- Top of page
- Patients and methods
The prognostic value of 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) in the assessment of post-treatment residual masses in patients with Hodgkin's disease (HD) or non-Hodgkin's lymphomas (NHL) was evaluated. We prospectively studied 58 patients with HD (n = 43) or NHL (n = 15) who had post-therapeutic complete remission with residual masses (CRu) indicated by computerized tomography. Analysis of 62 residual locations by FDG-PET was performed separately for HD and NHL. Patients with a PET-positive residual mass [standardized uptake value (SUV) > 3] had a recurrence rate of 62·5% (5/8 patients), whereas patients with PET-negative residual mass (SUV ≤ 3·0) showed a recurrence rate of 4% (2/50 patients, P = 0·004). A positive FDG-PET study correlated with a significantly poorer progression-free survival (P < 0·00001). No recurrence occurred in any of the 39 HD patients with a negative PET scan (negative predictive value, 100%). Four out of four NHL patients with a positive PET study relapsed (positive predictive value, 100%). In conclusion, FDG-PET is a suitable non-invasive method with a high degree of accuracy in the prediction of early recurrence in lymphoma patients with CRu.
Despite the good response to therapy, residual mass can be demonstrated radiologically in up to 80% of patients with Hodgkin's disease (HD) and in up to 40% of non-Hodgkin's lymphoma (NHL) patients after completion of treatment (Radford et al, 1988; Surbone et al, 1988). To account for the difficulty in assessing residual mass viability, the Cotswolds Meeting led to a new category for evaluation of treatment response: complete remission unconfirmed/uncertain (CRu) (Lister et al, 1989), for which the relapse risk is reported to be about 20% (Jochelson et al, 1985; Canellos, 1988; Radford et al, 1988; Surbone et al, 1988). The management of these residual masses is a dilemma for the oncologist. Some patients might receive unnecessary additional treatment and are thus exposed to avoidable early and late toxicity.
Owing to this diagnostic dilemma, there are high expectations of positron emission tomography (PET) with the radioactively labelled glucose analogue 2-[F-18]fluoro-2-deoxy-D-glucose (FDG) which enables quantitative measurement of glucose metabolism in tumour tissue (Haberkorn et al, 1993; Steinert et al, 1995; Rigo et al, 1996; Wahl, 1999; Avril et al, 2000; Flamen et al, 2000; Halter et al, 2000; Shah et al, 2000). The FDG accumulation is directly correlated with the raised glucose metabolism of malignant cells (Warburg, 1956; Gallagher et al, 1978). As it enables detection of viability, FDG-PET appears to be especially suitable for the differentiation of active lymphoma tissue and inactive scar tissue. To date, there have only been a few prospective studies on the evaluation of residual masses after lymphoma therapy (Kostakoglu & Goldsmith, 2000). To our knowledge, no study group has differentiated between HD and NHL to assess the diagnostic accuracy for detection of active residual lymphoma. Therefore, we initiated this prospective study to enable a separate investigation of the prognostic value of PET for both lymphoma entities.
- Top of page
- Patients and methods
Early evaluation of treatment response for malignant lymphoma is important for the prognosis and further individual patient management. The interpretation of persistent residual masses detected by CT is a major clinical problem, especially in mediastinal and abdominal bulk. The remission status is evaluated by CT according to standardized guidelines (Cheson et al, 1999). However, on the basis of morphology (ultrasound, CT, MRI), a residual mass cannot be differentiated from scar tissue because size is not closely related to the risk of recurrence (Rodriguez-Catarino et al, 2000) and therefore a residual mass does not necessarily entail residual disease (Fuks et al, 1982; Canellos, 1988; Surbone et al, 1988; Van den Berg et al, 2000). However, Jerusalem et al (1999a) described a higher rate of recurrence in the presence of a post-therapeutic residual mass.
MRI has not been able to fulfil the expectations of a reliable distinction between malignant and fibrotic or necrotic lymphoma tissue (Hill et al, 1993; Devizzi et al, 1997; Maisey et al, 2000). The diagnostic relevance of 67Gallium scintigraphy is controversial, but appears to be helpful, particularly in patients with a mediastinal residual mass (Front et al, 1990; Salloum et al, 1997; Setoain et al, 1997; Ulusakarya et al, 1999; Zinzani et al, 1999a). However, the need of pretherapeutic baseline investigation in order to identify a gallium-avid lymphoma is a disadvantage. PET has additional important advantages compared with 67Gallium scintigraphy. It has a higher detection rate in the abdomen, the patient's radiation exposure is significantly less, and the logistical requirement is relatively minor (Stroobants et al, 1997; Zinzani et al, 1999b).
