Assessment of Expressor Lymphoma Prevalence in High-Grade Large B-cell Lymphomas: An Observational Analysis From a Tertiary Care Centre

Abstract

Double or Triple Hit Lymphomas (DHL/THL) represent a subset of aggressive B-cell lymphomas with markedly poorer outcomes and survival rates compared to Diffuse Large B-cell Lymphomas, NOS (DLBCL, NOS). Identifying the Hit status in High-grade large B-cell lymphoma (HGBL)/ Diffuse Large B-cell Lymphomas (used synonymously henceforth) can aid in identifying cases with aggressive clinical course that may benefit from more intensive chemotherapy. Utilizing immunohistochemistry (IHC) to detect Expressor Lymphomas (EL) provides a simpler and cost-effective means to screen DLBCL cases for further molecular studies aimed at identifying potential Hit status. Our investigation aimed to assess the occurrence of Expressor Lymphoma cases in the Indian population and compare clinical (age, gender, site of tumor and B-symptoms) and pathological parameters {Ki67 proliferation index and Germinal Centre/Non-Germinal Centre (GC/Non-GC) phenotype} between Expressor and Non-Expressor Lymphomas (NEL). The present study included 131 patients diagnosed with HGBL. Men comprised 62.6% of patients, with 58% aged 60 years or younger. Clinical data on the occurrence of B-symptoms was accessible for 99 patients, with only 35 of them displaying such symptoms. Serum LDH levels were measured in 34 of the 131 total cases, and 30 of these showed elevated levels. Nodal disease was present in 26% of cases, while the remaining had extranodal involvement. The GC phenotype was observed in 58.8% of cases. Out of 131 cases, 22.2% were EL, while 77.8% were NEL. ELs exhibited a higher mean Ki67 value (p = 0.025) compared to NEL cases, indicating increased proliferation. No significant correlation was found between EL and NEL with respect to age, gender, tumor site, B-symptoms, LDH levels, or GC/non-GC phenotype. In the present study, out of a total of 131 cases, 29 were classified as EL (22.1%), while 102 cases (77.8%) were classified as NEL. It is recommended to perform IHC to detect expressor status in all cases of DLBCL, as a significant percentage of them may turn out to be ELs, indicating a potentially aggressive clinical course that warrants therapy intensification. Moreover, this IHC panel may serve as a screening tool to identify cases requiring further molecular studies for potential Hit status detection.

Introduction

Diffuse Large B-cell Lymphoma (DLBCL), a mature B-cell neoplasm, is the most common Non-Hodgkin Lymphoma (NHL) in adults, representing over 30% of cases worldwide [1, 2]. In India, the estimated NHL incidence rate was 2.2 per 10,000 population in 2012, with DLBCL being the most prevalent subtype (34–60% of cases) [3]. The classification of these aggressive B-cell lymphomas is dynamic, driven by their diverse morphological and molecular traits, genetic mutations and varying response to chemotherapy [4]. Klein et al. [5] discovered that the B-cell neoplasms harbor ‘somatic hypermutations’ in their immunoglobulin genes and concluded that the DLBCLs arise from B-cells which are either in the germinal center or post-germinal centre stage. Subsequently, other research teams validated these findings using diverse microarray platforms. They noted that the gene expression patterns of DLBCLs categorize them into Germinal Centre B-cell like DLBCL (GCB) or Activated B-cell like DLBCL (ABC) [6–8]. Hans et al. [9] tried to compare and validate IHC technique to determine cell of origin of DLBCLs against gold standard of cDNA microarray. They found that the IHC expression of CD10, BCL6 and MUM1 can be used to classify DLBCLs into GCB and non-GCB phenotypes, comparable to the classification using Gene Expression Profiling (GEP).

