MRD Detection in Mantle Cell Lymphoma: Flow Cytometry Is Comparable to Consensus Primer IgH-PCR at Initial Staging, but Less Sensitive Than PCR after Combined Immuno-Chemotherapy.

The significance of 4-color flow cytometry (4C-FC) and PCR for initial staging and follow-up in patients (pts.) with mantle cell lymphoma (MCL) is currently poorly defined.

We compared 4C-FC and consensus primer IgH-PCR (cPCR) for minimal residual disease (MRD) assessments in 180 samples from 75 MCL pts. treated with Rituximab-(R)-CHOP, R-FC or alternating R-CHOP/R-DHAP in the current EU MCL network trials. MCL was diagnosed by pathology in all cases and additionally confirmed by FISH of peripheral blood (PB) or bone marrow (BM) in 58% of pts. The MRD results were compared to histological BM examination and clinical staging of the pts. For 4C-FC, all 180 samples were stained with kappa/lambda/CD19/CD5 to assess the light chain restriction (LCR). LCR had a maximum sensitivity of up to 4 x E-4. 157 MCL samples were additionally stained with CD23/CD22/CD19/CD5 and CD79b/CD22/CD19/CD5. CD23 expression was low on all MCL cells. Compared to benign CD19+CD5+ lymphocytes, CD79b was low in 65% and overexpressed in 5% of samples, while CD22 was always underexpressed. Those phenotypic aberrations were combined to detect MCL cells at a maximum sensitivity of 9 x E-4 (CD23/CD22) and 5 x E-4 (CD79b/CD22). Overall, there was evidence for MCL cells by flow in 38 % (LCR), 43 % (CD79b/CD22), and 38 %(CD23/CD22) of samples, respectively. The MCL levels detected by LCR and by phenotypic aberrations correlated well (r =0.99, p<0.0001).

At initial diagnosis 88 samples were analyzed (67 pts., 63 PB, 25 BM). 80/88 samples (91%) were positive by 4C-FC. The 4C-FC+ samples included 22 samples from 17 pts. without histological BM involvement (26 % of all pts. ). Since 16 of those 17 pts. had clinical stage 2 or 3 4C-FC demonstrated more extensive disease than assumed by clinical staging. MRD levels were significantly higher in pts. with histological BM involvement (median: 8 x E-2) than in pts. without BM involvement (median: 4 x E-3, p < 0.0001). 75/88 initial diagnosis samples (85%) from 59 pts. were cPCR+. 4C-FC and cPCR yielded discordant results in 8 samples without histologically proven BM involvement (7 samples 4C-FC+/cPCR-; 1 sample 4C-FC-/cPCR+). Overall 4C-FC and cPCR were concordant at initial diagnosis in 80/88 samples (91%).

92 samples were obtained from 40 treated pts. during therapy or short-term follow up (maximum 7 months after end of therapy). The treatment led to a profound reduction in absolute B cell numbers in PB (from mean 2355/μl to mean 3/μl) and BM (mean 6632/μl to mean 25/μl). 20 of those 92 samples (21%) were positive by cPCR. 5/92 samples (7 %) showed evidence of residual MCL cells by 4C-FC (median MCL level: 1xE-3). These samples were positive by cPCR as well.

4C-FC approaches and cPCR are equally suitable to assess low level PB and BM involvement in MCL patients at initial diagnosis. Both methods detect low level disease even in pts. without histological BM involvement and might therefore improve staging accuracy in MCL. In the setting of B lymphopenia after Rituximab plus chemotherapy 4C-FC detects MCL cells in a minority of cases only and is less sensitive than cPCR.