In our study, we prospectively investigated the diagnostic relevance of FDG-PET in lymphoma patients with CT-documented residual masses. The most important result of the study is the prediction of an early recurrence in lymphoma patients with post-therapeutic residual mass (accuracy 91%). With a median follow-up of 37 months, patients with a PET-positive residual mass (SUV > 3) had a rate of recurrence of 62·5% (5/8 patients), whereas PET-negative patients (SUV ≤ 3) showed a rate of recurrence of only 4% (2/50 patients; P < 0·05). The recurrences occurred after a median follow-up of 11 months after the end of therapy (median 6 months after PET). Positive PET studies correlated with a significantly poorer PFS (P < 0·00001).
The great heterogeneity of malignant lymphoma necessitated a separate evaluation of HD and NHL. The number of patients with NHL investigated prospectively with appropriate clinical correlates is quite small in the present study. A high positive predictive value (PPV) of 100% for NHL residual masses (4/4 recurrences) contrasts with the low PPV of 25% for HD residual masses. Jerusalem et al (1999b) also described relatively high PPV for NHL residues (6/6 recurrences), Mikhaeel et al (2000) reported 5/5 recurrences in NHL, and Zinzani et al (1999b) reported 13/13 recurrences in NHL with abdominal residual mass in their prospective studies. Spaepen et al (2001) reported 26/26 relapses after first-line chemotherapy in NHL with persistent abnormal FDG uptake (12/26 residual disease according to conventional diagnostic methods); in relapsing patients, the PFS was significantly shorter after a positive scan than after a negative scan.
The low PPV in HD residues in our study can be explained both by the low rate of recurrence of 2% in our HD patients (1/43 patients developed recurrence) as well as by a relatively higher rate (7%) of false-positive PET studies (3/43 patients) in this lymphoma entity. Cremerius et al (1999) showed that the PPV rises with an increased risk of recurrence (stage IV, patients with relapse, > 2 cycles of chemotherapy). However, false-positive PET results (SUV > 3) were found in three HD patients in our study. The inclusion of clinical information in PET-positive cases helps to interpret questionable findings, especially when these are located outside the residual mass. In our study, the percentage of false-positive residual findings was 5% in the positive (SUV > 3) PET scans evaluated (3/58). In three HD patients (two positive, one questionable finding in the residual mass), scar tissue was found in two cases after sternotomy, and an eosinophilic granuloma was found in one case. In other studies inflammatory processes such as pneumonia, tuberculosis or abscesses were reported as the causes of false-positive findings (Kubota et al, 1992; Strauss, 1996; de Wit et al, 1997). The thymus-gland hyperplasia that has been described in children and young adults (Weinblatt et al, 1997) did not occur in any of our cases.
The clinical impact of the SUV is controversial owing to the large range of variation of SUV in malignant lymphomas (de Wit et al, 1997; Stumpe et al, 1998; Cremerius et al, 1999). Consequently, PET findings were established exclusively on the basis of visual criteria in restaging of malignant lymphomas in most PET studies. In our study, calculation of SUV for visually suspect FDG accumulations was carried out in all scans. The cut-off of SUV > 3 used was based on the experience of other authors with various tumour entities (Patz et al, 1993; Cheon et al, 1999). Whereas a negative interpretation of questionable findings had only a minor effect on the sensitivity (88% versus 75%, not significant), the prognostic value of PET was improved. In the Hodgkin's disease group the specificity rose from 64% to 93% (P < 0·05) and accuracy from 65% to 93% (P < 0·05). This cut-off proved to be reliable, especially in findings outside the residual mass. By reduction of the numerous false-positive findings, the PPV rose from 6% to 25% in HD patients. In our study there were no significant changes for the NHL group owing to the smaller number of cases evaluated. Negative PET findings were highly predictive for the absence of disease recurrence (HD 100%, NHL 88%) in both tumour entities in our study. This is consistent with the results of earlier studies (de Wit et al, 1997; Bangerter et al, 1998; Spaepen et al, 2001). One false-negative result which may be attributed to a diabetic situation (Lindholm et al, 1993) was found in a NHL patient. A negative PET scan is an important contribution in the management of these patients owing to its prognostic value. It may help to obviate further treatments and may prolong the intervals between CT scans. The fear of patients that their disease will recur at the site of the residual mass might be reduced by negative PET scan.
In summary, our results indicate that FDG-PET could be used as a helpful component of therapy assessment in patients with CRu. We propose that all PET scans should comprise attenuation correction including calculation of SUV in order to minimize false-positive findings, especially in HD patients. However, owing to the small number of prospective studies available, present conventional radiological investigations remain indispensable.