Contribution of PET imaging to the initial staging and prognosis of patients with Hodgkin s disease

Original article Annals of Oncology 15: 1699 1704, 2004 doi:10.1093/annonc/mdh426 Contribution of PET imaging to the initial staging and prognosis of patients with Hodgkin s disease R. Munker 1 *, J. Glass 1, L. K. Griffeth 2,3, T. Sattar 1, R. Zamani 4, M. Heldmann 4, R. Shi 1 & D. L. Lilien 5 Departments of 1 Medicine and 4 Radiology, Feist-Weiller Cancer Center and 5 PET Imaging Center of the Biomedical Research Foundation of Northwest Louisiana, at Louisiana State University Health Sciences Center, Shreveport, LA; 2 North Texas Clinical PET Institute, Department of Radiology, Baylor University Medical Center, Dallas, TX; 3 US Oncology, Inc., Dallas, TX, USA Received 23 March 2004; accepted 16 June 2004 Background: Positron emission tomographic (PET) scanning utilizing [ 18 F]fluorodeoxyglucose (FDG) is a new method of tumor imaging based on the increased glucose metabolic activity of malignant tumors. In Hodgkin s disease (HD), PET has proven value for the evaluation of residual masses following treatment and for the early diagnosis of relapse. In the initial staging of HD, PET frequently shows a higher stage than conventional methods (upstaging by PET). In the present study, we evaluated the frequency of stage changes by PET in a multicenter setting and determined its prognostic relevance. Patients and methods: A total of 73 patients with newly diagnosed HD were staged with both conventional methods and whole-body PET scanning. All histological types and stages were represented. The median time of follow-up after the initial diagnosis was 25 months (range 1 month to 5 years). The response to treatment was determined by standard clinical and diagnostic criteria. For the purpose of this analysis, data from a PET center associated with a university medical center and a PET center associated with a group oncology practice were combined. Results: A total of 21 patients (28.8%) were upstaged by PET compared with conventional methods. In two cases (2.7%), a lower stage was suggested by PET scanning. With one possible exception, the upstaging had no obvious clinical or biological correlate. Among 12 patients in stage I (A + B) by conventional methods, seven were upstaged by PET (58.3%), four to stage II, one to stage III and two to stage IV. Among 42 patients in stage II, eight were upstaged by PET (19.0%), six to stage III and two to stage IV. Among 12 patients in stage III, six (50%) were upstaged to stage IV by PET. If only early-stage patients and major changes are considered (stages IA IIB to III or IV), among 49, 10 were upstaged to III or IV, whereas in 39 staging was unchanged following PET. In the former group, three relapsed or were refractory compared with none in the latter group (P < 0.006). In advanced stage patients (IIIA or IIIB) a trend toward treatment failure was apparent in patients who were upstaged by PET. Conclusions: PET scanning is an interesting new modality for the accurate staging of patients with HD and frequently shows a higher stage than conventional methods. PET should be performed at initial diagnosis and should be included in prospective studies of patients with HD. Upstaging by PET may represent a risk factor for a more advanced stage or a biologically more aggressive tumor. Patients with early-stage disease as identified by conventional methods have a significant risk of treatment failure if a more advanced stage is indicated by PET. At present, major stage changes suggested by PET imaging should be confirmed by an independent diagnostic method. Key words: Hodgkin s disease, positron emission tomography, staging Introduction *Correspondence to: Dr R. Munker, Division of Hematology/Oncology, Louisiana State University, 1501 Kings Highway, Shreveport, LA 71130, USA. Tel: + 1-318-675-8770; Fax: + 1-318-675-5944; E-mail: rmunke@lsuhsc.edu Hodgkin s disease (HD) is a paradigm of a curable malignancy. With current therapy, more than 75% of patients will survive free of disease for more than 5 years, and the large majority of these will be cured [1 4]. These results have been obtained with judicious use of progressively improved staging methodology with time, and the stage-adapted use of radiation q 2004 European Society for Medical Oncology

