Immunophenotypic analysis of adult patients with T-cell lymphoblastic lymphoma treated with hyper-CVAD

ABSTRACT

Objectives: Immunophenotype is an important prognostic factor for childhood and adult T-cell acute lymphoblastic leukemia. However, immunophenotypic data from adult patients with T-cell lymphoblastic lymphoma (T-LBL) are scarcely available.

Methods: Subjects were unselected adult patients with T-LBL who were treated with intensive chemotherapy. Immunophenotyping of tumor cells was performed according to standard techniques.

Results: A total of eight patients with a median age of 31 years were analyzed who received hyper-CVAD treatment for LBL. Immunophenotypic analysis showed that the most common tumor type was cortical T-cell type [early T (n = 2), cortical T (n = 4), and medullary T (n = 2)]. Two patients diagnosed with early T-cell type had early disease progression.

Conclusions: Assessment of T-cell differentiation stages in malignant T lymphoblasts would be important in choosing treatment strategies for adult patients with T-LBL.

KEYWORDS:

• Lymphoblastic lymphoma
• T-cell
• immunophenotyping
• hyper-CVAD regimen
• flow cytometry

Introduction

T-cell lymphoblastic lymphoma (T-LBL) is a neoplasm of lymphoblasts committed to the T-cell lineage. The cell of origin is the precursor T-cell blasts whose differentiation is arrested at discrete stages of maturation [1,2]. Cell-marker profiling has been well studied in pediatric acute lymphoblastic leukemia (ALL). Early T-cell precursor ALL (ETP-ALL) has been identified as a high-risk subgroup of T-cell ALL (T-ALL) [1,3]. However, there are less data on immunophenotypic analysis confined only to adult patients with T-LBL, a disease entity similar to ALL. Here, immunophenotypic analysis was performed to evaluate the clinical impact of T-cell differentiation stages in adult patients with T-LBL.

Patients and methods

Patients and samples

This study was approved by our institutional review board. Subjects were unselected adult patients with T-LBL who were treated with intensive chemotherapy at Aichi Cancer Center Hospital from March 2005 to October 2015. Patients were diagnosed with T-LBL based on the presence of mediastinal masses and nodal lesions, regardless of whether there were 25% or more blasts in the bone marrow. All patients received hyper-CVAD alternating with high-dose methotrexate and cytarabine, followed by maintenance and intensification [4]. Dosage adjustments were made based on each physician’s judgment. Patients with mediastinal disease at presentation were scheduled to receive consolidative irradiation. None of the patient underwent hematopoietic stem cell transplantation.

Flow cytometric analysis and immunohistochemical staining

Surface immunophenotyping of tumor cells was performed according to standard techniques that we have previously described [5]. Immunophenotypic detection was mainly performed by three- and six-color flow cytometry. Immunohistochemistry was used to supplement diagnosis. Stages of T-cell differentiation were defined by the current World Health Organization (WHO) classification (Supplementary Table 1). Lymphoid-associated antibodies to the following cell surface markers were used: CD1a, CD2, cytoplasmic CD3, surface CD3, CD4, CD5, CD7, CD8, CD34, terminal deoxynucleotidyl transferase (TdT), and CD56. Early T-cell precursor (ETP) markers were examined using immunophenotypic techniques available for routine diagnosis. The diagnosis of early T-cell precursor lymphoblastic lymphoma (ETP-LBL) was made when the T lymphoblast-phenotype was CD1a−, CD8−, and CD5dim, in conjunction with expression of at least one of the following myeloid or stem cell antigens: CD13, CD33, CD34, CD117, or HLA-DR [1,3].

Results

Characteristics of patients

A total of eight patients were analyzed. The median age was 31 years (range: 18–68 years). Six patients were diagnosed with stage IV disease. Detailed patient characteristics are listed in Table 1. All patients expressed cytoplasmic CD3, CD5, CD7, and TdT. CD1a was positive in four cases and two cases showed dual positivity for CD4 and CD8. Immunophenotypic analysis of tumor cells led to the classification of two cases as medullary T-cell phenotype (UPN 2 and 5), four as cortical T-cell phenotype (UPN 1, 3, 4, and 8), one as pre-T-cell phenotype (UPN 7), and one as pro-T-cell phenotype (UPN 6) (Table 2). The patient with the pro-T-cell phenotype was diagnosed as ETP-LBL (Figure 1).

Figure 1. Immunophenotyping of ETP. Tumor samples from a patient (UPN 6) before induction chemotherapy showed ETP phenotype of CD1a−, CD8−, and CD5dim, in conjunction with expression of stem cell antigen CD34.

