Intensity of 18Fluorodeoxyglucose Uptake in Positron Emission Tomography Distinguishes Between Indolent and Aggressive Non-Hodgkin’s Lymphom
http://www.100md.com
《临床肿瘤学》
the Department of Radiology, Nuclear Medicine Service
Department of Medicine, Lymphoma Service
Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY
ABSTRACT
PURPOSE: 18Fluorodeoxyglucose positron emission tomography (FDG PET) is widely used for the staging of lymphoma. We investigated whether the intensity of tumor FDG uptake could differentiate between indolent and aggressive disease.
MATERIALS AND METHODS: PET studies of 97 patients with non-Hodgkin's lymphoma who were untreated or had relapsed and/or persistent disease and had not received treatment within the last 6 months were analyzed, and the highest standardized uptake value (SUV) per study was recorded. Correlations were made with histopathology.
RESULTS: FDG uptake was lower in indolent than in aggressive lymphoma for patients with new (SUV, 7.0 ± 3.1 v 19.6 ± 9.3; P < .01) and relapsed (SUV, 6.3 ± 2.7 v 18.1 ± 10.9; P = .04) disease. Despite overlap between indolent and aggressive disease in the low SUV range (indolent, 2.3 to 13.0; aggressive, 3.2 to 43.0), all cases of indolent lymphoma had an SUV 13. A receiver operating characteristic (ROC) analysis demonstrated that the SUV distinguished reasonably well between aggressive and indolent disease (area under ROC curve, 84.7%), and an SUV > 10 excluded indolent lymphoma with a specificity of 81%. With a higher cutoff for the SUV, the specificity would have been higher.
CONCLUSION: FDG uptake is lower in indolent than in aggressive lymphoma. Patients with NHL and SUV > 10 have a high likelihood for aggressive disease. This information may be helpful if there is discordance between biopsy and clinical behavior.
INTRODUCTION
Positron emission tomography (PET) with the glucose analog 18fluorodeoxyglucose (FDG) is now a clinically accepted and widely used imaging tool for the staging and treatment evaluation of non-Hodgkin’s lymphoma (NHL).1-6 The usefulness of FDG PET for the detection of recurrent disease has also been demonstrated.7 FDG PET can assess the extent of lymphoma with higher sensitivity than anatomic imaging modalities (ie, computed tomography [CT] and magnetic resonance imaging), frequently resulting in upstaging and change of patient management, affecting between 8% and 42% of patients.1,8-10
While FDG PET is clinically useful in a wide variety of lymphomas,11 there has been no demonstration that PET imaging can distinguish between different subgroups of lymphoma.12-14 If the probability for indolent versus aggressive lymphoma could be predicted noninvasively, PET could potentially provide a clinically important metabolic biopsy tool. For instance, high FDG uptake in a patient with biopsy-proven indolent lymphoma would support either a redirected biopsy or a more intensive treatment. Accordingly, this study evaluated whether FDG PET can reliably distinguish between different WHO subtypes of lymphoma in a large sample of patients.
PATIENTS AND METHODS
Patient Population
All FDG PET studies done in patients with NHL between April 1998 and July 2002 were evaluated retrospectively. Database and retrospective analysis were approved by the institutional review board. A total of 97 patients met the following entry criteria: PET before any therapy or at relapse more than 6 months after chemo-, immuno- or radiotherapy; histopathologic review at our institution; no diagnosis of transformed lymphoma; no other diagnosis of cancer within 5 years preceding the current PET exam; and time interval of less than 90 days between PET and histopathology.
Patients with a diagnosis of more than one histologically proven subtype of NHL or coexistence of Hodgkin’s disease and NHL were excluded from the initial analysis. A group of patients with discordant pathologies was subsequently evaluated separately.
The clinical stage of the disease was determined using the Ann Arbor classification.15
PET
Patients were injected with 12 to 15 mCi (444 to 555 MBq) of FDG intravenously. Images were acquired from the level of the skull base to the inguinal regions. The average time interval between FDG injection and image acquisition was 61minutes (standard deviation [SD], 24 minutes; range, 40 to 130 minutes; all but six studies were performed between 40 and 80 minutes postinjection). Plasma glucose measurements were routinely obtained, and ranged from 53 to 222 mg/dL (including one diabetic patient with plasma glucose > 140 mg/dL), with a mean of 85 mg/dL (SD, 18 mg/dL) for the entire study population.
Studies were performed using the GE Advance (GE Medical Systems, Waukesha, WI) or HR plus (Siemens/CTI, Knoxville, TN) PETs. The GE Advance has a transaxial spatial resolution of 4.9, 5.3, and 6.1 mm at 1, 10, and 20 cm off axis, respectively. The slice thickness or axial resolution is 4.2 mm at the center. The HR plus has a transaxial spatial resolution of 4.5 mm full width at half maximum (FWHM) at 1 cm off axis and 5.4 mm at 10 cm off axis. The axial resolution is 4.2 mm at the center of the system. The following acquisition parameters were used: Advance, 3 minutes transmission, 4 minutes emission per bed position (15.2 cm); HR plus, 7 to 8 minutes per bed position (14.2 cm), with 35% of this time for the transmission scan. On both tomographs, attenuation correction was performed using Ge-68 transmission rods (GE Medical Systems).
Since November of 2001, studies were also acquired on combined PET/CT tomographs (either Biograph [Siemens/CTI, Nashville,TN] or DiscoveryLS [GE Medical Systems, Waukesha, WI]). Both PET/CT tomographs consist of a combination of a spiral CT tomograph with a PET tomograph. The Discovery uses the Advance PET tomograph, and the Biograph uses the HR plus PET tomograph. With this equipment, a low-dose CT scan, which is used for attenuation correction of PET emission images as well as for anatomic localization of PET abnormalities, is acquired first using the following parameters: for the Biograph, scout view with 130 kVp and 30 mAs, followed by spiral CT using effective mAs of 50, 130 kVp, scan width 5 mm, collimation of 4 mm and a feed/rotation of 12 mm; for the Discovery LS, scout view with 30 mAs, 120 kVp, followed by a spiral CT at 0.8 seconds/rotation using 80 mAs, 140 kVp, a slice thickness of 5 mm, and a 4.25 mm interval in high sensitivity mode. PET emission images were acquired for 4 minutes per bed position. The total acquisition time varied between 25 and 35 minutes per patient.
Regardless of the PET tomograph used, PET images were reconstructed using iterative algorithms. Attenuation correction was routinely applied. PET images of patients who met the above entry criteria were reviewed by one of the nuclear medicine physicians (H.S., L.W., or H.W.Y.). Abnormal FDG uptake was defined as greater than background activity in surrounding tissue and unrelated to physiologic sites of tracer uptake (such as the myocardium) or excretion (such as excreted FDG in the ureters and urinary bladder). For the spleen, abnormal uptake was defined as FDG accumulation greater than in the liver, based on standardized uptake value (SUV) measurements. Areas of abnormal FDG uptake were identified, and the intensity of FDG uptake was quantified by calculating the SUV. For the calculation of SUV, circular regions of interest ( 70 pixels) were drawn on transaxial images around the areas with increased FDG uptake. The SUV was calculated as:
To minimize partial volume effects and assure reproducibility of measurements, the maximum SUV (SUVmax) in a large lesion was used. The highest SUV per patient was recorded and used for further analysis. To assess for potential loss of count recovery, lesion size was measured as described previously.16
Histologic Analysis and Correlation With PET Findings
At least one lymph node biopsy was obtained per patient. All tissue specimens were reviewed by experienced hematopathologists at our institution. The recent WHO classification for NHL was used.17,18 Accordingly, patients were then grouped as having indolent or aggressive lymphoma. Mantle-cell lymphoma could not be classified into one of these groups. In some cases where the WHO classification could not be applied (eg, in cases with small cleaved cell lymphoma), the histopathologic description was recorded.