1700 and chemotherapy. For relapsing patients, high-dose chemotherapy is an additional treatment option. In earlier years, a staging laparotomy was the gold standard of diagnostic procedures in newly diagnosed HD. Later, computed tomography (CT) became widely available and was at the basis of the Cotswold staging classification for HD [5]. The indications for staging laparotomy have become more restricted in recent years, because of the more widespread use of chemotherapy, and because newer imaging techniques such as ultrasonography and magnetic resonance imaging (MRI) have entered clinical use [6 8]. More recently, positron emission tomography (PET) imaging with [ 18 F]fluorodeoxyglucose (FDG), has become available at most major medical centers. PET imaging with FDG is based on the increased glucose metabolic activity of most malignant tumors and, in general, is a sensitive method for the diagnosis and follow-up after treatment of a wide variety of different tumors [9]. However, imaging with FDG PET may not be specific in all situations. PET scanning in HD already has proven its value in the assessment of residual masses after treatment and for the early diagnosis of relapse [10 12]. Further supporting the clinical value of PET, a study combining data on 19 patients with HD and 41 patients with non-hodgkin s lymphoma, has shown that PET negativity after high-dose therapy predicts relapse-free survival after an autologous stem cell transplant [13]. In the present study we investigated the contribution of PET scanning to the initial staging of HD and attempted to determine whether PET positivity or the change of stage by PET scanning had any prognostic impact. Patients and methods Initial staging evaluation and treatment Table 1. Patient characteristics (n = 73) n Age (years) Range 16 79 Mean 37.6 Median 36 Gender Male 42 Female 31 Histological subtypes of HD Lymphocyte predominance 9 Nodular sclerosis 42 Mixed cellularity 16 Other types, unclassified, unknown 6 Stage distribution a IA n =9 IB n =3 I n =12 IIA n = 35 IIB n =7 II n =42 IIIA n = 6 IIIB n = 6 III n =12 IVA n = 1 IVB n =6 IV n =7 a Staging as determined by conventional techniques. HD, Hodgkin s disease. This is a retrospective study based on record review of patients referred to two PET centers for FDG PET imaging as part of initial staging of HD. All patients had a histological diagnosis of HD, typically from lymphoid tissue. One patient was known to be HIV-positive. All underwent conventional staging procedures including history, physical examination, complete blood counts, serum chemistry profile, CT scan of the chest, abdomen and pelvis, and in all cases a bone marrow biopsy and aspiration. In selected cases, other imaging methods such as ultrasonography and MRI were performed. If available, the diagnostic CT scans were reviewed without knowledge of the results of the PET scans. For staging purposes, the Cotswold modification of the Ann Arbor classification was used. The clinical and histological characteristics of the patients studied are described in Table 1. The patients came from the PET facility of a large university medical center in Northwest Louisiana (Biomedical Research Foundation, BRF; n = 45) and the PET facilities of a large group practice in North Texas (North Texas Clinical PET Institute, NTCPI; n = 28). All consecutive, untreated patients who had PET scanning as part of their initial evaluation (May 1996 to January 2002) were included in the present study. Early-stage patients (IA, IB, IIA without risk factors) were generally treated with radiotherapy. Intermediate-stage patients were generally treated with combined modality [four to six cycles of ABVD (doxorubicin, bleomycin, vinblastine and darcarbacine) + involved field radiotherapy]. Advanced-stage patients (IIIB, IVA, IVB) were generally treated with six to eight cycles of ABVD. In case of major stages changes by PET (e.g. stage I or II to III or IV), the treatment plan was changed in some, but not all, cases, by giving more chemotherapy or only chemotherapy. Four patients underwent a stem cell transplantation after relapse (three autologous, one allogeneic). FDG PET imaging All patients were studied after a fast of at least 6 8 h duration, usually overnight. Fingerstick serum glucose levels were determined in all patients using commonly available portable monitoring devices. Diabetics were not excluded. At BRF, two diabetics were studied, with glucose levels at the time of FDG injection of 166 and 129 mg/dl; all other patients had glucose of <_127 mg/dl. At NTCPI, two diabetics were also included in the study group (maximum glucose levels 127 mg/dl). Some details of the methodology varied between the two participating institutions and, at different times, at the BRF. All BRF patients were hydrated orally with 500 cm 3 of water prior to FDG injection, and with 250 cm 3 0.9% saline solution intravenously after FDG injection. Furosemide, 10 mg, was injected intravenously 15 min after FDG injection. A standard 10 15 mci (370 555 Mbq) dose of FDG was utilized, and