Normalized CCND1 expression has prognostic value in mantle cell lymphoma



This article is corrected by:

  1. Errata: Erratum Volume 159, Issue 3, 380, Article first published online: 10 October 2012

Mantle cell lymphoma (MCL) is an aggressive B-cell non-Hodgkin lymphoma (NHL; Swerdlow et al, 2008). The molecular hallmark for diagnosis of MCL is the t(11;14)(q13;q32) translocation, resulting in constitutive over-expression of CCND1 (Cyclin D1; Campo et al, 1999; Swerdlow et al, 2008). The current median survival in patients with MCL is only 4–5 years (Herrmann et al, 2009), one of the poorest overall survival rates of all NHLs. Yet, some MCL patients have a more indolent course (Martin et al, 2009; Weigert et al, 2009). While several predictors of outcome for MCL have been proposed, their reliability and reproducibility are not well established (Rosenwald et al, 2003; Martin et al, 2009; Weigert et al, 2009). The chronic lymphocytic leukaemia overlap marker CD23 may be associated with a better prognosis in MCL, but this has not been confirmed (Kelemen et al, 2008). High Ki-67 expression and variant blastoid morphology are useful tissue correlates of poor prognosis, but these measures cannot be applied to patients diagnosed in the leukaemic state (Determann et al, 2008). Thus, at this time, there are no laboratory assays that stratify MCL patients into conservative versus aggressive treatment regimens.

The primary aim of this study was to evaluate the prognostic utility of normalized quantitative CCND1 mRNA values in the blood or marrow of patients with newly diagnosed MCL. This novel quantitative CCND1 assay has been validated for establishing a diagnosis of MCL in blood/bone marrow (Hui et al, 2003; Howe et al, 2004), but its use as a prognostic marker has not been evaluated. We further compared the normalized CCND1 assay with CD23 expression and circulating levels of malignant B-cells.

All cases tested between October 2002 and May 2010 that demonstrated an elevated normalized CCND1 mRNA value were eligible for inclusion (n = 74). Any cases in which the tissue source was not blood or marrow (n = 6) was subsequently excluded, and inclusion in the final study was restricted to only those patients with a new diagnosis of MCL and no prior treatment (n = 40). The institutional review board approved the review and use of patient information in this study.

The normalized CCND1 mRNA assay is a relative fold increase (RFI) determined against the background of B-cells (Howe et al, 2004). The RFI method has been shown to strongly correlate with results obtained using fluorescent in-situ hybridization (FISH) with CCND1 probes (Hui et al, 2003). The quantitative assay is briefly described here. RNA was extracted using RNeasy reagent (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. All primers and probes were synthesized as previously described (Howe et al, 2004) and reaction mixtures contained primers for CCND1 and CD19. Reverse transcription and polymerase chain reaction were performed using an ABI PRISM 7700 Sequence Detector (Applied Biosystems, Carlsbad, CA, USA). Critical threshold (C t) cycle numbers were obtained for amplification of CCND1 and CD19. The ΔC t values were then obtained by subtracting the C t value of CD19 from the C t value of CCND1.

The characteristics of our 40 patients with new a diagnosis of MCL in blood (n = 25) or marrow (n = 15) are presented in Table 1. All 40 patients met the molecular criteria for diagnosis of MCL as previously described (Howe et al, 2004). All 40 patients were confirmed to have CD19+ CD5+ monoclonal B-cell NHL using flow cytometry; the following analyte-specific reagent antibodies (BD Biosciences, Franklin Lakes, NJ, USA) were used in all cases: CD19, CD20, CD5, CD10, CD23, FMC-7, κ, and λ. We established a normalized mRNA RFI ratio cut-off of 75 was based on median CCND1 mRNA levels in the blood or marrow of the first ten patients assayed for evaluation. Patients with high CCND1 mRNA (RFI ≥ 75; n = 25) in blood or marrow, were found to have an overall median survival of 36 months from the time of diagnosis, while patients with RFI < 75 (n = 15) had not yet reached their median survival at 90 months (P = 0·0286; Fig 1A). Thus, normalized CCND1 mRNA levels in blood and marrow provided prognostic information for overall survival in newly-diagnosed, untreated MCL patients. For all 40 patients, there was no statistically significant difference in CCND1 mRNA levels between marrow and blood (141 ± 194 vs. 194 ± 218; mean ± standard deviation; P = 0·1988).