The revised 4th edition of WHO classification of haematolymphoid tumors (WHO-HAEM4R) introduced a category termed high-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements owing to the inferior prognosis of these neoplasms characterized by such dual or triple rearrangements [4]. Co-rearrangement of MYC with either BCL2 or BCL6, was termed as Double Hit Lymphoma (DHL) or Triple Hit Lymphoma (THL) respectively. There is no reliable criteria to predict DLBCL/HGBL-MYC/BCL2 on the basis of morphology. Since cytogenetic/molecular studies are costly and less accessible especially in resource limited settings, immunohistochemistry emerged as a practical, convenient, and accessible technique for screening purposes, offering potential insights into identifying patients who could benefit from Fluorescence-In-Situ Hybridization (FISH) testing to detect DHL/THLs. Tumors with immunohistochemical overexpression of MYC and BCL2 proteins, irrespective of the rearrangements, were termed Double Expressor Lymphoma (DEL) [4]. This DEL group also serves as a potential indicator of aggressive behavior and likely adverse prognosis.

DLBCL/HGBL-MYC/BCL2 is postulated to be a mature GC phenotype, in contrast DLBCL/HGBL-MYC/BCL6 with absence of BCL2 translocation is frequently a non-GCB phenotype, and having ABC-like gene expression profiles. The latter group has varied patterns of presentation and mutations which are linked to ABC-DLBCL, thus in the latest 5th edition of WHO blue book, this group is exluded from Hit lymphoma category and is now included in HGBL, NOS or DLBCL, NOS category and not in DLBCL/HGBL-MYC/BCL2 category, with a rider that this category is subject to re-classification in future depending on results of further studies [10].

The latest WHO has stated that using GCB-like MYC/BCL2 double expressors (on IHC) as criterion reduces the FISH testing to around 13% with a 35% positive predictive value [10]. Thus, the present study was undertaken to showcase IHC protein expression analysis enabling the differentiation between expressor and non-expressor lymphoma cases, and to compare different clinico-pathological parameters (some of which are independent prognostic indicators themselves) between these two groups. Both the current and previous editions of WHO have mentioned the lymphomas with immunohistochemical overexpression of MYC and BCL2 proteins, irrespective of the rearrangements as Double Expressor Lymphoma (DEL). Few studies in published literature have also mentioned Triple Expressor Lymphomas (TELs), expressing all three mutated proteins (MYC, BCL2, and BCL6). However, both these WHO editions do not mention TEL as a separate category. The present study has not made a distinction between DEL and TEL, and labelled ‘both’ under a common designation of ‘Expressor Lymphomas’. DHL comprises approximately 7–10% of DLBCL cases, whereas the prevalence of DEL typically ranges from 11–34% in the various studies [11–18] Expressor status in lymphoma cases has been shown to confer independent prognostic implications. The prognosis of DELs is worse than that of DLBCL, NOS. However, it demonstrates a better prognosis compared to DHLs and THLs. Most hit lymphomas cases fall within expressor spectrum, but the reverse isn’t true as the majority are activated B-cell type, and do not harbour translocations. To our knowledge, very few studies have looked into various aspects of this newly defined prognostic category (Expressor lymphomas) of HGBL, especially from Indian subcontinent; therefore, in this study we evaluated the frequency of expressor phenotype of HGBL in Indian population and compared clinical (age, gender, site of tumor and B-symptoms and pathological parameters {Ki67 proliferation index and Germinal Centre/non-Germinal Centre (GC/non-GC) phenotype} between Expressor and non-Expressor cases.

Materials and Method

This descriptive study included retrospective and prospective cases of B-NHL cases with high grade morphology {high-grade B-NHLs / Diffuse large B-cell lymphoma (HGBL/DLBCL) terms used interchangeably} over a period of 3 years. Cases of Follicular lymphoma, Small lymphocytic lymphoma/Chronic lymphocytic leukemia, Mantle cell lymphoma, Marginal zone lymphoma and Lymphoblastic lymphoma were excluded from this study. The clinical data including age, gender, B-symptoms, nodal or extranodal presentation and serum LDH levels (whether elevated or within the normal range of 125–220 IU/L) were meticulously recorded from patients’ records. Ethical clearance was obtained from the institutional ethical board.

