Clinicopathological features and outcome of chronic lymphocytic leukaemia in Chinese patients

Chronic lymphocytic leukaemia (CLL) is uncommon in Chinese population and its biology, genetics and treatment outcome in Chinese patients have not been comprehensively investigated. In this study, we studied the clinicopathological features and outcome of 212 Chinese patients with newly diagnosed CLL in Hong Kong and Singapore. The median age at diagnosis was 64 years. The majority of patients presented with early-stage disease (Binet stage A, 56.1%). Del(13)(q14) was the most frequent abnormality (41.7%) detected by fluorescence in situ hybridization (FISH) analysis. Del(17p) and TP53 gene mutations were detected in 7.8% and 8.2% of patients, respectively. MYD88 mutations were found at a higher frequency (11.5%) than expected. CLL with unmutated variable region of the immunoglobulin heavy chain genes (IGHV) occurred in only 31.2% of cases, and was associated with advanced-stage disease (p <0.01) and adverse FISH abnormalities (p<0.01). With a median follow-up of 39 months, the median overall survival (OS) was 108 months. The presence of del(17p) or TP53 mutations was associated with a significantly shorter time to first treatment and an inferior OS (p <0.01). Unmutated IGHV was also associated with a significantly shorter time to treatment (p <0.01). Among patients who required treatment, the median OS and progression-free survival (PFS) were 107 and 23 months, respectively. The presence of del(17p) was associated with a significantly inferior OS and PFS (p <0.01). In summary, Chinese CLL patients had similar genetic aberrations at diagnosis compared with those of Western populations. FISH abnormalities are major factors affecting outcome.


INTRODUCTION
Chronic lymphocytic leukaemia (CLL) is a mature B-cell neoplasm.It is characterized by an increased number of circulating clonal B-cells (>5 x 10 9 /L) with typical morphology and immunophenotype (CD5 + , CD23 + , CD79b weak/-, FMC7 weak/-, surface immunoglobulin dim ) [1].It is the most common leukaemia in the Western world, accounting for 25% of all leukaemias in adults.The ageadjusted annual incidence rate (AAIR) of CLL in USA is 4.75 per 100,000 persons [2].In contrast, CLL is much rarer in Asian populations, with an AAIR of less than 0.5 per 100,000 persons [3,4].Epidemiological studies also revealed a similarly low incidence of CLL in Asian migrants in North America, underscoring the importance of genetics in the pathogenesis of CLL [4,5].Nonetheless, the genetic basis for such ethnical differences remains undefined [6].
CLL is clinically heterogeneous.In some patients, the disease remains indolent and asymptomatic for years.Other patients, however, experience rapid disease progression even after aggressive treatment.The clinical staging systems devised by Rai et al [7] and Binet et al [8] are widely adopted for risk stratification and as an indicator for treatment in the International Workshop on CLL (iwCLL) guidelines [9].More recently, recurrent cytogenetic and molecular aberrations identified in CLL have provided important information on the genomic landscape of the disease.With karyotyping and fluorescence in situ hybridization (FISH), a hierarchal prognostic model has been proposed [10].Deletion of the short arm of chromosome 17, del(17p), is associated with the worst prognosis, which is not ameliorated by conventional chemoimmunotherapy [11].The mutation status of the immunoglobulin heavy chain variable region (IGHV) genes is also a powerful predictor of outcome, with mutated IGHV being associated with superior survival [12,13].The use of conventional and/ or next generation sequencing has also identified genetic mutations in the TP53, NOTCH1, BIRC3 and SF3B1 genes, which are of prognostic significance [14][15][16][17][18][19][20][21].
Currently, existing information on cytogenetic and molecular studies has been derived from Caucasian populations, and available data for CLL in Chinese patients are very limited.Comparing the cytogenetic and molecular features of CLL in Chinese and Caucasian patients may help to explain the marked ethnical disparities in disease incidence.Furthermore, it is also necessary to define the clinical significance and impact of prognostic or predictive biomarkers derived from Western patients when applied to Asian populations.To address these issues, we studied the clinical, pathological and molecular characteristics of CLL in a cohort of Chinese patients from Hong Kong and Singapore.

Demographical and clinical characteristics
Two hundred and twelve patients (men, N=154; and women, N=58) at a median age of 64 (26 -94) years were studied (Hong Kong cohort, retrospective group, N=44, prospective group, N=95; Singapore cohort, N=73).More than half of the patients presented with early-stage disease (Binet stage A, 56.1%).The median white blood cell count and lymphocyte count at the time of diagnosis were 28 x10 9 /L (6.2 -453 x10 9 /L) and 20.8 x 10 9 /L (5.2 -448.3 x10 9 /L).The demographical and baseline clinical characteristics of the patients were summarized in Table 1.