the patients were comfortably situated in a quiet, dimly lit room during the uptake period. They were instructed to minimize motion and talking during the initial 30 min of the uptake period. Many patients were given 5 mg diazepam orally to minimize FDG accumulation at muscle insertion sites and within brown fat, or because of potential claustrophobia or anxiety. Patients initially studied at the BRF underwent scanning utilizing an Advance scanner (General Electric, Milwaukee, WI, USA) with a delay of 1 h after FDG injection. After October 2000, the delay prior to scanning was increased to 90 min. In all patients, whole-body emission scans were performed, and prior to September 2000 all but three patients had attenuation-corrected scans of the area of interest. After September 2000, whole-body attenuation-corrected scans utilizing segmented attenuation correction and iterative reconstruction techniques were performed. At NTCPI, patients were studied at ambient hydration state, without diuresis. The delay before scanning varied between 45 and 60 min,

and imaging was performed on an ECAT EXACT scanner (CTI, Knoxville, TN, USA). All images were reconstructed utilizing segmented attenuation correction and standard iterative reconstruction algorithms supplied by the vendor. All other details were essentially identical and are typical of methodology in common use at the majority of PET centers in the USA. At both sites, scans were qualitatively interpreted by experienced PET physicians, utilizing live display on the scanner workstations. At the discretion of the PET reader, review of non-attenuation-corrected images or semi-quantitative measurement of focal tracer uptake (SUV or standard uptake value) were utilized as ancillary tools at NTCPI. Interpreting physicians had access to clinical information and the results of other imaging studies. For the purposes of this study, the final PET stage was based on the dictated reports generated at the time the scans were initially performed and interpreted. All studies were performed prior to initiation of therapy and, in many patients, additional examinations during or following therapy or at later intervals for surveillance for recurrence also were performed but are not further considered in this study. Follow-up Patients were seen regularly in clinic and the status of their disease was determined. Patients who were not followed by the institution that started treatment were contacted and asked for health information (general state of health, relapse, treatment complications). In patients who had died, the likely cause of death was determined from medical records. The range of follow-up was 1 month to 5 years; the median time of follow-up from diagnosis was 25 months. In the total series of patients, 66 reached a complete remission or were free from progression if residual abnormalities persisted. Five patients failed the induction treatment and two were not evaluable (one death from a treatment complication and one in whom the induction chemotherapy was not yet finished). The initial response was evaluated after four to six cycles of chemotherapy, if chemotherapy was given. If residual masses persisted, those masses were followed closely and follow-up biopsies obtained if felt clinically necessary; many patients had FDG PET imaging as part of their follow-up and information from that study was also utilized in assessing results of therapy. At last followup, six patients had died and five had relapsed after having previously reached complete remission. At last follow-up, one patient had developed a second malignant neoplasm (breast cancer) and one patient had relapsed with a composite lymphoma. Follow-up PET scans were performed in most patients following completion of therapy and at 3- to 6-month intervals. Statistical analysis Fisher s exact test was used to test the association between the proportion of changes in stage with different histological subtype, etc. A two-sided P value was used to test our null hypothesis that there is no association. If the P value was < 0.05 a statistically significant association was concluded. Survival and freedom from treatment failure were analyzed using Kaplan Meier plots. Results A total of 73 patients with newly diagnosed HD were included in this study (see Table 1 for clinical and histological details). Altogether, in 21 patients (28.8%) a higher stage was indicated by PET than by conventional staging. In two patients, a lower stage was suggested by PET. In one of these two cases, an enlarged spleen was negative by PET. Later, after polychemotherapy, this patient entered complete remission, without further splenic abnormalities. In the other case, a