Sample Preparation and Immunohistochemical Staining

Biopsy samples underwent thorough examination, with sections stained using hematoxylin and eosin and basic IHC to confirm the HGBL diagnosis. Upon confirmation, these slides were stained with antibodies targeting CD10, BCL6, MUM1, CMYC and BCL2 as per standard protocol. Immunohistochemistry staining was done on Ventana BenchMark Ultra, Roche Diagnostics, fully automated staining instrument.

Determination of following was done:

• Tumor proliferation index (percentage positivity on Ki67 IHC marker).

• GC/non-GC phenotype (by interpretation of CD10, MUM1 and BCL6 IHC markers as per Hans algorithm) [9].

• DEL/TEL categorization on the basis of c-MYC, BCL2 and BCL6 IHC markers.

Interpretation of Immunohistochemical Markers

IHC for different markers was taken as positive if showing definitive, moderate to strong expression in tumor cells with cut off values as follows:

• Membranous staining of ≥ 30% tumor cells for CD10

• Nuclear staining of ≥ 30% tumor cells for BCL6 & MUM1

• Nuclear staining of ≥ 40% tumor cells for CMYC

• Membranous staining of ≥ 50% tumor cells for BCL2

Ki67 proliferation index was expressed in percentage positivity.

Figure 1 depicts the algorithmic approach adopted in NHL cases showing Blastoid morphology.

Fig. 1.

Algorithm to approach Non-Hodgkin lymphoma cases showing Blastoid morphology. BL Burkitt lymphoma, NHL non-Hodgkin’s lymphoma, DLBCL diffuse large B-cell lymphomas

Histomorphological diagnosis of HGBL was followed by IHC study using a panel comprising of CD10, BCL6 and MUM1 to determine the GC/non-GC phenotype of the cases. EL/NEL status was determined on the basis of c-MYC, BCL2 and BCL6 expression. Both DEL and TEL cases were together referred as ELs and analysed as one entity in this study. Lymphoma cases not fulfilling the criteria for DEL/TEL were labeled as Non-Expressors. The current investigation assessed the occurrence of the expressor phenotype in HGBL, while also examining the clinical attributes (such as age, gender, tumor location, presence of B-symptoms) and pathological characteristics (including Ki67 proliferation index and Germinal Centre/non-Germinal Centre phenotype) in both Expressor and non-Expressor cases. The follow-up period spanned from 2 months to 3 years.

Statistical analysis was conducted using SPSS version 28.0 (SPSS, Chicago, Illinois). Continuous variables are presented as mean ± SD, and categorical variables as absolute numbers and percentages. Categorical variables were assessed using chi-square or Fisher’s exact tests. Normality of data was confirmed prior to analysis. Normally distributed continuous variables were compared using unpaired t-tests, while the Mann–Whitney U test was used for non-normally distributed variables. Kaplan–Meier curves were employed to visualize time-to-event variables, with comparison conducted using the Log rank test between Expressor and non-Expressor lymphomas. A significance level of p < 0.05 was adopted for all tests.