Statistical Analysis
SUVs were analyzed considering the various histologies and grades, and separately for newly diagnosed and recurrent disease. Two-sample comparisons were performed using the Wilcoxon test. A P value less than .05 was considered significant.
The accuracy of PET scans was evaluated in two different ways: one using the interpretation of the radiologist (positive/negative) and the other using the SUV. Because the nuclear medicine physician’s interpretation was dichotomous, sensitivity (true-positive fraction) and specificity (true-negative fraction) were computed using histopathology as the gold standard19.
To evaluate the accuracy of SUV, which is a continuous variable, we used receiver operating characteristic (ROC) curve analysis,19 again using histopathologic findings as the gold standard. The area under the curve was computed using the trapezoidal rule and used as a summary measure of accuracy. This analysis for SUV used only the subset of aggressive or indolent NHL patients for whom the site of biopsy was the same as the site at which SUVmax was obtained.
RESULTS
The study population (Table 1) consisted of 97 individuals; there were 51 males and 46 females (mean age, 59 years; SD, 15 years; range, 15 to 86 years). Sixty-eight patients presented with newly diagnosed disease. Twenty-nine patients had been treated previously and presented with relapsed or persistent disease. Prior treatment modalities included chemotherapy, immunotherapy and radiation.
Among patients with aggressive NHL (n = 63), the following specific histologies were identified: 55 diffuse large-cell lymphoma (53 of 55 patients with B-cell type); seven follicular lymphoma, grade 3; and one peripheral T-cell lymphoma.
Among patients with indolent NHL (n = 28), the following histologies were identified: one diffuse small cleaved; one unspecified follicular lymphoma; 10 follicular lymphoma, grade 1; four follicular lymphoma, grade 2; one lymphocytic plasmocytic lymphoma; four marginal zone lymphoma; one small cleaved-cell lymphoma; one small lymphocytic lymphoma; and five small-cell lymphoma.
Six additional patients presented with mantle cell lymphoma. These were analyzed separately as the nature of this disease does not fit the categories of indolent or aggressive disease and varies widely among patients.
Intensity of FDG Uptake
SUV in patients with NHL ranged from 3.2 to 43.0. In patients with indolent lymphoma, the SUV ranged from 2.3 to 13.0 (mean, 6.7; SD, 2.9). In patients with aggressive lymphoma, SUV ranged from 3.2 to 43.0 (mean, 17.2; SD, 9.7; P < .01 v indolent). In patients with mantle cell lymphoma, the SUV ranged from 4.7 to 8.7, with a mean of 6.4 (SD, 1.7). The intensity of FDG uptake in the various histologic subgroups is listed in Table 2, and patient examples are shown in Figures 1 and 2. The diameter of the lesion used for SUV analysis ranged from 1.0 to 18 cm.
Box plots for indolent and aggressive NHL are shown in Figure 3. FDG uptake in aggressive disease was significantly higher than in indolent disease (P < .01). A high FDG uptake was suggestive of aggressive disease since all of the indolent lymphomas showed an SUV 13. On the other hand, a low FDG uptake was not always an indicator of indolent disease: of the 63 patients with aggressive disease, 22 (35%) had lesions with an SUV < 13.
To further investigate whether SUV can distinguish between aggressive and indolent lymphoma, we used ROC curve analysis. The primary ROC analysis was restricted to the 69 patients for whom the site of biopsy was the same as the site at which SUVmax was measured. The area under the ROC curve was 84.7% (SD, 4%), which suggests that the SUV is a reasonable predictor of diagnosis (Fig 4). The various sensitivity and specificities that were obtained as the cutoff value for the SUV was varied are shown in Table 3. An SUV of 10 provided a fair balance, with a specificity of 81% and a sensitivity of 71% to detect an aggressive lymphoma. A higher cutoff would capture more indolent lymphomas. Conversely, a lower cutoff would capture a higher percentage of aggressive lymphomas.
In 22 patients (11 with indolent and 11 with aggressive lymphoma) the biopsy was not the site of highest SUV in the PET scan (mean difference between highest SUV and SUV at biopsy site, 40%; range, 5% to 60%; Table 4). Using an SUV of 10 as the cutoff to predict aggressive NHL, two patients with biopsy-proven indolent lymphoma were misclassified as having aggressive lymphoma. Similarly, six cases of aggressive lymphoma would have been misclassified as indolent lymphoma. The latter reflects the overlapping SUVs in the lower range.
Patients With Discordant Histologies
There were 12 patients with biopsies demonstrating discordant lymphoma pathologies at separate time points (Table 5) . Eight patients had previous diagnoses of indolent lymphomas (n = 7) or Hodgkin’s disease (n = 1), but at the time of the PET scan had biopsy proven aggressive lymphomas. There were four additional patients with previous diagnoses of aggressive lymphomas who had biopsy-proven indolent lymphomas at the time of the PET scan. PET imaging was not available at the time of the original diagnosis in any of these patients.
For the eight patients with transformed lymphoma, SUVs ranged from 4.8 to 29.8. Five of the eight patients had an SUV > 10: one patient had an SUV of 10.2, one 18.8, and three others > 25.
Among four patients with previous aggressive lymphoma and indolent disease at the time of the PET scan, two had an SUV < 10, whereas the other two patients had SUVs of 14.6 and 15.7, respectively. The patient with an SUV of 14.6 had clinically aggressive disease with five prior therapy regimens and clinical deterioration secondary to lymphoma after the PET study. For the patient with an SUV of 15.7, the biopsy was performed before the PET scan and came from a region where the SUV was only 6.9. It is not possible to rule out that the area with an SUV of 15.7 had a higher-grade lymphoma.
DISCUSSION
This study demonstrates that FDG PET imaging can distinguish between patients with indolent and aggressive lymphoma, and that the likelihood for aggressive disease increases in parallel with increases in SUV. This is in concordance with PET studies in a number of other malignancies demonstrating that higher FDG uptake in a given malignant tumor indicates a more aggressive course of the disease and is frequently associated with poor prognosis.20-26
Previous studies have demonstrated a considerable overlap in the intensity of FDG uptake between patients with indolent and aggressive lymphoma in the low SUV range.12-14 While we also noted such an overlap in the lower SUV range, the current study nevertheless demonstrates that the intensity of FDG uptake is generally lower in indolent as compared with aggressive non-Hodgkin’s lymphoma and that the likelihood for aggressive disease increases in parallel with increases in SUV. An SUV < 10 captured 81% of indolent lymphomas, while an SUV 13 was virtually indicative of aggressive disease.
