Overall survival and EZH2 expression in mantle cell lymphoma

A gain of function mutation in the enhancer of zeste homolog 2 (EZH2) has been associated with the development of both follicular lymphoma and diffuse large B-cell lymphoma (DLBCL) of germinal center B-cell-like type. EZH2 is the catalytic center of the polycomb repressive complex 2 (PRC2) that is responsible for gene silencing via histone methylation.¹ You can read more about this topic, here.

Inhibition of EZH2 has demonstrated promising clinical activity in patients with FL, contributing to the plethora of novel targeted agents in FL. However, there remains little information regarding the role of EZH2 expression in mantle cell lymphoma (MCL), and so Diana Martinez-Baquero and colleagues sought to evaluate the prevalence of EZH2 mutations in a MCL population.¹

Study design

• Systematic analysis of EZH2 expression across samples from 166 patients with MCL
• Immunohistochemical staining of EZH2 was performed on the two tissue microarrays constructed from tumor rich regions of each patient sample that were identified via review of hematoxylin and eosin-stained slides
• With the aim to correlate EZH2 expression with proliferation rate, alternative PRC2 components, genomic markers, and clinical outcome, the investigators also obtained immunohistochemical readouts for CD5, MHC-I, CD10, CD20, Ki-67, and p53, cyclin D1, H3K27me3, EED, SUZ12, and SOX11
• Gene expression profiling—using RNA derived from formalin-fixed, paraffin-embedded tissue—was performed on 37 patients

Baseline characteristics

Of the 166 patients with MCL who were identified, 29 (17%) were women; the median age of the cohort was 61 years (39‒88). Table 1 lists the baseline characteristics for the total cohort.

Table 1. Baseline characteristics*

CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisolone; CVAD, cyclophosphamide, vincristine, doxorubicin, and dexamethasone; MIPI, mantle cell lymphoma-international prognostic index. *Adapted from Martinez-Baquero et al.1
Characteristic	Total cohort (N = 166)
Age, years, n (%)
<60	75 (45)
≥60	91 (55)
Site, n (%)
Lymph node	125 (75)
Spleen	5 (3)
Tonsils	14 (8)
Extranodal sites	22 (13)
Stage, n (%)
I/II	13 (8)
III/IV	146 (92)
MIPI score, n (%)
Low	59 (45)
Intermediate	42 (32)
High	29 (22)
Pathology, n (%)
Classic	99 (60)
Aggressive	67 (40)
Pattern, n (%)
Mantle zone	2 (1)
Diffuse	88 (53)
Nodular	76 (46)
Treatment regimen, n
hyperCVAD ± rituximab	106
CHOP ± rituximab	31
Bendamustine + rituximab	5
Observation	14
Not available	10

Treatment response was available for 142 of the 166 patients:

• Complete remission was observed in 109 patients, 68 (62.4%) of whom eventually relapsed
• Partial remission was seen in 8 (5.6%) patients
• Stable disease was seen in 5 (3.5%) patients
• Progressive disease was seen in 11 (7.7%) patients

Results

EZH2 expression: Clinical significance

EZH2 expression (≥40%) was seen in 57 (38%) of the 150 patients in whom it was assessed; when compared with EZH2-negative patients, those with EZH2-positive MCL had more aggressive histology, high Ki-67 proliferation rate, and p53 overexpression (all p <0.001; Table 2). No significant associations were observed between EZH2 expression and other patient/disease characteristics.