Results

We conducted a study involving 131 patients diagnosed with HGBL spanning a period of 3 years. The age range of the patients varied from 11 to 84 years, with the majority falling below 60 years of age (mean age being 53.2 years). Among them, 49 were female and 82 were male, resulting in a male to female ratio (M:F) of 1.7:1. Clinical data regarding the presentation of B-symptoms was available for 99 cases. Among these, only 35 patients exhibited B-symptoms. Serum LDH levels were available for 34 out of total 131 cases, out of which 30 exhibited elevated levels. The LDH levels among patients ranged from 178 to 20,429 IU/L, against a normal range of 125–220 IU/L. Biopsy specimens were received in 80 cases, while the remaining cases involved resection specimens. Some patients underwent biopsy followed by lesional resection. We selected either the biopsy or the resection specimen, depending on which sample provided well preserved sufficient material for immunohistochemical analysis. Thus our study did not include any case duplications. Among the 131 cases examined, extranodal involvement was observed in 97 cases (74%). Of these, 20 cases exhibited gastrointestinal involvement, 14 cases showed brain involvement, 9 cases each demonstrated mediastinal and bone involvement. Additionally, soft tissue and liver involvement were noted in 6 cases each, while tonsil and spleen involvement were observed in 3 cases each. The remaining 27 cases displayed involvement at various miscellaneous sites. Out of the 34 cases with nodal involvement, cervical was the most common site (14 cases), followed by axillary (7 cases) and inguinal (5 cases), with abdominal, iliac and supraclavicular site involvement constituting 2 cases each. Infrequently obturator and retroperitoneal lymph nodes involvement was identified (one patient each). The cases were assigned GC or non-GC phenotype according to the expression of CD10, BCL6 and MUM1 on IHC as per Hans algorithm. Out of total 131 cases of HGBL, GC phenotype was observed in 54 patients (41.2%), and 77 cases (58.8%) had non-GC phenotype. The cases were then evaluated according to their expression of c-MYC, BCL2 and BCL6 on IHC. Of the total 131 cases, 29 were classified as EL (22.1%) and 102 (77.8%) as NEL (Fig. 2a–f, 3a–f). Table 1 outlines the clinicopathological parameters for EL and NEL cases. Information regarding treatment history was gathered from both patients as well as their treating clinicians. However, data was unavailable for 32 cases. Among the 99 cases with documented treatment history, 40 had completed their treatment, while 46 patients had begun treatment. This latter group encompassed patients currently undergoing treatment as well as those who passed away before completing their treatment. The remaining 13 cases categorized as “treatment not received” included patients unable to undergo treatment due to severe comorbidities, advanced disease, or death shortly after diagnosis. Follow-up data was available for 102 patients. The follow-up duration varied from 2 months to 3 years. In 3 out of 102 cases, relatives provided the final outcome (whether the patient was alive or deceased) but declined to provide further clinical history or treatment details. Therefore, there is a discrepancy in the number of patients studied regarding some of the parameters, with outcome analysis available for 102 cases, while information on the presence/absence of B-symptoms and treatment history being accessible for 99 cases. The “Alive” category included patients actively undergoing treatment and those who had completed treatment by the end of the follow-up period. Till the conclusion of the study, 12 out of 21 EL patients (57.1%) and 47 out of 81 NEL patients (58%) were alive. There was no statistically significant correlation between the outcomes of the EL and NEL groups (p = 0.941). We compared the mean and median survival period in months between the EL and NEL cases. Even though the EL cases had a noticably lower mean (17.3 vs 20.23 months) and median (12 vs 19 months) duration of survival than the NEL cases, the difference was not statistically significant (p = 0.824) (Fig. 4). Molecular analysis using FISH was available in 11 out of 29 EL cases to ascertain their ‘Hit’ status. Among these cases, 2 were identified to harbor mutations in MYC, BCL2, and/or BCL6, categorizing them as ‘Hit’ lymphomas. Specifically, 1 case was classified as Double Hit lymphoma, while the other was characterized as Triple Hit lymphoma. These cases are detailed as Case 7 and Case 8 in Table 2.

Fig. 2.

Illustrations demonstrating GC phenotype in a case of Triple Expressor Lymphoma a H & E stain showing large tumor cells arranged in a diffuse pattern in an ileal mass [Extranodal NHL] (200X); b Tumor cells with positive CD10 staining (200X); c Positive BCL6 staining in tumor cells (200X); d Positive MUM1 staining in tumor cells (200X); e Tumor cells having positive CMYC staining (200X) f Positive BCL2 staining in tumor cells (200X) [Inset- High Ki67 proliferation index (200X)]

Fig. 3.

Illustrations demonstrating Non-GC phenotype in a case of Triple Expressor Lymphoma a H & E section showing diffuse sheets of large tumor cells (200X); b Negative CD10 staining of tumor cells (200X); c Tumor cells showing positive MUM1 staining (200X); d Tumor cells showing positive BCL6 staining (200X); e Positive CMYC staining in tumor cells (200X); f Positive BCL2 staining in tumor cells (200X) [Inset- High Ki67 proliferation index (200X)]

Table 1.