Our results suggest that SUV may alter clinical management. For example, if a patient with previously documented indolent disease had a PET scan showing unexpectedly high SUVs, a transformation to more aggressive disease might be suspected and a biopsy considered. This would be particularly true in cases with very high SUV (> 13). Transformation occurs in roughly 25% of patients with indolent follicular lymphoma, is unpredictable, and does not necessarily involve all sites of known disease.27 Our series of patients with transformation is small, but 5 of 8 patients would have had PET scans suggestive of transformation if an SUV cutoff of 10 had been applied. In contrast, a low SUV in patients with previously aggressive disease is somewhat inconclusive because of the overlap between aggressive and indolent histologies in the lower SUV range.
PET scanning may also be helpful in directing biopsies. Although typically biopsies are obtained from areas that are easily accessible, an unexpectedly high SUV may suggest a transformation to aggressive lymphoma and a biopsy of the site with the highest SUV may be considered.
This is the largest study to date evaluating the relationship between histopathology and intensity of FDG uptake in lymphoma patients. Some previous smaller studies suggested that the intensity of FDG uptake in lymphoma correlated with the histologic grade of malignancy.12,13,28 Rodriguez et al12 studied 23 patients with non-Hodgkin’s lymphoma (11 high grade, nine low grade: three transformed, respectively) with FDG PET. Three different parameters for the assessment of tumor FDG uptake (SUV, mass influx, and transport rate) were significantly higher for high-grade than for low-grade lymphoma. However, only the transport rate appeared to distinguish between low- and high-grade disease, whereas tumor SUV (15.9 ± 7.7 v 8.7 ± 6.6) and mass influx showed a marked overlap. It should be noted that PET measurement of glucose transport rates is a cumbersome and time-consuming research technique directed at one field of view, usually 15 cm of the torso. Lapela et al13 studied 22 patients with untreated NHL and FDG PET: seven with low-grade lymphoma, 11 with intermediate grade, and four with high grade. SUV and metabolic rate of glucose were analyzed in one lesion per patient. The median SUV for all patients was 8.5 (range, 3.5 to 31). SUV for low-grade lymphoma ranged from 4 to 10; for intermediate grade, 3 to 25; and for high grade, 12 to 31. Differences in SUV were significant, but there was again considerable overlap between subgroups.
Our findings are in concordance with data reported in these earlier studies. However, by using ROC analysis we determined meaningful cutoffs above which indolent lymphomas are highly unlikely.
There is evidence that indolent lymphoma is not always detectable by FDG PET.10,11,28-30 However, a wide range in detection rates (40% to 98%) has been reported. Varying intensity of FDG uptake in lymphomas of different histologic subtypes has been cited as one explanation. For example, while Jerusalem et al10 found that FDG PET showed 40% more nodal lesions than did conventional imaging in patients with follicular lymphoma, only 58% of abnormal lymph nodes by CT size criteria were detected in patients with small lymphocytic lymphoma. Other studies showed lower PET detection rates for extranodal marginal zone lymphoma (MZL) and peripheral T-cell lymphoma.11,29 In one recent study, only eight of 12 MZLs (67%) were detected by PET using an SUV of 2.5 as the cutoff.11 In contrast, all four MZLs were visualized in our study, possibly as a result of improved PET technology or referral bias. Indeed, there could have been an overall referral bias to obtain a PET scan in younger patients or those with more clinically aggressive disease. While a high detection rate for indolent lymphoma was also noted in a recent study from Australia,30 a prospective study may be needed to define the ultimate role of PET in these patients.
Only the highest SUV per patient was used in the analysis. Variability in SUV from lesion to lesion in any particular patient is not addressed by the current investigation. Although lesions in patients with aggressive disease may show variability in SUV numbers, inclusion of all lesions with FDG uptake would likely blur the differences between histologic subgroups of lymphoma. In addition, lesions with the highest FDG uptake likely represent sites of the most aggressive disease. For classification into indolent or aggressive disease it would be important not to underestimate the metabolic activity and aggressiveness of the disease. Moreover, because lesions with higher tracer accumulation tend to be larger in size (data not shown), these SUV measurements are less prone to partial volume effects that may cause underestimation of the true activity concentration.
In some patients, the location of highest SUV was discordant from the location of biopsy. For practical purposes, biopsies are generally obtained from lymph nodes and tissues that are easily accessible. It is then generally assumed that the tissue obtained is representative of the disease elsewhere. However, our study suggests in patients with indolent NHL, in whom transformation is suspected, a repeat biopsy of an area with high SUV may be indicated in the appropriate setting.
The SUV is a relative measure of local radiotracer accumulation in tissue. It varies with biologic factors,31 the method of analysis,32 and PET image reconstruction parameters.33 Our results, therefore, only apply when the uptake period between FDG injection and imaging is at least 40 minutes and when attenuation correction and iterative reconstruction are used for generating the PET images.
In lesions smaller than twice the resolution (FWHM) of the PET tomograph, the activity concentration will be underestimated. Although the spatial resolution of current dedicated PET tomographs is quoted as between 4 and 5 mm, in our clinical setting (considering matrix and reconstruction parameters), the spatial resolution of PET tomographs used is approximately 7 to 8 mm FWHM. In lesions smaller than 2x FWHM (ie, 1.5 cm diameter) the true activity concentration (and hence, the SUV) in this lesion will therefore be underestimated. The reason for this is a loss of count recovery. However, such lack of count recovery did not affect the overall results in this study. Although 11 lesions had a minimum diameter of 1.5 cm, this occurred to a similar degree in group with indolent lymphoma (4 of 28 lesions; SUV range, 3.2 to 5.5) and the group with aggressive lymphoma (7 of 63 lesions; SUV range, 5.5 to 10.0).
Studies were acquired with different types of PET tomographs, reflecting our clinical practice over the past several years. In phantom studies (unpublished results) we have noted an approximately 10% difference in quantitative measurements between Advance and Biograph tomographs. However, in patient studies we did not observe any significant difference between SUV measurements obtained with different PET tomographs for patients with either aggressive or indolent disease. It should also be noted that technical and patient-related factors could potentially influence SUV measurements to a greater degree than the observed 10% difference between tomographs. In clinical practice it may be difficult to standardize all potential parameters that can affect SUV measurements. In addition, our study shows overlap of data in the lower SUV range. Overall, it is therefore impossible to define one specific SUV number that will distinguish with 100% accuracy between aggressive and indolent lymphoma. Rather, our data show that the likelihood for aggressive disease increases in parallel with rises in SUV, and an SUV > 13 was virtually indicative of aggressive disease.
Although our study showed higher FDG uptake in aggressive lymphoma as a group, a wide range of SUVs was observed among both patients with an indolent and aggressive subtype of the disease. Future studies will have to address whether FDG PET can identify subsets of patients with better or worse clinical course and prognosis among patients with aggressive disease. Future studies will also have to investigate whether imaging with other radiotracers, such as the proliferation marker 18F fluoro-thymidine, have superior predictive value.