histologically positive bone marrow was missed by PET. The percentage of patients who had a higher stage following PET was comparable in the patients from the group practice PET center and the university PET center (32% versus 27%, difference not statistically significant). In Table 2, the phenomenon of upstaging by PET was correlated with sex, early versus later stage, presence or absence of B symptoms and age. No statistically significant differences could be found. As the only possible exception, the histological subtype of mixed cellularity had a higher incidence of upstaging by PET compared with the subtype of nodular sclerosis (borderline statistically significant). Table 3 shows details of the patients who had a higher stage following PET. In stage I, both patients who had stage IV based on PET had bone involvement. In one of these cases, the bone involvement was confirmed histologically; in the other case, persistent B symptoms made stage IV clinically plausible. Among the seven patients who changed their stage from I or II to III, four had a supradiaphragmatic and three had an infradiaphragmatic presentation by conventional staging. In patients who did not have a stage change, the pattern of positivity given by conventional staging was, in general, confirmed by PET (data not shown). In four of the seven patients with stage IV disease by conventional staging, additional major sites of disease were identified by PET (one case each of subcutaneous, splenic, bone and liver involvement). The outcome and prognostic implications of stage changes following PET are shown in Table 4. If the different stages are analyzed separately, no statistical difference in freedom from Table 2. Higher stage following PET scanning n (%) Total group of patients 21/73 (28.8%) Sex Male 8/31 (25.8%) 0.79 (NS) Female 13/42 (31.0%) Histological subtype Nodular sclerosis 9/42 (21.4%) 0.05 (NS) Mixed cellularity 8/16 (50.0%) Lymphocyte predominance 3/9 (33.3%) Other types, unclassified, unknown 1/6 (16.7%) Early stages versus late stages Stages IA, IB, IIA, IIB 15/54 (27.8%) 0.17 (NS) Stages IIIA, IIIB 6/12 (50.0%) Stage IA, IB 7/12 (58.3%) 0.15 (NS) Stage IIA, IIB 8/42 (19.0%) Presence or absence of B symptoms (only stages I III) No B symptoms 17/51 (33.3%) 0.26 (NS) B symptoms 4/22 (18.2%) Age at diagnosis < 37 years 8/39 (20.5%) 0.12 (NS) >_ 37 years 13/34 (38.2%) PET, positron emission tomography; NS, not significant. P 1701

1702 Table 3. Stage changes following PET Stage prior to PET Total number of patients Stage changes following PET [n (%)] Comments IA + B 12 7 (58.3) four to stage II, one to stage III, two to stage IV IIA + B 42 8 (19.0) six to stage III, two to stage IV IIIA + B 12 6 (50.0) All to stage IV, two cases of bone, two cases of bone marrow, two cases of liver involvement PET, positron emission tomography. Table 4. Prognostic implications of stage changes following PET treatment failure could be demonstrated between the patients who maintained their stage and patients who moved to a higher stage (Kaplan Meier analysis, data not shown). However, when patients in stages I and II were combined and patients moving from stages I or II to III or IV were compared with patients remaining in stages I or II following PET, a significant difference in freedom from treatment failure could be demonstrated (three of 10 patients failed versus none of 39 patients; P < 0.006). Upstaging by PET indicates more widespread disease and probably a worse prognosis. This phenomenon occurs in all stages, but may be especially frequent in stages I or III. Patients who advance from stage I or III to stage IV may have a prognosis as bad as if they were a priori in stage IV (see Table 4). Indeed, two-thirds of patients who went from stage III to stage IV by PET were refractory or relapsed early. Liver involvement, even if visualized only by PET, may herald a bad prognosis. Indeed, in one of two patients with liver involvement visualized only by PET, the liver was the only site of relapse and cause of ultimate treatment failure. As mentioned, PET missed one of three cases of bone marrow involvement. Owing to the generally good prognosis of our patients with HD and the limited time of followup, no difference in survival could be demonstrated between patients who remained in the same stage and patients who were upstaged by PET (Kaplan Meier analysis, data not shown). Discussion Stage (by conventional staging) Stage I Stage II Stage III Stage IV Total number of patients 12 42 12 7 Same stage following PET 5 34 6 a 7 Complete remission 5 34 4 5 Refractory/relapsed 0 0 1 2 Higher stage following PET 7 8 a 6 NA Complete remission 5 6 2 NA Refractory/relapsed 2 1 4 NA a