analysis of FDG-PET images in pulmonary abnormalities. J Nucl Med 35:1771-1776, 1994
Juweid ME, Wiseman GA, Menda Y, et al: Integrated PET based response classification for non-Hodgkin's lymphoma (NH). J Clin Oncol 22:565S, 2004 (abstr 6533)
Juweid ME, Wiseman G, Menda Y, et al: Integrated PET plus CT-based response assessment of aggressive NHL. Blood 102:629a, 2003 (abstr 2321)
Kostakoglu L, Leonard JP, Coleman M, et al: Evaluation of FDG-ET in the detection of bone marrow involvement in lymphoma. J Nucl Med 41:71P, 2000 (abstr 279)
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, NY, John Wiley & Sons, 1980
Cox DR: Regression models and life tables (with discussion). J R Stat Soc B 34:187-220, 1972
Allison PD: Survival Analysis Using the SAS System: A Practical Guide. Cary, NC, SAS Institute Inc, 1995
Shipp MA, Harrington DP, Anderson JR, et al: A predictive model for aggressive non-Hodgkin's lymphoma. N Engl J Med 329:987-994, 1993
Coiffier B, Lepage E, Briere J, et al: CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med 346:235-242, 2002
Lardinois D, Weder W, Hany F, et al: Staging of non-small-cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 348:2500-2507, 2003
Freudenberg LS, Antoch G, Schutt P, et al: FDG-PET/CT in re-staging of patients with lymphoma. Eur J Nucl Med Mol Imaging 31:325-329, 2004
Allen-Auerbach M, Quon A, Weber WA, et al: Comparison between 2-deoxy-2-[(18)F]fluoro-D-glucose positron-emission tomography and positron-emission tomography/computed tomography hardware fusion for staging of patients with lymphoma. Mol Imaging Biol 6:411-416, 2004
Goerres GW, von Schulthes GK, Steinert HC: Why most of lung and head and neck cancer will be PET/CT. J Nucl Med 45:66S-71S, 2004(Malik E. Juweid, Gregory )