Regardless of the findings in this study, biopsy remains the standard procedure for establishing an unequivocal diagnosis in patients with lymphoma. However, in selected cases where biopsies from easily accessible peripheral lymph nodes are inconsistent with the clinical course of the disease, or where the PET scan shows additional lesions with significantly higher metabolic activity, an effort should be made to obtain additional biopsies from these sites of intense FDG uptake on PET. In unique situations, it may be advisable to treat for aggressive disease if a biopsy cannot be safely obtained.
The intensity of FDG uptake on PET images correlates with tumor aggressiveness in NHL. Low intensity of FDG uptake can be observed in both indolent and aggressive disease, but increasingly higher SUV have a higher specificity for the detection of aggressive disease. Although FDG PET alone cannot reliably diagnose transformation from indolent to aggressive lymphoma, PET findings may be useful to guide biopsies of lesions with the highest SUV when clinically appropriate.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Hoh CK, Glaspy J, Rosen P, et al: Whole-body FDG-PET imaging for staging of Hodgkin's disease and lymphoma. J Nucl Med 38:343-348, 1997
Carr R, Barrington SF, Madan B, et al: Detection of lymphoma in bone marrow by whole-body positron emission tomography. Blood 91:3340-3346, 1998
Moog F, Bangerter M, Diederichs CG, et al: Extranodal malignant lymphoma: Detection with FDG PET versus CT. Radiology 206:475-481, 1998
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
Buchmann I, Reinhardt M, Elsner K, et al: 2-(fluorine-18)fluoro-2-deoxy-D-glucose positron emission tomography in the detection and staging of malignant lymphoma: A bicenter trial. Cancer 91:889-899, 2001
Romer W, Hanauske AR, Ziegler S, et al: Positron emission tomography in non-Hodgkin's lymphoma: Assessment of chemotherapy with fluorodeoxyglucose. Blood 91:4464-4471, 1998
Spaepen K, Stroobants S, Dupont P, et al: Early restaging positron emission tomography with (18)F-fluorodeoxyglucose predicts outcome in patients with aggressive non-Hodgkin's lymphoma. Ann Oncol 13:1356-1363, 2002
Moog F, Bangerter M, Kotzerke J, et al: 18-F-fluorodeoxyglucose-positron emission tomography as a new approach to detect lymphomatous bone marrow. J Clin Oncol 16:603-609, 1998
Schoder H, Meta J, Yap C, et al: Effect of whole-body (18)F-FDG PET imaging on clinical staging and management of patients with malignant lymphoma. J Nucl Med 42:1139-1143, 2001
Jerusalem G, Beguin Y, Najjar F, et al: Positron emission tomography (PET) with 18F-fluorodeoxyglucose (18F-FDG) for the staging of low-grade non-Hodgkin's lymphoma (NHL). Ann Oncol 12:825-830, 2001
Elstrom R, Guan L, Baker G, et al: Utility of FDG-PET scanning in lymphoma by WHO classification. Blood 101:3875-3876, 2003
Rodriguez M, Rehn S, Ahlstrom H, et al: Predicting malignancy grade with PET in non-Hodgkin's lymphoma. J Nucl Med 36:1790-1796, 1995
Lapela M, Leskinen S, Minn HR, et al: Increased glucose metabolism in untreated non-Hodgkin's lymphoma: A study with positron emission tomography and fluorine-18-fluorodeoxyglucose. Blood 86:3522-3527, 1995
Filmont J, Ko Yang F, Yap C, et al: Prediction of malignancy grade using 18FDG PET in Non Hodgkin's lymphoma patients: An urban legend J Nucl Med 43:79, 2002 (abstr)
Rosenberg S: Validity of the Ann Arbor staging classification for the non-Hodgkin's lymphomas. Cancer Treat Rep 61:1023-1027, 1977
Erdi YE, Mawlawi O, Larson SM, et al: Segmentation of lung lesion volume by adaptive positron emission tomography image thresholding. Cancer 80:2505-2509, 1997
Harris NL, Jaffe ES, Diebold J, et al: Lymphoma classification—From controversy to consensus: The R.E.A.L. and WHO Classification of lymphoid neoplasms. Ann Oncol 11:3-10, 2000
Harris NL, Jaffe ES, Diebold J, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee meeting, Airlie House, Virginia, November 1997. J Clin Oncol 17:3835-3849, 1999
Pepe M: The statistical evaluation of medical tests for classification and prediction. Oxford, England Oxford University Press, 2003, pp 66-94
Ahuja V, Coleman RE, Herndon J, et al: The prognostic significance of fluorodeoxyglucose positron emission tomography imaging for patients with nonsmall cell lung carcinoma. Cancer 83:918-924, 1998
Wong RJ, Lin DT, Schoder H, et al: Diagnostic and prognostic value of [(18)F]fluorodeoxyglucose positron emission tomography for recurrent head and neck squamous cell carcinoma. J Clin Oncol 20:4199-4208, 2002
Eary JF, O'Sullivan F, Powitan Y, et al: Sarcoma tumor FDG uptake measured by PET and patient outcome: A retrospective analysis. Eur J Nucl Med Mol Imaging 29:1149-1154, 2002
Vansteenkiste JF, Stroobants SG, Dupont PJ, et al: Prognostic importance of the standardized uptake value on (18)F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: An analysis of 125 cases: Leuven Lung Cancer Group. J Clin Oncol 17:3201-3206, 1999
Fukunaga T, Okazumi S, Koide Y, et al: Evaluation of esophageal cancers using fluorine-18-fluorodeoxyglucose PET. J Nucl Med 39:1002-1007, 1998
Oshida M, Uno K, Suzuki M, et al: Predicting the prognoses of breast carcinoma patients with positron emission tomography using 2-deoxy-2-fluoro[18F]-D-glucose. Cancer 82:2227-2234, 1998
Nakata B, Chung YS, Nishimura S, et al: 18F-fluorodeoxyglucose positron emission tomography and the prognosis of patients with pancreatic adenocarcinoma. Cancer 79:695-699, 1997
Bastion Y, Sebban C, Berger F, et al: Incidence, predictive factors, and outcome of lymphoma transformation in follicular lymphoma patients. J Clin Oncol 15:1587-1594, 1997
Leskinen-Kallio S, Ruotsalainen U, Nagren K, et al: Uptake of carbon-11-methionine and fluorodeoxyglucose in non-Hodgkin's lymphoma: A PET study. J Nucl Med 32:1211-1218, 1991
Hoffmann M, Kletter K, Becherer A, et al: 18F-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) for staging and follow-up of marginal zone B-cell lymphoma. Oncology 64:336-340, 2003
Blum RH, Seymour JF, Wirth A, et al: Frequent impact of [18F]fluorodeoxyglucose positron emission tomography on the staging and management of patients with indolent non-Hodgkin's lymphoma. Clin Lymphoma 4:43-49, 2003
Huang SC: Anatomy of SUV: Standardized uptake value. Nucl Med Biol 27:643-646, 2000
Yeung HW, Sanches A, Squire OD, et al: Standardized uptake value in pediatric patients: An investigation to determine the optimum measurement parameter. Eur J Nucl Med Mol Imaging 29:61-66, 2002
Schoder H, Erdi YE, Chao K, et al: Clinical implications of different image reconstruction parameters for interpretation of whole-body PET studies in cancer patients. J Nucl Med 45:559-566, 2004(Heiko Schder, Ariela Noy,)
Department of Medicine, Lymphoma Service
Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY
ABSTRACT
PURPOSE: 18Fluorodeoxyglucose positron emission tomography (FDG PET) is widely used for the staging of lymphoma. We investigated whether the intensity of tumor FDG uptake could differentiate between indolent and aggressive disease.