One patient each inevaluable. PET, positron emission tomography; NA, not applicable. Our study is one of the largest series reported to date of patients with HD who had PET scanning as part of their initial work-up. In previous studies, a total of 15 44 patients were investigated by PET and conventional staging modalities [14 18]. In these previous studies, PET imaging resulted in a higher stage than conventional imaging in 9% to 41% of patients. In our series, 21/73 patients (28.8%) had a higher stage following PET. In the previous studies, no or very limited clinical follow-up was reported. A recent study with 77 patients suggested stage changes in 20% of cases and changes of treatment in 18% of cases [19]. However, these changes were only hypothetical in the context of several doseintense protocols. Our series is unique since we present a careful clinical follow-up of the patients treated for HD with standard chemotherapy and/or radiotherapy. The clinical implications of these stage changes by PET are not always clear-cut [20]. Depending on the treatment strategies, a minor stage change (e.g. from stage I to II) may or may not have clinical relevance. In patients with early stages, PET staging may help tailor radiation therapy, an additional benefit that may not be reflected in change of stage. Major stage changes (e.g. from stage II to IV) indicate more widespread disease. Our data show a worse prognosis for the cases which went to stage III or IV from stage I or II by PET, and possibly also for the cases which went from stage III to stage IV. It is not surprising that no statistical differences in survival could be detected in these cases (limited number of cases in each category, availability of salvage treatments, generally good prognosis). In non-small-cell lung cancer, a tumor with mostly unfavorable prognosis, PET scanning has allowed the stratification of patients for radical surgery and radiochemotherapy [21]. Gallium-67 scanning is another method for the metabolic staging of malignant lymphomas, including HD, and has been used for many years in the initial evaluation and more particularly in the follow-up of HD. Gallium imaging has several disadvantages relative to FDG PET, including lower spatial resolution of the scanning technique and lower tumor to background activity levels with resultant poorer image quality. Furthermore, gallium scanning may require up to 1 week to 10 days to complete, compared with FDG PET imaging, which is completed in several hours. Recent data show that the accuracy of gallium scanning compares unfavorably with FDG PET and should therefore no longer be of routine use if FDG PET is available [22]. FDG PET positivity generally indicates a metabolically active tumor, but in some cases the tracer uptake may be nonspecific or unrelated to active tumor. According to the literature, many inflammatory or infectious lesions may also show FDG PET positivity; sarcoidosis and some systemic chronic

granulomatous infections may be particularly difficult to distinguish from HD based solely on scan appearance [23, 24]. Since the cellular infiltrate in lesions of HD is composed of tumor cells as well as reactive lymphoid and stromal cells, it is likely that FDG uptake in reactive cells is also, at least in part, responsible for scan positivity. This does not decrease the value of PET scanning at initial diagnosis and for suspected relapse. Another application in later follow-up might be the detection of second malignant tumors in patients with HD. A potential shortcoming of FDG PET imaging is limited spatial resolution. Microscopic disease (as in limited bone marrow infiltration in patients without obvious signs of marrow compromise) is not evident on PET. On the other hand, FDG PET may show focal marrow infiltration that is not evident or easily assessed by other methods. PET may also occasionally miss a focal site of disease activity in HD, especially in normal sized nodes (as defined by clinical or CT criteria) [16, 17, 24]. It is clear, however, that FDG PET may be positive in normal sized but involved nodes by these criteria in a variety of malignancies, well documented in nonsmall-cell lung cancer [21]. Different from most cases of HD, low-grade non-hodgkin s lymphomas may be negative for FDG uptake [25, 26]. In our series we describe an interesting biological phenomenon: upstaging by PET may be more frequent in the subtype of mixed cellularity than in nodular sclerosis. This finding needs confirmation. A possible rationale might be that the lesions in the subtype of mixed cellularity are metabolically more active or occur at sites poorly evaluated by CT. The finding does not automatically indicate