High Incidence of False Positive PET Scans in Patients with Aggressive Non-Hodgkins Lymphoma Treated with Rituximab-Containing Regimens.

Introduction: 18F-Fluorodeoxyglucose positron emission tomography (PET) is increasingly used for initial staging and response assessment in patients with Non-Hodgkins lymphoma (NHL) and is a powerful predictor of relapse and survival in this setting. All reported PET studies, however, were performed in the pre-rituximab era. Little is known about the predictive abilities of PET in rituximab-treated patients.

Patients and Methods: Patients with aggressive NHL treated at the University of Miami between September 2002 and January 2006 that had baseline and follow-up PET studies were included in the study. Clinical characteristics at presentation, PET and CT scan results, and outcomes were reviewed. PET studies were defined as positive if greater than physiological activity was observed.

Results: A total of 33 patients with the following histologies were included: diffuse large B cell lymphoma (23 patients), Mantle-cell lymphoma (5), Peripheral T-cell lymphoma (3) and NK/T cell lymphoma (2). The median age was 54 years (range 21–92) and 19 were male (58%). Six patients had an IPI of 0, 13 an IPI of 1, 10 an IPI of 2, 3 an IPI of 3, and 1 an IPI of 4. All patients with CD 20+ lymphomas were treated with rituximab containing regimens. Median follow up was 18 months (range 6–47). All patients had positive PET scans at diagnosis. Nine of 25 (36%) mid-therapy PET studies were positive. Four of these patients exhibited persistent PET positivity, of which 2 eventually died of progressive disease. In the other 2 patients, disease recurrence was biopsy-proven in one patient, while the other received further chemotherapy and is presently alive and in CR. Upon therapy completion, the remaining 5 converted to negative studies and 4 of these patients are in continuous complete remission (CR) for 8 to 19 months. The fifth patient relapsed 6 month after treatment. The relapse positive predictive value (PPV), negative predictive value (NPV), sensitivity (Se) and specificity (Sp) of the mid-therapy PET were 55.5% (95% CI 23–85%), 81.2% (95% CI 53–95%), 62.5% (95% CI 26–90%), and 76.4% (95% CI 50–92%), respectively. Upon therapy completion, 5 patients with residual masses on CT scans had positive PET scans, including 2 patients with positive mid-therapy PET scans. Biopsy was performed in 1 and did not demonstrate lymphoma, 1 was followed-up for 10 months without evidence of disease and 3 demonstrated disease progression and were salvaged with additional therapy. Seven patients with a negative post-therapy PET demonstrated residual masses by CT scan criteria. Two of these patients relapsed while the remaining 5 are in CR for 8 to 19 months. In two of these patients in CR, PET scans performed 3 to 6 months after therapy completion became positive. However, biopsy revealed inflammatory changes with no evidence of lymphoma. Collectively, for the post-therapy PET: PPV, NPV, Se, Sp were 71.4% (95% CI 30–95%), 79.2% (95% CI 57–92%), 50% (95% CI 20–80%), and 90.4% (95% CI 68–98%), respectively.

Conclusions: Our study demonstrated an acceptable NPV and Sp but low PPV and Se for mid- and post- therapy PET in patients with aggressive NHL treated with rituximab. Furthermore, negative mid-therapy PET did not assure continuous response to therapy. Our findings also emphasize the importance of confirmatory biopsies in patients with post-therapy positive PET scans prior to initiation of second line therapy.