Richter's transformation
Clinical information on Richter's transformation was available in 139 patients.With a median followup period of 36 months, 5 patients developed Richter's

DISCUSSION
Our study reports on one of the largest series of Chinese CLL patients.There are several important observations.Firstly, the median age at diagnosis of our patients is 64, which is much lower than the median age of 72 reported for Caucasian CLL patients [22].Interestingly, our result is also comparable with the median age at diagnosis reported in other studies of CLL in Asian populations, which ranges from 57 to 66 [23][24][25][26].Furthermore, the lower median age at diagnosis is also observed in Chinese residents in the Western world, suggesting that there may be a genetic basis for CLL to present at a younger age in Chinese population [4].Nonetheless, the preferential referral of younger CLL patients to the participating specialist centres in this study may also partly contribute to the observed difference.
On the other hand, the prevalence and pattern of cytogenetic abnormalities in Chinese patients were similar to those of Caucasian patients.This observation was similarly reported in previous series of Chinese CLL patients [24,27] (Table 8).Isolated del(13q) (q14) was the most prevalent cytogenetic abnormality, occurring in 41.7% of our cases.This frequency is similar to that of 36% in the original study that documented a favorable prognostic impact of this aberration [10].In a recent study from the CLL research consortium [28] revisiting the hierarchal prognostic model in the era of chemoimmunotherapy, isolated 13q14 deletion occurred with a similar frequency of 39%, and was also associated with a outcome.
Mutations of NOTCH1, SF3B1, and FBXW7 in our cohort were again similar in frequencies compared with CLL patients from Western countries (Table 9).In contrast, mutations of MYD88 were more frequently identified in Chinese (10.8%) than Caucasians patients [29][30][31].Interestingly, a similarly higher frequency of MYD88 mutations in Chinese CLL patients was also observed in another study [25].MYD88 is an important protein involved in the toll-like receptor complex.Mutations in MYD88 lead to constitutive activation of NF-kB [30].It also increases the release of interleukin-6 and chemokine ligands 2, 3 and 4, resulting in recruitment of macrophages and T lymphocytes [32].All these changes create a favourable microenvironmental niche for the survival of CLL cells.MYD88 mutations have also been implicated as an early event in the pathogenesis of CLL [33], and associated with mutated IGHV and better overall survival [34].The rate of TP53 mutations was slightly higher than expected compared with Caucasian CLL, considering the lower prevalence of IGHV unmutated CLL among Chinese patients [35].The overall prognostic impact of MYD88 and TP53 mutations in Chinese CLL patients, however, requires further evaluation in future studies.
The correlations among genetic mutations, IGHV mutational status and FISH abnormalities reported in Caucasians [17,29] were also similarly seen in our cohort.In particular, NOTCH1 mutations were associated with trisomy 12.In Caucasian patients, NOTCH1 mutations were enriched from an overall frequency of 10% to that of 24% in patients having trisomy 12.The presence of NOTCH1 mutations in trisomy 12 also confers an inferior survival [19].The prognostic effect of NOTCH1 mutations was also observed in our Chinese patients.
The prevalence of mutated IGHV in the Western population was reported to be around 55-64% [13,29,36].Early reports have demonstrated a high prevalence of mutated IGHV in Chinese patients [37], and this was confirmed in a larger cohort of Chinese patients [35].In the current study, 68% of our patients had mutated IGHV, slightly higher than that in the Western populations.Further studies are required to confirm if this higher prevalence of IGHV may be associated with a better outcome in Chinese patients with CLL.# patients with more than one abnormality were counted more than once.Remark: P<0.01 comparing patient cohort from Jeromin et al and the current study.Chi square was not performed for comparison with other groups as some of the mutations were not reported.
Cytogenetic abnormalities significantly impact on TTFT.In our study, patients with del(13)(q14) had a significantly longer TTFT compared with the rest of the cohort.However, the difference among the remaining three cytogenetic groups was not apparent.Van Dyke et al [28] have shown similar findings, with the TTFT from diagnosis being 35 months, 30 months, 22 months and 22 months for patients with normal FISH, trisomy 12, del(11)(q23) and del(17p), respectively.These findings illustrate that CLL with del(17p) may initially run an indolent course, and treatment should not be commenced based only on cytogenetic findings.In contrast, CLL with del(17p) was associated with a significantly worse survival outcome, reflecting its poor response to current chemoimmunotherapy once treatment is needed.
We also showed that patients with mutated IGHV had a longer TTFT.This is in accordance with results reported in Western populations [38,39].A superior, though not statistically significant, OS was also seen in patients with mutated IGHV, indicating that the prognostic relevance of IGHV mutation was also observed in Chinese patients.
However, there were some inherent limitations of the study.Firstly, the percentage of our patients presenting with Binet stage A (56%) was lower than that reported in the Western World [40].Routine health-check including blood tests is not commonly practiced in the region especially during the early recruitment period of the study, and this may lead to the under-diagnosis of asymptomatic Binet stage A patients.In addition, the heterogeneity in treatments would affect the outcome of patients in this study.In Hong Kong and Singapore, healthcare system is mostly government subsidized, and accessibility to expensive medications is dependent on reimbursement policies.Chlorambucil was thus the mainstay of treatment, especially in early study period.As a result, a large proportion of patients with Binet stage A were given chlorambucil.Furthermore, the lack of standardized criteria for treatment initiation and the imbalance in baseline characteristics of the two treatment groups (FCR vs chlorambucil) probably had introduced bias in our analysis.In the FCR group, there was a higher proportion of patients with unmutated IGHV, as compared with the chlorambucil.This might explain the lack of statistical significant difference in OS between IGHV mutated and unmutated patient groups.
In conclusion, in a population where the incidence of CLL is only about 10% of that in Western countries, we have shown in one of the largest cohorts of Chinese patients studied at diagnosis that FISH and genetic abnormalities are similar to those previously reported in Western patients.MYD88 mutations and mutated IGHV were apparently more frequent.The prognostic significance of hierarchal FISH abnormalities, genetic mutations and IGHV mutation status in refining the prognostic stratification is applicable to Chinese patients.