Table 2. Clinicopathologic features associated with EZH2 expression in patients with MCL*

MIPI, mantle cell lymphoma-international prognostic index. *Adapted from Martinez-Baquero et al.1
Characteristic	EZH2+ (n = 57)	EZH2− (n = 93)	p value
Female gender, n	10	15	0.8214
Age, years, median (range)	61 (42‒87)	60 (39‒88)	0.347
<60, n	28	47	0.8664
≥60, n	29	46
Site, n			0.6317
Lymph node	44	69
Spleen	3	2
Tonsils	4	8
Extranodal sites	6	14
Stage, n (%)			0.7675
I/II	5 (9)	7 (8)
III/IV	50 (91)	82 (92)
MIPI, n (%)			0.2839
Low	18 (42)	38 (50)
Intermediate	13 (30)	26 (34)
High	12 (28)	12 (16)
Pathology, n (%)			<0.0001
Classic	20 (35)	66 (71)
Aggressive	37 (65)	27 (29)
Pattern, n (%)			0.2826
Mantle zone	0 (0)	2 (2)
Diffuse	33 (58)	44 (47)
Nodular	24 (42)	47 (51)
Biomarkers, n (%)
SOX11+	48/49 (98)	64/67 (96)	0.4774
Ki-67 high	41/57 (72)	17/91 (19)	<0.0001
P53 overexpression	15/35 (43)	1/48 (2)	<0.0001
CD5+	57/57 (100)	86/91 (95)	0.1568
CD10+	2/38 (5)	0/71 (0)	0.1194
MHC-I loss	1/40 (3)	0/56 (0)	0.1734
MHC-II loss+	3/36 (8)	4/50 (8)	>0.9999

EZH2 expression: Weak or no correlation

When compared with other components of the PRC2 complex—embryonic ectoderm development (EED) and suppressor of zeste 12 (SUZ12), specifically—there was a weak correlation with EZH2 expression. No correlation was found between EZH2 expression and expression of trimethylated histone 3 at lysine 27 (H3K27me3), another component of PRC2, nor between EZH2 expression and loss of MHC-I or MHC-II (Table 2).

EZH2 expression: Association with overall survival

Survival data was available for 148 patients, with a median follow up of 5.3 years. Inferior overall survival (OS) was seen in patients with EZH2 expression compared to those without EZH2 expression (median OS, 3.9 years and 9.4 years, respectively; p <0.001), though there was no significant difference in relapse-free survival (RFS) between these groups (median RFS, 3.8 years and 2.7 years, respectively; p = 0.7430).

On the subgroup analysis by treatment regimen:

• In the group treated with hyperCVAD ± rituximab, patients who expressed EZH2 had inferior OS compared to patients who did not express EZH2 (median OS, 5 years and 11.7 years, respectively; p = 0.0008).
• A similar trend was observed in the group of patients treated with CHOP ± rituximab (p = 0.0721)

On the subgroup analysis by histology:

• In patients with MCL with classic histology, EZH2 expression was associated with a poorer outcome compared with non-expression of EZH2 (median OS, 4.6 years and 9.6 years, respectively; p <0.001).
• In patients with MCL with aggressive histology, EZH2 expression was also associated with a poorer outcome compared with non-expression of EZH2 (median OS, 3.7 years and 7.9 years, respectively; p = 0.0458).
• There was a trend toward poorer prognosis in patients with low Ki-67 MCL and EZH2 expression, and no correlation between high Ki-67 and EZH2 expression.

Multivariate analysis

EZH2 expression, high (≥30%) Ki-67 proliferation rate, p53 overexpression (≥50%), and aggressive histologic variants showed increased hazard ratios for OS in univariate analysis. However, in multivariate analysis, only Ki-67 proliferation rate demonstrated a marginal p value (p = 0.063); EZH2 expression, p53 expression, and aggressive histology failed to remain as independent predictors of OS.

Conclusion

In this analysis, EZH2 expression in MCL was associated with a high Ki-67 proliferation rate, more aggressive histology, p52 overexpression, and poorer OS; in the case of aggressive MCL variants, this association was significant. However, EZH2 expression was not an independent risk factor of poorer OS in multivariate analysis. Despite this, the authors note that EZH2 may be useful for the identification of patients with high-risk MCL, especially in the setting of high proliferation rate, and this can identify patients who may benefit from treatment with EZH2 inhibitors.