Clinico-pathological parameters of EL and NEL cases

	Age N (%)		Gender N (%)		Site N (%)		B-symptoms N (%)		LDH levels N (%)			Phenotype N (%)		Mean Ki67 (%)
	≤ 60 yrs	> 60 yrs	Male	Female	Nodal	Extra nodal	Present	Absent	Elevated	Not Elevated	Median Value (IU/L)	GC	Non- GC
EL (N = 29)	17 (58.6%)	12 (41.4%)	20 (69%)	9 (31%)	7 (24.1%)	22 (75.9%)	7 (33.3%)	14 (66.7%)	9 (100%)	0 (0%)	291.0	11 (37.9)	18 (62.1)	77.9
NEL (N = 102)	59 (57.8%)	43 (42.2%)	62 (60.8%)	40 (39.2%)	27 (26.5%)	75 (73.5%)	28 (35.9%)	50 (64.1%)	21 (84%)	4 (16%)	287.0	43 (42.2)	59 (57.8)	70.7
Total cases	76	55	82	49	34	97	35	64	30	4	_	54	77	131
	131		131		131		99		34			131
P value	0.94		0.422		0.8		0.827		0.201			0.846		0.025

N number, LDH lactate dehydrogenase, GC germinal centre, EL expressor lymphomas, NEL non-expressor lymphomas, DHL/THL double or triple hit lymphomas

Fig. 4.

Kaplan–Meier curve of the survival of the EL vs NEL cases

Table 2.

Clinico-pathological parameters of the 11 EL cases that underwent molecular study (FISH)

	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6	Case 7	Case 8	Case 9	Case 10	Case 11
Age	26	38	62	40	42	73	62	71	66	27	24
Gender	Female	Male	Female	Female	Female	Male	Male	Male	Female	Female	Female
B-symptoms	Absent	Present	Absent	Present	Present	Present	Absent	Present	Present	Absent	Absent
Serum LDH levels	Elevated	Elevated	–	Elevated	Not elevated	Elevated	-	Elevated	–	–	–
Treatment	Completed	Completed	Started	Completed	Completed	Started	Started	Started	Started	Started	Started
Outcome	Alive	Alive	Dead	Alive	Alive	Dead	Alive	Alive	Alive	Alive	Alive
Survival	22 months	21 months	17 months	20 months	17 months	2 months	5 months	2 months	10 months	10 months	15 days
Ki67 (%)	80	70	60	80	90	40	90	50	70	70	80
Phenotype	Non-GC	Non-GC	Non-GC	Non-GC	GC	GC	GC	GC	GC	Non-GC	Non-GC
HPE Dx	TEL	NE	NE	TEL	DEL	NE	DEL	TEL	DEL	DEL	DEL
FISH result	Neg	Neg	Neg	Neg	Neg	Neg	DHL	THL	Neg	Neg	Neg

LDH lactate dehydrogenase, GC germinal centre, DEL double expressor lymphomas, TEL triple expressor lymphomas, NEL non-expressor lymphomas, HPE Dx histopathological diagnosis, FISH fluorescence-in-situ hybridization, Neg negative