Patients
CLL patients of Chinese ethnicity diagnosed and treated in participating haematology centres from Hong Kong and Singapore were analysed.Records of ethnicities of the patients were retrieved from the case notes.Patients of other ethnic groups were excluded from the study.Diagnosis of CLL was based on standard morphological and immunophenotypical criteria.Indications for treatment followed those of the iwCLL 2008 guidelines [9] and the initiation of treatment was at the discretion of the treating physicians.
There were two groups of patients from the Hong Kong cohort.The first was a retrospective group identified between January 2003 to June 2010.The second group was a prospective group collected between July 2010 to June 2013.Patients from Singapore were from a retrospective group identified from January 1999 to July 2015.All samples were taken from treatment-naïve patient with newly diagnosed CLL and were collected at the time of diagnosis.

IGHV mutation status
The analysis of IGHVDJ genes was carried out on peripheral blood leukaemic cells.Polymerase chain reaction (PCR) amplification and sequence analysis of IGHVDJ rearrangements were performed on either genomic DNA (gDNA) or cDNA using sense family specific VH primers (framework region 1 [FR1] or VH leader primers), combined with consensus JH primers as previously described [41].PCR products were sequenced directly or after a cloning procedure, using 3130 Genetic Analyzer (Life Technologies, Carlsbad, CA).Productive rearrangements were analyzed by the IMGT database (http://www.imgt.org,Montepellier, France) [42] and the IMGT/V-QUEST tool (version 3.3.0)[43].Percentage of identity to the closest germline IGHV allele was evaluated [41].

Figure 1 :
Figure 1: Correlation between time to first therapy (TTFT) and genetic abnormalities.(A) TTFT stratified according to the FISH abnormlities.(B) TTFT stratified according to the IGHV mutation status.

Figure 2 :
Figure 2: Overall survival (OS) of all patients.(A) OS of the whole cohort of CLL patients.(B) OS stratified according to the FISH abnormalities.(C) OS stratified according to the NOTCH1 mutations.(D) OS stratefied according to the IGHV mutation status.

Figure 3 :
Figure 3: Overall survival (OS) and progression free survival (PFS) of patients who received treatment.(A) OS of the treated patients.(B) PFS of the treated patients.(C) OS of the treated patients stratified according to the therapies received (FCR vs chlorambucil).(D) PFS of the treated patients stratified according to the therapies received (FCR vs chlorambucil).(E) OS of the treated patients stratified according to the FISH abnormalities.(F) PFS of the treated patients stratified according to the FISH abnormalities.

Table 2 : Results of FISH, karyotyping, IGHV analysis and targeted gene sequencing FISH abnormalities N (%) # Current study N (%) Dohner et al [10] N (%) Van Dyke et al [28]
[10]rk: hierarchal model proposed by Dohner et al was used[10].Those with more than 1 FISH abnormalities would be assigned to a group with worse prognosis.

Table 3 : Correlation between FISH abnormalities and stage
*Statistical analysis was performed for stage A vs stage B/C.(p<0.01)(Table7).Del(17p) (N=10) was associated with a significantly inferior survival (p<0.01),compared with other FISH abnormalities, with patients having a median OS of 33 months and a PFS of only 1 month (Figure3Eand 3F).