MATERIALS AND METHODS: PET studies of 97 patients with non-Hodgkin's lymphoma who were untreated or had relapsed and/or persistent disease and had not received treatment within the last 6 months were analyzed, and the highest standardized uptake value (SUV) per study was recorded. Correlations were made with histopathology.
RESULTS: FDG uptake was lower in indolent than in aggressive lymphoma for patients with new (SUV, 7.0 ± 3.1 v 19.6 ± 9.3; P < .01) and relapsed (SUV, 6.3 ± 2.7 v 18.1 ± 10.9; P = .04) disease. Despite overlap between indolent and aggressive disease in the low SUV range (indolent, 2.3 to 13.0; aggressive, 3.2 to 43.0), all cases of indolent lymphoma had an SUV 13. A receiver operating characteristic (ROC) analysis demonstrated that the SUV distinguished reasonably well between aggressive and indolent disease (area under ROC curve, 84.7%), and an SUV > 10 excluded indolent lymphoma with a specificity of 81%. With a higher cutoff for the SUV, the specificity would have been higher.
CONCLUSION: FDG uptake is lower in indolent than in aggressive lymphoma. Patients with NHL and SUV > 10 have a high likelihood for aggressive disease. This information may be helpful if there is discordance between biopsy and clinical behavior.
INTRODUCTION
Positron emission tomography (PET) with the glucose analog 18fluorodeoxyglucose (FDG) is now a clinically accepted and widely used imaging tool for the staging and treatment evaluation of non-Hodgkin’s lymphoma (NHL).1-6 The usefulness of FDG PET for the detection of recurrent disease has also been demonstrated.7 FDG PET can assess the extent of lymphoma with higher sensitivity than anatomic imaging modalities (ie, computed tomography [CT] and magnetic resonance imaging), frequently resulting in upstaging and change of patient management, affecting between 8% and 42% of patients.1,8-10
While FDG PET is clinically useful in a wide variety of lymphomas,11 there has been no demonstration that PET imaging can distinguish between different subgroups of lymphoma.12-14 If the probability for indolent versus aggressive lymphoma could be predicted noninvasively, PET could potentially provide a clinically important metabolic biopsy tool. For instance, high FDG uptake in a patient with biopsy-proven indolent lymphoma would support either a redirected biopsy or a more intensive treatment. Accordingly, this study evaluated whether FDG PET can reliably distinguish between different WHO subtypes of lymphoma in a large sample of patients.
PATIENTS AND METHODS
Patient Population
All FDG PET studies done in patients with NHL between April 1998 and July 2002 were evaluated retrospectively. Database and retrospective analysis were approved by the institutional review board. A total of 97 patients met the following entry criteria: PET before any therapy or at relapse more than 6 months after chemo-, immuno- or radiotherapy; histopathologic review at our institution; no diagnosis of transformed lymphoma; no other diagnosis of cancer within 5 years preceding the current PET exam; and time interval of less than 90 days between PET and histopathology.
Patients with a diagnosis of more than one histologically proven subtype of NHL or coexistence of Hodgkin’s disease and NHL were excluded from the initial analysis. A group of patients with discordant pathologies was subsequently evaluated separately.
The clinical stage of the disease was determined using the Ann Arbor classification.15
PET
Patients were injected with 12 to 15 mCi (444 to 555 MBq) of FDG intravenously. Images were acquired from the level of the skull base to the inguinal regions. The average time interval between FDG injection and image acquisition was 61minutes (standard deviation [SD], 24 minutes; range, 40 to 130 minutes; all but six studies were performed between 40 and 80 minutes postinjection). Plasma glucose measurements were routinely obtained, and ranged from 53 to 222 mg/dL (including one diabetic patient with plasma glucose > 140 mg/dL), with a mean of 85 mg/dL (SD, 18 mg/dL) for the entire study population.
Studies were performed using the GE Advance (GE Medical Systems, Waukesha, WI) or HR plus (Siemens/CTI, Knoxville, TN) PETs. The GE Advance has a transaxial spatial resolution of 4.9, 5.3, and 6.1 mm at 1, 10, and 20 cm off axis, respectively. The slice thickness or axial resolution is 4.2 mm at the center. The HR plus has a transaxial spatial resolution of 4.5 mm full width at half maximum (FWHM) at 1 cm off axis and 5.4 mm at 10 cm off axis. The axial resolution is 4.2 mm at the center of the system. The following acquisition parameters were used: Advance, 3 minutes transmission, 4 minutes emission per bed position (15.2 cm); HR plus, 7 to 8 minutes per bed position (14.2 cm), with 35% of this time for the transmission scan. On both tomographs, attenuation correction was performed using Ge-68 transmission rods (GE Medical Systems).
Since November of 2001, studies were also acquired on combined PET/CT tomographs (either Biograph [Siemens/CTI, Nashville,TN] or DiscoveryLS [GE Medical Systems, Waukesha, WI]). Both PET/CT tomographs consist of a combination of a spiral CT tomograph with a PET tomograph. The Discovery uses the Advance PET tomograph, and the Biograph uses the HR plus PET tomograph. With this equipment, a low-dose CT scan, which is used for attenuation correction of PET emission images as well as for anatomic localization of PET abnormalities, is acquired first using the following parameters: for the Biograph, scout view with 130 kVp and 30 mAs, followed by spiral CT using effective mAs of 50, 130 kVp, scan width 5 mm, collimation of 4 mm and a feed/rotation of 12 mm; for the Discovery LS, scout view with 30 mAs, 120 kVp, followed by a spiral CT at 0.8 seconds/rotation using 80 mAs, 140 kVp, a slice thickness of 5 mm, and a 4.25 mm interval in high sensitivity mode. PET emission images were acquired for 4 minutes per bed position. The total acquisition time varied between 25 and 35 minutes per patient.
Regardless of the PET tomograph used, PET images were reconstructed using iterative algorithms. Attenuation correction was routinely applied. PET images of patients who met the above entry criteria were reviewed by one of the nuclear medicine physicians (H.S., L.W., or H.W.Y.). Abnormal FDG uptake was defined as greater than background activity in surrounding tissue and unrelated to physiologic sites of tracer uptake (such as the myocardium) or excretion (such as excreted FDG in the ureters and urinary bladder). For the spleen, abnormal uptake was defined as FDG accumulation greater than in the liver, based on standardized uptake value (SUV) measurements. Areas of abnormal FDG uptake were identified, and the intensity of FDG uptake was quantified by calculating the SUV. For the calculation of SUV, circular regions of interest ( 70 pixels) were drawn on transaxial images around the areas with increased FDG uptake. The SUV was calculated as:
To minimize partial volume effects and assure reproducibility of measurements, the maximum SUV (SUVmax) in a large lesion was used. The highest SUV per patient was recorded and used for further analysis. To assess for potential loss of count recovery, lesion size was measured as described previously.16
Histologic Analysis and Correlation With PET Findings
At least one lymph node biopsy was obtained per patient. All tissue specimens were reviewed by experienced hematopathologists at our institution. The recent WHO classification for NHL was used.17,18 Accordingly, patients were then grouped as having indolent or aggressive lymphoma. Mantle-cell lymphoma could not be classified into one of these groups. In some cases where the WHO classification could not be applied (eg, in cases with small cleaved cell lymphoma), the histopathologic description was recorded.