Discussion

DLBCLs often possess diverse genetic mutations, with studies indicating MYC gene translocations in around 3–16% of cases [13, 19]. MYC acts as a nuclear transcription factor, regulating genes involved in DNA repair and cell cycle. Around 29–35% of DLBCL cases involve BCL6 translocations, and 15–30% display t(14;18) translocations, leading to elevated BCL2 expression [6–8]. BCL6, akin to p53, acts as a transcriptional repressor, impeding apoptosis when overexpressed in response to DNA damage. Conversely, BCL2 acts as an anti-apoptotic agent in normal cells by inhibiting the p53 pathway, thereby promoting cell survival. It is crucial to note that immunohistochemical expression alone isn’t a reliable indicator of double-hit cytogenetic/molecular status. While many DHLs exhibit immunophenotypic expression of DEL, most DELs however lack translocations on molecular level [20]. As a possible explanation, mutated proteins may be expressed through pathways unrelated to gene mutations. Approximately 65% of identified DHL patients show MYC with BCL2 rearrangements, 14% show MYC and BCL6, and 21% show all three rearrangements (THL) [12]. Surprisingly, up to 20% of DHL cases don’t show immunohistochemical expression of MYC and BCL2, possibly indicating better outcomes [21, 22]. Most studies have used the terms ‘DLBCL’ and ‘HGBL’ interchangeably and have included all the B-NHL cases with high grade morphology for the evaluation of Expressor and/or Hit status.

Prevalence of DELs has been reported ranging from 11.6 to 34% amongst DLBCLs and 3–11% amongst DHL/THLs by various authors [14, 16, 17, 21–26]. In a systemic review and meta-analysis, the pooled proportion of DELs amongst DLBCLs was found to be 23% (with a range of 20–26%) worldwide [15]. In Indian scenario, Mehta et al. reported DELs to be 11.6% amongst DLBCLs in India [14]. Ananthamurthy’s recent study revealed double expression of C-MYC and BCL2 in 27.5% of DLBCL cases [27]. Parkhi et al., [28] in their analysis of 117 primary CNS diffuse large B-cell lymphoma cases, identified seven cases (5.9%) of double/triple-expressor lymphomas (six double-expressor and one triple-expressor). Thirunavukkarasu et al. [29] investigated a series of 109 systemic high-grade B-cell NHL cases and observed an incidence rate of 11.9% for DHL among the North Indian population.

Studies have shown that only a fraction of the Expressor lymphomas harbour genetic mutations (‘Hits’) as mechanisms like gene amplification, upregulation or epigenetic changes can result in overexpression of the transcripted proteins in the absence of actual rearrangements. Swerdlow reported that less than 20% of DELs actully proved to be DHLs on molecular testing. On the other hand, about 80–90% of ‘Hit’ lymphomas were reported to be DELs on IHC [20]. Since, DHL/THL patients are known to respond unfavourably to the standard treatment regimes, therefore it is recommended that all DLBCL cases should undergo molecular detection to determine their Hit status. Study by Hwang et al. [15] has suggested that the screening of Expressor status in all high grade lymphomas maybe be done, especially in a resource constrained set-up, for selection of appropriate cases (Double and Triple Expressors) for molecular confirmation to determine presence or absence of ‘Hit’ status (DHL/THL). The Expressor lymphomas have also been shown to have independant prognostic significance and poor outcome. The current cinical practice is to give R-CHOP therapy to all CD20 positive DLBCLs. DHL/THL and EL show poor response to this therapy with higher relapse rates [14, 16, 19]. Studies have shown that Hit as well as Expressor lymphomas benefit from more aggressive therapy than the non-Hit and non-Expressor cases respectively [13, 15]. One study has compared the effect of R-CHOP and a more intense DA-EPOCH-R therapy on DEL cases and have found better progression-free suvival in the patients receiving the intensified therapy [30].

