18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography Following Chimeric Antigen Receptor T-cell Therapy in Large B-cell Lymphoma

Abstract

Purpose: ¹⁸F-Fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) is a well-established imaging modality to assess responses in patients with B-cell neoplasms. However, there is limited information about the utility of FDG PET/CT after chimeric antigen receptor T-cell (CART) therapies for large B-cell lymphomas. In this retrospective analysis, we aimed to evaluate how FDG PET/CT performs in patients receiving commercially available anti-CD19 CART therapies for relapsed/refractory (r/r) large B-cell lymphomas. In addition, we examined the time to repeat scan and the rate of pseudoprogression within this population. Lastly, the rates of radiographic response to CART therapy using FDG PET/CT are reported.

Procedures: The pre-treatment and post-treatment scans were analyzed from a selected cohort of 43 patients from a single institution. Patients were stratified by diagnosis of either a first occurrence of diffuse large B-cell lymphoma: de novo diffuse large B-cell lymphoma (DLBCL); or a transformed diffuse large B-cell lymphoma arising from indolent non-Hodgkin lymphoma (t-iNHL).

Results: More patients received CART therapy for DLBCL than t-iNHL (65 % vs 35 %). FDG PET/CT had a 99 % sensitivity and 100 % specificity for detecting recurrent disease in this group. The median time to initial response assessment was 86 days (IQR 79-91; full range 24-146) after infusion. There were no biopsy-proven cases of pseudoprogression identified. In this selected group of patients, the overall response rate by Lugano 2014 criteria was 56 %. All patients with a partial response (N = 6) eventually progressed despite additional therapy.

Conclusions: Due to its excellent test characteristics and ability to detect asymptomatic disease, routine surveillance with PET/CT at 3 months after CART infusion is supported by our data. Earlier PET/CT may be of value in select situations as we did not find any cases of pseudoprogression.

Keywords: Chimeric antigen receptor T-cell; FDG PET/CT; Immune imaging; Lymphomas; Reporter gene imaging.

Figure 1.

The lesion identities and overall characteristics of the final patient cohort. (A) Outlines how the identity of each FDG-avid index lesion was determined. N represents the number of lesions within that category. If a patient had a biopsy showing disease progression, all other growing index lesions for that patient were counted as biopsy-proven disease progression. (B) A table describing the general characteristics of the study population. Ranges provided in the table represent the interquartile range. Full range for median time to follow up after CAR T was 2-20 months, for median time to CAR T infusion after baseline scan was 6-400* days and for median time to response assessment was 24-146 days. *400 days time to CART was a patient outlier related to the pre-treatment PET timing. The patient received a CT abdomen with contrast 46 days prior to CART initiation because the extent of their disease was a singular mesenteric lymph node. On that cross-sectional imaging exam, the patient did not have any new areas of lymphadenopathy and their known lymph node was similar in size (1.4 x 1 cm from 1.5 x 1.2 cm).

Figure 2.

2 x 2 table showing the number of true positive (TP), false positive (FP), false negative (FN), and true negative (TN) lesions after infusion with anti-CD19 CARTs.

Figure 3.

A diagnostically challenging case for a patient with DLBCL. (A) The first column depicts a patient’s baseline non-contrast CT and FDG PET/CT fusion before anti-CD19 CART therapy. The second column depicts the same region 3 months after treatment. Two foci of FDG uptake on the pretreatment scan suggested active lymphoma in the right hilum and left lung. On the post-treatment scan, the left lung lesion no longer demonstrates increased FDG uptake. However, FDG uptake in the right hilum persisted. Transbronchial biopsy was performed. (B) H&E stain showing a cluster of large atypical cells with abnormal nuclei. (C) PAX5 stain showing that most cells within the atypical cluster are B-cells. (D) CD3 stain showing minimal T-cells within the atypical cell cluster.

Figure 4.

Patient response to CART infusion on initial assessment and data cutoff. (A) A table listing all responses by PET/CT on initial reassessment. (B) A plot illustrating the “all-or-none” response pattern by comparing each patient’s initial response assessment after CART therapy to the last recorded response before data cutoff. The endpoints stratify patients by their last known survival status up to data cutoff. The majority of patients with a CR on their initial response assessment had a CR on their most recent PET/CT. Three patients with CR developed PD. Additionally, four patients with initial PD converted to CR after treatment with additional therapy. The 4 patients who converted from post-CART PD to CR w/ salvage therapy received: 1) Lenalidomide + Rituximab, 2) Radiation, pembrolizumab, obinutuzumab, polatuzumab, 3) Brentuximab vedotin, 4) Rituximab + bendamustine + polatuzumab Every patient with an initial PR advanced to PD despite additional therapy in majority of these patients.

Scheme 1.

The cellular mechanism and time course for pseudoprogression. Following immunotherapy, various immune cells infiltrate a tumor. At this stage, the tumor could increase in size and metabolic activity due to the increased cellularity and local immune activation. Imaging interpretation at this time can be challenging given the overlap in imaging findings. After the inflammatory infiltrate subsides, the tumor can either demonstrate a response to treatment or disease progression.

Scheme 2.

A PET reporter gene could serve as a therapeutic biomarker portending response to CART therapy and guide use of concurrent therapy. After infusion with CARTs labeled with a PET reporter gene, an early follow up scan could be used to stratify patients in three different degrees of response. Patients with moderate levels of reporter expression may have CARTs present in the tumor, but with anergic/exhausted phenotype limiting treatment efficacy. Therefore, these patients could be selected to receive a concurrent therapy, such as bispecific antibodies, immune checkpoint inhibitors (ICIs), to augment their CARTs and potentiate successful treatment.