Statistical Analysis
SUVs were analyzed considering the various histologies and grades, and separately for newly diagnosed and recurrent disease. Two-sample comparisons were performed using the Wilcoxon test. A P value less than .05 was considered significant.
The accuracy of PET scans was evaluated in two different ways: one using the interpretation of the radiologist (positive/negative) and the other using the SUV. Because the nuclear medicine physician’s interpretation was dichotomous, sensitivity (true-positive fraction) and specificity (true-negative fraction) were computed using histopathology as the gold standard19.
To evaluate the accuracy of SUV, which is a continuous variable, we used receiver operating characteristic (ROC) curve analysis,19 again using histopathologic findings as the gold standard. The area under the curve was computed using the trapezoidal rule and used as a summary measure of accuracy. This analysis for SUV used only the subset of aggressive or indolent NHL patients for whom the site of biopsy was the same as the site at which SUVmax was obtained.
RESULTS
The study population (Table 1) consisted of 97 individuals; there were 51 males and 46 females (mean age, 59 years; SD, 15 years; range, 15 to 86 years). Sixty-eight patients presented with newly diagnosed disease. Twenty-nine patients had been treated previously and presented with relapsed or persistent disease. Prior treatment modalities included chemotherapy, immunotherapy and radiation.
Among patients with aggressive NHL (n = 63), the following specific histologies were identified: 55 diffuse large-cell lymphoma (53 of 55 patients with B-cell type); seven follicular lymphoma, grade 3; and one peripheral T-cell lymphoma.
Among patients with indolent NHL (n = 28), the following histologies were identified: one diffuse small cleaved; one unspecified follicular lymphoma; 10 follicular lymphoma, grade 1; four follicular lymphoma, grade 2; one lymphocytic plasmocytic lymphoma; four marginal zone lymphoma; one small cleaved-cell lymphoma; one small lymphocytic lymphoma; and five small-cell lymphoma.
Six additional patients presented with mantle cell lymphoma. These were analyzed separately as the nature of this disease does not fit the categories of indolent or aggressive disease and varies widely among patients.
Intensity of FDG Uptake
SUV in patients with NHL ranged from 3.2 to 43.0. In patients with indolent lymphoma, the SUV ranged from 2.3 to 13.0 (mean, 6.7; SD, 2.9). In patients with aggressive lymphoma, SUV ranged from 3.2 to 43.0 (mean, 17.2; SD, 9.7; P < .01 v indolent). In patients with mantle cell lymphoma, the SUV ranged from 4.7 to 8.7, with a mean of 6.4 (SD, 1.7). The intensity of FDG uptake in the various histologic subgroups is listed in Table 2, and patient examples are shown in Figures 1 and 2. The diameter of the lesion used for SUV analysis ranged from 1.0 to 18 cm.
Box plots for indolent and aggressive NHL are shown in Figure 3. FDG uptake in aggressive disease was significantly higher than in indolent disease (P < .01). A high FDG uptake was suggestive of aggressive disease since all of the indolent lymphomas showed an SUV 13. On the other hand, a low FDG uptake was not always an indicator of indolent disease: of the 63 patients with aggressive disease, 22 (35%) had lesions with an SUV < 13.
To further investigate whether SUV can distinguish between aggressive and indolent lymphoma, we used ROC curve analysis. The primary ROC analysis was restricted to the 69 patients for whom the site of biopsy was the same as the site at which SUVmax was measured. The area under the ROC curve was 84.7% (SD, 4%), which suggests that the SUV is a reasonable predictor of diagnosis (Fig 4). The various sensitivity and specificities that were obtained as the cutoff value for the SUV was varied are shown in Table 3. An SUV of 10 provided a fair balance, with a specificity of 81% and a sensitivity of 71% to detect an aggressive lymphoma. A higher cutoff would capture more indolent lymphomas. Conversely, a lower cutoff would capture a higher percentage of aggressive lymphomas.
In 22 patients (11 with indolent and 11 with aggressive lymphoma) the biopsy was not the site of highest SUV in the PET scan (mean difference between highest SUV and SUV at biopsy site, 40%; range, 5% to 60%; Table 4). Using an SUV of 10 as the cutoff to predict aggressive NHL, two patients with biopsy-proven indolent lymphoma were misclassified as having aggressive lymphoma. Similarly, six cases of aggressive lymphoma would have been misclassified as indolent lymphoma. The latter reflects the overlapping SUVs in the lower range.
Patients With Discordant Histologies
There were 12 patients with biopsies demonstrating discordant lymphoma pathologies at separate time points (Table 5) . Eight patients had previous diagnoses of indolent lymphomas (n = 7) or Hodgkin’s disease (n = 1), but at the time of the PET scan had biopsy proven aggressive lymphomas. There were four additional patients with previous diagnoses of aggressive lymphomas who had biopsy-proven indolent lymphomas at the time of the PET scan. PET imaging was not available at the time of the original diagnosis in any of these patients.
For the eight patients with transformed lymphoma, SUVs ranged from 4.8 to 29.8. Five of the eight patients had an SUV > 10: one patient had an SUV of 10.2, one 18.8, and three others > 25.
Among four patients with previous aggressive lymphoma and indolent disease at the time of the PET scan, two had an SUV < 10, whereas the other two patients had SUVs of 14.6 and 15.7, respectively. The patient with an SUV of 14.6 had clinically aggressive disease with five prior therapy regimens and clinical deterioration secondary to lymphoma after the PET study. For the patient with an SUV of 15.7, the biopsy was performed before the PET scan and came from a region where the SUV was only 6.9. It is not possible to rule out that the area with an SUV of 15.7 had a higher-grade lymphoma.
DISCUSSION
This study demonstrates that FDG PET imaging can distinguish between patients with indolent and aggressive lymphoma, and that the likelihood for aggressive disease increases in parallel with increases in SUV. This is in concordance with PET studies in a number of other malignancies demonstrating that higher FDG uptake in a given malignant tumor indicates a more aggressive course of the disease and is frequently associated with poor prognosis.20-26
Previous studies have demonstrated a considerable overlap in the intensity of FDG uptake between patients with indolent and aggressive lymphoma in the low SUV range.12-14 While we also noted such an overlap in the lower SUV range, the current study nevertheless demonstrates that the intensity of FDG uptake is generally lower in indolent as compared with aggressive non-Hodgkin’s lymphoma and that the likelihood for aggressive disease increases in parallel with increases in SUV. An SUV < 10 captured 81% of indolent lymphomas, while an SUV 13 was virtually indicative of aggressive disease.