In general, men are noted to be more affected by HGBLs. Our study reflected this trend, with males accounting for 62.6% of all patients. Within the EL and NEL groups, males comprised 69% and 60.8%, respectively, aligning with findings from other studies [8, 11, 16, 30]. However, Hashmi et al. reported a female predominance (66.7%) in ELs [31]. In our study, no correlation was observed between patients’ gender and EL/NEL subtypes. We categorized our patients into two age groups: one group consisting of patients aged less than or equal to 60 years (non-geriatric group) and the other group aged over 60 years (geriatric group). Despite the elderly population typically being more affected by all NHLs in many studies, patients over 60 years of age constituted a smaller proportion (42%) in our study. This trend is consistent with demographic data from regional cancer registries in India. The relatively younger age group observed in our study can be attributed to the overall younger population distribution in our country. Within EL cases, 58.6% belonged to the non-geriatric group, whereas the geriatric group comprised 57.8% of NEL cases. Various studies have demonstrated a significant correlation between higher age in EL compared to NEL patients. We also found a relatively older mean age of 56.97 years in the EL group compared to 52.1 years in the NEL group, although this difference was not statistically significant (p value = 0.164). While approximately two-thirds of DLBCLs typically present with nodal involvement, our study showed extranodal site involvement in nearly 2/3rd (74%) of the cases. The gastrointestinal tract (n = 20) and brain (n = 14) were the most common sites, followed by mediastinum, bone, soft tissue, liver, tonsil, and spleen. Nodal involvement was observed in only 26% of cases, with cervical lymph node involvement being the most common site, followed by axillary and inguinal lymph nodes. No association was found between site of lymphoma involvement and EL-NEL subtypes. The elevated rate of extranodal presentation observed in both EL and NEL cases aligns with the results reported by Istiadi et al., Han B et al., and Mehta et al. [14, 32, 33]. Limited studies have examined the presence or absence of B-symptoms. In our study, only 35.4% of the entire HGBL patient population exhibited B-symptoms, consistent with findings from Han et al. [33]. When comparing EL and NEL patients, Mehta et al. [14] reported a higher prevalence of B-symptoms in EL patients (60%) compared to NEL patients (49.7%). However, our study found nearly equal incidence of B-symptom presence in EL and NEL cases (33.3% and 35.9%, respectively), with no association between B-symptoms and EL/NEL status. Serum LDH levels were available for 34 out of 131 cases, including 9 EL and 25 NEL cases. Most cases, including all ELs (9 out of 9) and the majority of NELs (21 out of 25), exhibited elevated LDH levels. Only 4 NEL patients (16%) had LDH within the normal range. While ELs showed a higher tendency for elevated LDH levels (100% vs. 84%), no significant association was observed between LDH levels in ELs and NELs (p = 0.201). Many studies have reported similar trends, with Han et al. [14, 16, 30] also noting a link between higher LDH levels and poorer prognosis in EL patients, suggesting potential clinical implications. Castro D et al. [34] reported lower odds of complete response to therapy in DLBCL cases with elevated LDH levels. However, we did not examine the impact of LDH levels on patient prognosis and response to treatment.

The GC and non-GC phenotypes serve as prognostic indicators in DLBCL, where cases with the GC phenotype exhibit notably improved overall survival and failure-free survival compared to those in the non-GC group. We attempted to analyze the expression of both the Ki67 proliferation index and the GC/Non-GC phenotype, comparing them between EL and NEL groups. Our study revealed that over half (58.8%) of all cases exhibited a non-GC phenotype, with a distribution between the EL and NEL groups being 62.1% and 57.8% respectively; however no statistical correlation being observed in phenotype distribution between the EL and NEL groups (p value = 0.846). Several studies have also noted a predominance of non-GC phenotype among ELs (ranging from 63 to 89.5%) [14, 16, 21, 35]. Conversely, a separate study indicated that ELs commonly exhibit a GC phenotype (85%) [24]. The discrepancies in the distribution of GC and non-GC cases within both EL and NEL groups across various studies may be attributed to differences in the overall phenotype distribution within the respective study populations. In our study, the Ki67 proliferation index ranged from 20 to 90%, with a mean of 70.57%. The mean Ki67 proliferation index was higher in the EL group (77.9%) compared to the NEL group (70.7%), which was found to be statistically significant with a p-value of 0.025. Few other studies have also recorded a higher mean Ki67 proliferation index in ELs compared to NELs (81.7% and 78.1%, respectively) [14, 31].