that disease in this subtype is more widespread or aggressive. Several authors have suggested that treatment should change according to PET results [14, 16, 17, 19]. To some extent, this may already have happened in clinical practice. However, treatment according to PET positivity under all circumstances is not supported by the presently available data. PET scanning may permit a more accurate diagnosis but major stage changes (e.g. from stage I to stage IV with consequent therapeutic consequences) should be investigated by further diagnostic studies and, if necessary, a biopsy should be performed. According to our data, true stage I may be rare using a sensitive method like PET scanning. A diagnostic test is only clinically useful if the results indicate particular clinical features, predict the treatment outcome or correlate with other established diagnostic tests. Therefore, FDG PET scanning at initial diagnosis should be included in prospective studies of HD. It is reasonable to assume that upstaging by PET is an indicator of more widespread disease and of worse prognosis. Upstaging by PET may be a risk factor like increased levels of soluble CD30, sedimentation rate and b 2 -microglobulin [27, 28]. As mentioned previously, a further use of pretreatment PET scanning may be in better planning of radiotherapy ports and in the choice of patients for different treatment approaches such as up-front high-dose therapy or immunotherapy. However, these uses of PET should be validated by clinical studies. In the future, technical advances like new tracers other than FDG, perhaps with greater specificity, scanners with improved spatial resolution, and scanners that combine PET and CT may further enhance the value of PET in the initial diagnosis and follow-up of patients with HD. Acknowledgements The authors would like to thank all physicians referring patients, especially Drs Nagen Bellare, Harry Butler, Laura Campbell, Joyce Feagin, Alan Grosbach, Robert Massingill and Robert Nickelson, and the US Oncology group, especially Kathleen Chilton, for data collection and follow-up information. Presented in part at the 39th Annual Meeting of the American Society of Clinical Oncology, May 31 to June 3, 2003, Chicago, IL, USA. References 1703 1. Linch DC, Gosden RG, Tuland T et al. Hodgkin s lymphoma: choice of therapy and late complications. Hematology 2000; 205 221. 2. Bonnadonna G. Historical review of Hodgkin s disease. Br J Haematol 2000; 110: 504 511. 3. Josting A, Wolf J, Diehl V. Hodgkin disease: prognostic factors and treatment strategies. Curr Opin Oncol 2000; 12: 403 411. 4. Donaldson SS, Hancock SL, Hoppe RT. The Janeway lecture. Hodgkin s disease finding the balance between cure and late effects. Cancer J Sci Am 1999; 5: 325 333. 5. Crowther D, Lister R. The Cotswolds report on the investigation and staging of Hodgkin s disease. Br J Cancer 1990; 62: 551 552. 6. Castellino RA. Diagnostic imaging studies in patients with newly diagnosed Hodgkin s disease. Ann Oncol 1992; 3 (Suppl 4): S45 S47. 7. Munker R, Stengel A, Staebler A et al. Diagnostic accuracy of ultrasound in the staging of Hodgkin s disease: verification by laparotomy in 100 cases. Cancer 1995; 76: 1460 1466. 8. Bangerter M, Griesshammer M, Bergmann L. Progress in medical imaging of lymphoma and Hodgkin s disease. Curr Opin Oncol 1999; 11: 339 342. 9. Czernin J, Phelps ME. Positron emission tomography scanning: current and future applications. Ann Rev Med 2002; 53: 89 112. 10. Jerusalem G, Beguin Y, Fassotte MF et al. Whole-body positron emission tomography using 18 F-fluorodeoxyglucose for post-treatment evaluation in Hodgkin s disease and non-hodgkin s lymphoma has higher diagnostic and prognostic value than classical computed tomography scan imaging. Blood 1999; 94: 429 433. 11. Weihrauch MR, Re D, Scheidhauer K et al. Thoracic positron emission tomography using 18F-fluorodeoxyglucose for the evaluation of residual mediastinal Hodgkin disease. Blood 2001; 98: 2930 2934. 12. Jerusalem, Beguin Y, Fassotte MF et al. Early detection of relapse by whole-body emission tomography in the follow-up of patients with Hodgkin s disease. Ann Oncol 2003; 14: 123 130. 13. Spaepen K, Stroobants S, Dupont P et al. Prognostic value of pretransplant positron emission tomography using fluorine-18-fluoro-deoxyglucose in patients with aggressive lymphoma treated with high-dose chemotherapy and stem cell transplantation. Blood 2003; 102: 53 59. 14. Partridge S, Timothy A, O Doherty MJ et al. 2-Fluorine-18-fluoro-2- deoxy-d.glucose positron emission tomography in the pretreatment staging of Hodgkin s disease: influence on patient management in a single institution. Ann Oncol 2000; 11: 1273 1279.

