Abstract

Abstract 2646

Background:

PET/CT scan is a widespread modality in both clinical settings and clinical trials in the evaluation of diffuse large B-cell lymphoma (DLBCL). The concordance among interpreters of PET/CT scan is important for basing clinical decisions on these results. In order to act properly on the results, the response evaluation according to interpretation of PET/CT reports has to correlate with specific clinical end points including outcome. As of today we have no evidence for that and clinical studies are only based on the assessment of nuclear medicine-physicians.

Aim:

To evaluate the clinician-based interpretation of PET/CT reports in newly diagnosed DLBCL at mid-therapy(I-PET) and end-therapy(E-PET) in terms of concordance and prognostic impact.

Method:

Patients were considered eligible for inclusion in this retrospective study if they fulfilled the following criteria: I) newly diagnosed with de novo DLBCL, II) age ≥15 years, III) treated with R-CHOP or R-CHOP like treatment with or without addition of CNS prophylaxis and radiotherapy and IV) evaluated with PET/CT at mid-therapy and/or end-therapy. Patients with primary CNS lymphoma and composite lymphoma histology, HIV-associated lymphoma and transplant related lymphoproliferative disease were excluded from this study. All DLBCL patients diagnosed between September 2005 and December 2009 at eight Danish hematology centers were screened for eligibility.

Nine expert hematologists were asked to interpret PET/CT reports. Each report was independently evaluated by three hematologists. The assessments of reports were performed without any clinical information. PET/CT reports were labeled positive or negative if all three interpreters independently agreed. All others were considered indeterminate.

Results:

A total of 434 patients and 617 PET/CT reports were included in the study. The distribution of PET/CT report interpretation is shown in Table 1.

The median follow-up time was 3.4 years (range 0.3 – 6.7 years). The progression free survival (PFS) and overall survival (OS) for the I-PET reports were not significantly different between the indeterminate and negative results (p=0.6). However, patients with an indeterminate result had a worse PFS and OS according to the E-PET reports (p=0.09 and p=0.006, respectively). Patients with a positive I-PET and/or E-PET report both had a significantly lower PFS (p<0.0001) and OS (p<0.0001) compared to the two other groups. Survival curves are shown in Figure 1.

Discussion:

PET/CT evaluation in newly diagnosed DLBCL needs not only acceptable predictive values for outcome prediction to reinforce the decision to modify treatment strategy, but also high concordance between the clinicians who make the decisions. Failing to achieve this will make PET an unsuitable tool for this purpose. In the present study, we found a high number of indeterminate evaluations according to both I-PET and E-PET reports. There was no significant prognostic difference between the negative and the indeterminate group of the I-PET. However, we observed a significant difference in outcome between patients with a negative and indeterminate E-PET. Patients with a positive E-PET and I-PET had a very poor prognosis.

This study demonstrates that binary visual PET response criteria for de novo DLBCL leads to a considerable amount of indeterminate interpretations in the clinician based setting. The study also underlines the importance of consistent PET reports and a multidisciplinary approach for the evaluation of treatment response in DLBCL.

Table 1:

Distribution of 617 PET/CT reports by hematologist interpretation

IndeterminateNegativePositive
I-PET (n=241) 59% (n=142) 30% (n=73) 11% (n=26) 
E-PET (n=376) 37% (n=186) 41% (n=153) 10% (n=37) 
IndeterminateNegativePositive
I-PET (n=241) 59% (n=142) 30% (n=73) 11% (n=26) 
E-PET (n=376) 37% (n=186) 41% (n=153) 10% (n=37) 
Table 2:

Survival estimates according to results of I-PET and E-PET

I-PET (n=241)Indeterminate (n=142)Negative (n=73)Positive (n=26)
2YRS PFS estimate (95% CI) 85% (78-90%) 90% (81-95%) 52% (31-69%) 
2 YRS OS estimate (95% CI) 87% (81-91%) 89% (79-94%) 58% (37-74%) 
E-PET (n=376) Indeterminate (n=186) Negative (n=153) Positive (n=37) 
2YRS PFS estimate (95% CI) 86% (80-90%) 95% (90-97%) 36% (21-52%) 
2 YRS OS estimate (95% CI) 89% (84-93%) 97% (92-99%) 41% (25-56%) 
I-PET (n=241)Indeterminate (n=142)Negative (n=73)Positive (n=26)
2YRS PFS estimate (95% CI) 85% (78-90%) 90% (81-95%) 52% (31-69%) 
2 YRS OS estimate (95% CI) 87% (81-91%) 89% (79-94%) 58% (37-74%) 
E-PET (n=376) Indeterminate (n=186) Negative (n=153) Positive (n=37) 
2YRS PFS estimate (95% CI) 86% (80-90%) 95% (90-97%) 36% (21-52%) 
2 YRS OS estimate (95% CI) 89% (84-93%) 97% (92-99%) 41% (25-56%) 
Figure 1.

Overall survival for I-PET and E-PET according to interpretation of PET/CT reports

Figure 1.

Overall survival for I-PET and E-PET according to interpretation of PET/CT reports

Disclosures:

No relevant conflicts of interest to declare.

This icon denotes a clinically relevant abstract

Author notes

*

Asterisk with author names denotes non-ASH members.