ELs are known for their poor prognosis and limited response to therapy. In our study, follow-up data was available for 102 patients, including 21 EL cases and 81 NEL cases, with follow-up durations ranging from 2 to 36 months. Among the 21 EL cases, 12 (57.1%) were alive at the last follow-up, comparable to the 47 out of 81 NEL cases (58% survival rate among NEL patients). The mean survival duration for ELs was 17.3 months, which was lower than the 20.3 months observed for NELs. However, no statistically significant correlation was found (p value = 0.824), possibly due to the relatively small sample size and shorter follow-up duration. Further research with larger cohorts and longer follow-up periods is needed for more accurate estimation. Mehta et al. [14] found 4-year overall survival rates of 46.9% for ELs versus 71.9% for NELs. Green et al. [16] observed lower complete response rates (70% vs. 89%) and 3-year overall survival (54% vs. 91%) in ELs compared to NELs. Istiadi et al. [32] reported 3-year OS rates of 36% for ELs and 80% for NELs. Hatzl et al. [35] noted a median OS of 1 year for ELs versus 2.3 years for NELs, with ELs showing a 13 times higher risk of disease progression. Herrera et al. [22] found a higher relapse risk in ELs (50% vs. 38%). Savage et al. [17] noted a higher 2-year CNS relapse risk in DELs (9.7% vs. 2.2%). DEL status indicated worse progression-free and overall survival in relapsed DLBCLs compared to NELs [22, 36]. Some of these can be combined. Numerous studies have compared EL cases based on the therapy administered, including CHOP or R-CHOP, along with other aggressive treatments. Across these studies, EL cases consistently exhibited a significantly poorer prognosis, treatment resistance, higher relapse rates, reduced progression-free survival, and increased mortality risk when subjected to standard therapy [16, 18, 30, 37]. Studies have also reported better response rate of the DELs to novel therapeutic agents or more aggressive therapies [30, 38–41]. In our study, all cases received the standard therapy for DLBCL, namely R-CHOP, irrespective of their Expressor status. Therefore, no comparative analysis of therapy could be made in the present study.

Out of the 29 patients with Expressor lymphoma, molecular analysis using FISH was conducted on 11 cases to detect mutations in MYC, BCL2, and/or BCL6. One case was identified as DHL, and another as THL. Both cases were recently diagnosed and are currently alive till the conclusion of the study. To the best of our knowledge, our study represents the fourth investigation into ELs within the Indian population. Our study, conducted as a pilot, involved a relatively small sample size of 131 cases, which may limit the precision of our conclusions. Larger-scale studies are needed to confirm any significant findings. Additionally, there was a lack of sufficient data available concerning molecular studies in ELs. A more comprehensive molecular analysis could have yielded valuable data for correlation analysis.

Conclusion

All HGBL/DLBCL cases can be evaluated using IHC (DEL/TELs) to facilitate molecular testing for ‘Hit’ status (DHL/THL), which necessitates more aggressive treatment compared to other HGBL patients. EL cases also demonstrate poor prognosis and are more aggressive than DLBCL, though less so than the Hit lymphoma subtype. Our study identified a prevalence of 22.2% for Expressor lymphoma among all HGBLs. A significant statistical correlation was observed between ELs and mean Ki67 values compared to NEL cases, with ELs exhibiting a higher proliferation index. However, no significant correlations were found between EL and NEL concerning age, gender, tumor location, presence or absence of B-symptoms, serum LDH levels, or GC/non-GC phenotype. Consequently, our findings align with the majority of previous studies indicating that ELs tend to have a poorer prognosis compared to NELs. Further studies are needed to better evaluate the clinico-pathological profiles of these EL patients.

Author Contributions

Conceptualization: SR, SB, NG. Data curation: SnR, SR. Project administration: SR, SB, NG. Writing-original draft: MAO, SnR. Writing-review & editing: MAO, SnR, SR. Approval of final manuscript: SR, SB, NG, MAO, SnR.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declarations

Conflict of interest

The authors declare that there is no conflict of interest.

Ethical Approval

All procedures performed in the current study were approved by the Ethics Committee of Sir GangaRam Hospital, New Delhi, India (EC/12/20/1767). Informed consent was waived by the Institutional Review Board.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.