Our results suggest that SUV may alter clinical management. For example, if a patient with previously documented indolent disease had a PET scan showing unexpectedly high SUVs, a transformation to more aggressive disease might be suspected and a biopsy considered. This would be particularly true in cases with very high SUV (> 13). Transformation occurs in roughly 25% of patients with indolent follicular lymphoma, is unpredictable, and does not necessarily involve all sites of known disease.27 Our series of patients with transformation is small, but 5 of 8 patients would have had PET scans suggestive of transformation if an SUV cutoff of 10 had been applied. In contrast, a low SUV in patients with previously aggressive disease is somewhat inconclusive because of the overlap between aggressive and indolent histologies in the lower SUV range.
PET scanning may also be helpful in directing biopsies. Although typically biopsies are obtained from areas that are easily accessible, an unexpectedly high SUV may suggest a transformation to aggressive lymphoma and a biopsy of the site with the highest SUV may be considered.
This is the largest study to date evaluating the relationship between histopathology and intensity of FDG uptake in lymphoma patients. Some previous smaller studies suggested that the intensity of FDG uptake in lymphoma correlated with the histologic grade of malignancy.12,13,28 Rodriguez et al12 studied 23 patients with non-Hodgkin’s lymphoma (11 high grade, nine low grade: three transformed, respectively) with FDG PET. Three different parameters for the assessment of tumor FDG uptake (SUV, mass influx, and transport rate) were significantly higher for high-grade than for low-grade lymphoma. However, only the transport rate appeared to distinguish between low- and high-grade disease, whereas tumor SUV (15.9 ± 7.7 v 8.7 ± 6.6) and mass influx showed a marked overlap. It should be noted that PET measurement of glucose transport rates is a cumbersome and time-consuming research technique directed at one field of view, usually 15 cm of the torso. Lapela et al13 studied 22 patients with untreated NHL and FDG PET: seven with low-grade lymphoma, 11 with intermediate grade, and four with high grade. SUV and metabolic rate of glucose were analyzed in one lesion per patient. The median SUV for all patients was 8.5 (range, 3.5 to 31). SUV for low-grade lymphoma ranged from 4 to 10; for intermediate grade, 3 to 25; and for high grade, 12 to 31. Differences in SUV were significant, but there was again considerable overlap between subgroups.
Our findings are in concordance with data reported in these earlier studies. However, by using ROC analysis we determined meaningful cutoffs above which indolent lymphomas are highly unlikely.
There is evidence that indolent lymphoma is not always detectable by FDG PET.10,11,28-30 However, a wide range in detection rates (40% to 98%) has been reported. Varying intensity of FDG uptake in lymphomas of different histologic subtypes has been cited as one explanation. For example, while Jerusalem et al10 found that FDG PET showed 40% more nodal lesions than did conventional imaging in patients with follicular lymphoma, only 58% of abnormal lymph nodes by CT size criteria were detected in patients with small lymphocytic lymphoma. Other studies showed lower PET detection rates for extranodal marginal zone lymphoma (MZL) and peripheral T-cell lymphoma.11,29 In one recent study, only eight of 12 MZLs (67%) were detected by PET using an SUV of 2.5 as the cutoff.11 In contrast, all four MZLs were visualized in our study, possibly as a result of improved PET technology or referral bias. Indeed, there could have been an overall referral bias to obtain a PET scan in younger patients or those with more clinically aggressive disease. While a high detection rate for indolent lymphoma was also noted in a recent study from Australia,30 a prospective study may be needed to define the ultimate role of PET in these patients.
Only the highest SUV per patient was used in the analysis. Variability in SUV from lesion to lesion in any particular patient is not addressed by the current investigation. Although lesions in patients with aggressive disease may show variability in SUV numbers, inclusion of all lesions with FDG uptake would likely blur the differences between histologic subgroups of lymphoma. In addition, lesions with the highest FDG uptake likely represent sites of the most aggressive disease. For classification into indolent or aggressive disease it would be important not to underestimate the metabolic activity and aggressiveness of the disease. Moreover, because lesions with higher tracer accumulation tend to be larger in size (data not shown), these SUV measurements are less prone to partial volume effects that may cause underestimation of the true activity concentration.
In some patients, the location of highest SUV was discordant from the location of biopsy. For practical purposes, biopsies are generally obtained from lymph nodes and tissues that are easily accessible. It is then generally assumed that the tissue obtained is representative of the disease elsewhere. However, our study suggests in patients with indolent NHL, in whom transformation is suspected, a repeat biopsy of an area with high SUV may be indicated in the appropriate setting.
The SUV is a relative measure of local radiotracer accumulation in tissue. It varies with biologic factors,31 the method of analysis,32 and PET image reconstruction parameters.33 Our results, therefore, only apply when the uptake period between FDG injection and imaging is at least 40 minutes and when attenuation correction and iterative reconstruction are used for generating the PET images.
In lesions smaller than twice the resolution (FWHM) of the PET tomograph, the activity concentration will be underestimated. Although the spatial resolution of current dedicated PET tomographs is quoted as between 4 and 5 mm, in our clinical setting (considering matrix and reconstruction parameters), the spatial resolution of PET tomographs used is approximately 7 to 8 mm FWHM. In lesions smaller than 2x FWHM (ie, 1.5 cm diameter) the true activity concentration (and hence, the SUV) in this lesion will therefore be underestimated. The reason for this is a loss of count recovery. However, such lack of count recovery did not affect the overall results in this study. Although 11 lesions had a minimum diameter of 1.5 cm, this occurred to a similar degree in group with indolent lymphoma (4 of 28 lesions; SUV range, 3.2 to 5.5) and the group with aggressive lymphoma (7 of 63 lesions; SUV range, 5.5 to 10.0).
Studies were acquired with different types of PET tomographs, reflecting our clinical practice over the past several years. In phantom studies (unpublished results) we have noted an approximately 10% difference in quantitative measurements between Advance and Biograph tomographs. However, in patient studies we did not observe any significant difference between SUV measurements obtained with different PET tomographs for patients with either aggressive or indolent disease. It should also be noted that technical and patient-related factors could potentially influence SUV measurements to a greater degree than the observed 10% difference between tomographs. In clinical practice it may be difficult to standardize all potential parameters that can affect SUV measurements. In addition, our study shows overlap of data in the lower SUV range. Overall, it is therefore impossible to define one specific SUV number that will distinguish with 100% accuracy between aggressive and indolent lymphoma. Rather, our data show that the likelihood for aggressive disease increases in parallel with rises in SUV, and an SUV > 13 was virtually indicative of aggressive disease.
Although our study showed higher FDG uptake in aggressive lymphoma as a group, a wide range of SUVs was observed among both patients with an indolent and aggressive subtype of the disease. Future studies will have to address whether FDG PET can identify subsets of patients with better or worse clinical course and prognosis among patients with aggressive disease. Future studies will also have to investigate whether imaging with other radiotracers, such as the proliferation marker 18F fluoro-thymidine, have superior predictive value.
Regardless of the findings in this study, biopsy remains the standard procedure for establishing an unequivocal diagnosis in patients with lymphoma. However, in selected cases where biopsies from easily accessible peripheral lymph nodes are inconsistent with the clinical course of the disease, or where the PET scan shows additional lesions with significantly higher metabolic activity, an effort should be made to obtain additional biopsies from these sites of intense FDG uptake on PET. In unique situations, it may be advisable to treat for aggressive disease if a biopsy cannot be safely obtained.