Table 1. Demographics of MCL patients based on CCND1 subsets
VariablesHigh CCND1 mRNA RFI (≥75)Low CCND1 mRNA RFI (<75)
Number (%) of patientsNumber (%) of patients
  1. a

    Data not available for all patients. There were no statistically significant differences in any variable between the patient groups.

  2. RFI, relative fold increase.

 Male14 (56)9 (60)
 Female11 (44)6 (40)
Median (range) age at diagnosis, years70 (42–92)62 (43–86)
Ann Arbor Stagea
Increased Lactate Dehydrogenasea
Performance Statusa
Figure 1.

(A) Overall survival in MCL patients with a high CCND1 mRNA level is significantly worse than MCL patients with low CCND1 mRNA level. The CCND1 hazard ratio for survival was 2·853 (95% confidence interval = 1·116–7·293). The average duration of follow-up for the high and low groups was 23·5 and 28 months, respectively (range 1–92 months). (B) Quantitative CCND1 mRNA levels do not correlate with the absolute number of monoclonal CD19+ CD5+ lymphocytes in blood. (C) The mean CCND1 mRNA level in CD23+ MCL patients, 115 ± 127 (standard deviation, SD), did not differ significantly from the mean in CD23− MCL patients, 214 ± 254 (SD). (D) The median survival for CD23+ MCL patients (48 months), was not significantly different from the median survival of those with a 23− immunophenotype (50 months).

Twenty-one of 25 patients with blood involvement had an available absolute CD19+ CD5+ lymphocyte count, which was examined relative to the quantitative CCND1 mRNA value (Fig 1B). There was no correlation (R 2 = 0·082; P = 0·288); hence, the quantitative measure of CCND1 mRNA was independent of the absolute number of circulating malignant MCL cells.

It has been previously suggested (Kelemen et al, 2008) that patients with CD23+ MCL (n = 18 patients in our study) have a superior outcome to those with CD23− disease (n = 22 patients in our study). When we compared these two subgroups based on CD23 expression, we found that their normalized CCND1 mRNA levels were not significantly different (P = 0·076; Fig 1C), nor did either group demonstrate a significant difference in overall survival (P = 0·779; Fig 1D).

The limitations in our study include a small sample size and the retrospective nature of our data collection. A prospective study with larger patient numbers is needed to validate and confirm our normalized cut-off value. Moreover, tissue studies are required in order to extend this observation to all MCL patients. Nonetheless, our preliminary data suggest that normalized CCND1 analysis in blood or marrow may be a valuable tool in assessing prognosis in newly-diagnosed, untreated MCL.

In addition to the prognostic information gleaned from quantitative CCND1 analysis, this assay has a similar sensitivity and specificity to traditional immunohistochemical stains or FISH probes for the characteristic t(11;14) translocation (Hui et al, 2003). Thus, normalized CCND1 mRNA assessment in tissue has the potential to prove superior to immunohistochemistry or FISH analysis because of its possible prognostic use.

In summary, this pilot study demonstrated that normalized CCND1 mRNA levels in blood and marrow provide prognostic information in patients with newly diagnosed MCL; by contrast, CD23 expression does not appear to correlate with overall survival.


Grant support: None.

Author contributions

AJS wrote the paper, collected, and analysed data. RT edited the paper and created the figures. HMR designed the research and wrote the paper. BRS designed the research and edited the paper. JGH performed the research. CAT designed the research and wrote the paper.

Conflicts of interest

None of the authors (AJS, RT, HMR, BRS, JGH, or CAT) report any conflict of interest with the data presented herein.