1704 15. Jerusalem G, Beguin Y, Fassotte MF et al. Whole-body positron emission tomography using 18 F-fluorodeoxyglucose compared to standard procedures for staging patients with Hodgkin s disease. Haematologica 2001; 86: 266 273. 16. Wirth A, Seymour JF, Hicks RJ et al. Fluorine-18 fluorodeoxyglucose positron emission tomography, gallium-67 scintigraphy, and conventional staging for Hodgkin s disease and non-hodgkin s lymphoma. Am J Med 2002; 112: 262 268. 17. Weihrauch MR, Re D, Bischoff S et al. Whole-body positron emission tomography using 18F-fluorodeoxyglucose for initial staging of patients with Hodgkin s disease. Ann Hematol 2002; 81: 20 25. 18. Menzel C, Dobert N, Mitrou P et al. Positron emission tomography for the staging of Hodgkin s lymphoma increasing the body of evidence in favor of the method. Acta Oncol 2002; 41: 430 436. 19. Naumann R, Beuthien-Baumann B, Reiss A et al. Substantial impact of FDP PET imaging on the therapy decision in patients with earlystage Hodgkin s lymphoma. Br J Cancer 2004; 90: 620 625. 20. Kaplan LD. Fluorine-18 fluorodeoxyglucose positron emission tomography for lymphoma: incorporating new technology into clinical care. Am J Med 2002; 112: 320 321. 21. MacManus MP. The role of positron emission tomography in the prognosis and in restaging of non-small cell lung cancer. In Perry MC (ed.). American Society of Clinical Oncology 2003; Educational Book. Alexandria (VA): 790 796. 22. Kostakoglu L, Leonard JP, Kuji I et al. Comparison of fluorine-18 fluorodeoxy-glucose positron emission tomography and Ga-67 scintigraphy in evaluation of lymphoma. Cancer 2002; 94: 879 888. 23. Sandherr M, von Schilling C, Link T et al. Pitfalls in imaging Hodgkin s disease with computed tomography and positron emission tomography using fluorine-18-fluorodeoxy-glucose. Ann Oncol 2001; 12: 719 722. 24. Karapetis CS, Strickland AH, Yip D et al. PET and PLAP in suspected testicular cancer relapse: beware sarcoidosis. Ann Oncol 2001; 12: 1485 1488. 25. Jerusalem G, Silvestre RM, Beguin Y et al. Does 18 F-FDG PET replace bone marrow biopsy (BMB) in patients with Hodgkin s disease (HD) or non-hodgkin s lymphoma (NHL)? Blood 2002; 100 (Suppl): 768a (Abstr 3039). 26. 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 2001; 12: 825 830. 27. Nadali G, Tavecchia L, Zanolin E et al. Serum levels of the soluble form of the CD30 molecule identify patients with Hodgkin s disease at high risk of unfavorable outcome. Blood 1998; 91: 3011 3016. 28. Specht L. Prognostic factors in Hodgkin s disease. Semin Radiat Oncol 1996; 6: 146 161.