The intensity of FDG uptake on PET images correlates with tumor aggressiveness in NHL. Low intensity of FDG uptake can be observed in both indolent and aggressive disease, but increasingly higher SUV have a higher specificity for the detection of aggressive disease. Although FDG PET alone cannot reliably diagnose transformation from indolent to aggressive lymphoma, PET findings may be useful to guide biopsies of lesions with the highest SUV when clinically appropriate.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Hoh CK, Glaspy J, Rosen P, et al: Whole-body FDG-PET imaging for staging of Hodgkin's disease and lymphoma. J Nucl Med 38:343-348, 1997
Carr R, Barrington SF, Madan B, et al: Detection of lymphoma in bone marrow by whole-body positron emission tomography. Blood 91:3340-3346, 1998
Moog F, Bangerter M, Diederichs CG, et al: Extranodal malignant lymphoma: Detection with FDG PET versus CT. Radiology 206:475-481, 1998
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
Buchmann I, Reinhardt M, Elsner K, et al: 2-(fluorine-18)fluoro-2-deoxy-D-glucose positron emission tomography in the detection and staging of malignant lymphoma: A bicenter trial. Cancer 91:889-899, 2001
Romer W, Hanauske AR, Ziegler S, et al: Positron emission tomography in non-Hodgkin's lymphoma: Assessment of chemotherapy with fluorodeoxyglucose. Blood 91:4464-4471, 1998
Spaepen K, Stroobants S, Dupont P, et al: Early restaging positron emission tomography with (18)F-fluorodeoxyglucose predicts outcome in patients with aggressive non-Hodgkin's lymphoma. Ann Oncol 13:1356-1363, 2002
Moog F, Bangerter M, Kotzerke J, et al: 18-F-fluorodeoxyglucose-positron emission tomography as a new approach to detect lymphomatous bone marrow. J Clin Oncol 16:603-609, 1998
Schoder H, Meta J, Yap C, et al: Effect of whole-body (18)F-FDG PET imaging on clinical staging and management of patients with malignant lymphoma. J Nucl Med 42:1139-1143, 2001
Jerusalem G, Beguin Y, Najjar F, et al: Positron emission tomography (PET) with 18F-fluorodeoxyglucose (18F-FDG) for the staging of low-grade non-Hodgkin's lymphoma (NHL). Ann Oncol 12:825-830, 2001
Elstrom R, Guan L, Baker G, et al: Utility of FDG-PET scanning in lymphoma by WHO classification. Blood 101:3875-3876, 2003
Rodriguez M, Rehn S, Ahlstrom H, et al: Predicting malignancy grade with PET in non-Hodgkin's lymphoma. J Nucl Med 36:1790-1796, 1995
Lapela M, Leskinen S, Minn HR, et al: Increased glucose metabolism in untreated non-Hodgkin's lymphoma: A study with positron emission tomography and fluorine-18-fluorodeoxyglucose. Blood 86:3522-3527, 1995
Filmont J, Ko Yang F, Yap C, et al: Prediction of malignancy grade using 18FDG PET in Non Hodgkin's lymphoma patients: An urban legend J Nucl Med 43:79, 2002 (abstr)
Rosenberg S: Validity of the Ann Arbor staging classification for the non-Hodgkin's lymphomas. Cancer Treat Rep 61:1023-1027, 1977
Erdi YE, Mawlawi O, Larson SM, et al: Segmentation of lung lesion volume by adaptive positron emission tomography image thresholding. Cancer 80:2505-2509, 1997
Harris NL, Jaffe ES, Diebold J, et al: Lymphoma classification—From controversy to consensus: The R.E.A.L. and WHO Classification of lymphoid neoplasms. Ann Oncol 11:3-10, 2000
Harris NL, Jaffe ES, Diebold J, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee meeting, Airlie House, Virginia, November 1997. J Clin Oncol 17:3835-3849, 1999
Pepe M: The statistical evaluation of medical tests for classification and prediction. Oxford, England Oxford University Press, 2003, pp 66-94
Ahuja V, Coleman RE, Herndon J, et al: The prognostic significance of fluorodeoxyglucose positron emission tomography imaging for patients with nonsmall cell lung carcinoma. Cancer 83:918-924, 1998
Wong RJ, Lin DT, Schoder H, et al: Diagnostic and prognostic value of [(18)F]fluorodeoxyglucose positron emission tomography for recurrent head and neck squamous cell carcinoma. J Clin Oncol 20:4199-4208, 2002
Eary JF, O'Sullivan F, Powitan Y, et al: Sarcoma tumor FDG uptake measured by PET and patient outcome: A retrospective analysis. Eur J Nucl Med Mol Imaging 29:1149-1154, 2002
Vansteenkiste JF, Stroobants SG, Dupont PJ, et al: Prognostic importance of the standardized uptake value on (18)F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: An analysis of 125 cases: Leuven Lung Cancer Group. J Clin Oncol 17:3201-3206, 1999
Fukunaga T, Okazumi S, Koide Y, et al: Evaluation of esophageal cancers using fluorine-18-fluorodeoxyglucose PET. J Nucl Med 39:1002-1007, 1998
Oshida M, Uno K, Suzuki M, et al: Predicting the prognoses of breast carcinoma patients with positron emission tomography using 2-deoxy-2-fluoro[18F]-D-glucose. Cancer 82:2227-2234, 1998
Nakata B, Chung YS, Nishimura S, et al: 18F-fluorodeoxyglucose positron emission tomography and the prognosis of patients with pancreatic adenocarcinoma. Cancer 79:695-699, 1997
Bastion Y, Sebban C, Berger F, et al: Incidence, predictive factors, and outcome of lymphoma transformation in follicular lymphoma patients. J Clin Oncol 15:1587-1594, 1997
Leskinen-Kallio S, Ruotsalainen U, Nagren K, et al: Uptake of carbon-11-methionine and fluorodeoxyglucose in non-Hodgkin's lymphoma: A PET study. J Nucl Med 32:1211-1218, 1991
Hoffmann M, Kletter K, Becherer A, et al: 18F-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) for staging and follow-up of marginal zone B-cell lymphoma. Oncology 64:336-340, 2003
Blum RH, Seymour JF, Wirth A, et al: Frequent impact of [18F]fluorodeoxyglucose positron emission tomography on the staging and management of patients with indolent non-Hodgkin's lymphoma. Clin Lymphoma 4:43-49, 2003
Huang SC: Anatomy of SUV: Standardized uptake value. Nucl Med Biol 27:643-646, 2000
Yeung HW, Sanches A, Squire OD, et al: Standardized uptake value in pediatric patients: An investigation to determine the optimum measurement parameter. Eur J Nucl Med Mol Imaging 29:61-66, 2002
Schoder H, Erdi YE, Chao K, et al: Clinical implications of different image reconstruction parameters for interpretation of whole-body PET studies in cancer patients. J Nucl Med 45:559-566, 2004(Heiko Schder, Ariela Noy,)