Table 1. Patients’ characteristics.

UPN	Age (years)	Sex	PS	LDH	Clinical stage	B symptoms	CNS	Mediastinal tumor	PE	BM	PB
1	19	Male	1	High	IV	−	−	+	+	+	+
2	30	Female	3	High	IV	−	−	+	+	+	+
3	29	Female	0	High	IV	−	−	−	−	+	+
4	31	Male	0	Normal	IV	−	−	+	+	−	−
5	55	Female	0	Normal	IV	−	−	+	+	−	−
6	35	Male	1	High	IV	+	−	+	+	−	−
7	68	Male	0	High	I	−	−	−	−	−	−
8	18	Male	3	High	I	+	−	+	+	−	−

Abbreviations: BM, bone marrow; B symptoms (present or absent of night sweat, more than 10% weight loss over 6 months, or recurrent fever >38.3°C); CNS, central nervous system; LDH, lactate dehydrogenase; PB, peripheral blood stem cells; PS, performance status; PE, pleural effusion; UPN, unique patient number.

Table 2. Immunophenotypic analysis of tumor cells.

UPN	CD1a	CD2	cCD3	sCD3	CD4	CD5	CD7	CD8	CD33	CD34	CD56	CD57	TdT	CD19	CD20	CD79a
1	+	−	+	−	dim	+	+	+	−	−	−	−	+	−	NE	+
2	−	+	+	dim	−	+	+	−	−	−	−	−	+	−	−	NE
3	+	+	+	dim	+	+	+	−	−	−	−	+	+	−	−	+
4	+	+	+	−	+	+	+	+	−	NE	−	dim	+	NE	−	NE
5	−	+	+	+	−	+	+	−	NE	+	NE	NE	+	NE	−	−
6	−	−	+	−	−	dim	+	−	−	+	+	−	+	+	−	+
7	−	+	+	−	−	+	+	dim	−	dim	+	−	+	NE	−	−
8	+	+	+	−	−	+	+	+	−	−	+	+	+	NE	−	NE

Abbreviations: CD, cluster of differentiation; cCD3, cytoplasmic CD3; NE, not evaluated; sCD3, surface CD3; UPN, unique patient number.

Clinical characteristics of patients with progressive disease: identification of early T-cell precursor phenotype

Over a median follow-up of 40 months (range: 10–93 months), two patients experienced treatment failure, whereas the remaining six were still in remission. The two patients with treatment failure had pro-T and pre-T-cell immunophenotypes (UPN 6 and 7). The tumor cells were positive for CD56 and CD99 and negative for CD57. The patient with ETP-LBL (UPN 6) had disease progression during intensive chemotherapy. Despite the administration of salvage chemotherapy, the patient died of lymphoma 10 months after initial diagnosis. The other patient, an elderly man (UPN 7) with the pre-T-cell phenotype, had a complete radiological remission after three cycles of the regimen. However, therapy was discontinued after four cycles owing to toxicity and poor performance status. We chose irradiation as a consolidation therapy, but as radiation therapy was being planned, rapid progression was observed at the site of the original tumor. After that, he underwent salvage chemoradiotherapy [6]. Although in complete remission after salvage therapy, the patient died of a secondary myelodysplastic syndrome.

Discussion

We analyzed eight consecutive LBL patients who received hyper-CVAD treatment. Immunophenotypic analysis showed that the most common tumor type was cortical T-cell type. Expression of CD4 and CD8, or CD4/CD8 double positive cases were less frequently observed compared to those in a Children’s Oncology Group [7]. Two cases were diagnosed as medullary T-cell phenotype, four as cortical T-cell phenotype, one as pre-T-cell phenotype, and one as pro-T-cell phenotype. The patient with the pro-T-cell phenotype was diagnosed as ETP-LBL.

ETPs are considered to be a subset of early thymic immigrants from the bone marrow and have multi-lineage differentiation potential [1]. ETP-ALL/LBL accounts for 5–21% of all T-ALL/LBL in children and 13–32% in children, adolescents, and adults (Table 3). According to different populations, ETP-ALL/LBL accounts for 5% in Japan and for 11–32% in Europe or America (Table 3) [3,7–21]. Researchers at a pediatric institution analyzed T-ALL patients using microarray and identified cases with ETP features [3]. In their study, ETP-ALL had a poor prognosis compared with that of typical ALL. ETP-ALL is well characterized in pediatric cases; however, immunophenotyping in adult T-LBL has not been as well reported. Here, we analyzed the presenting features of adult patients with T-LBL. T-LBL was stratified into stages of T-cell differentiation as defined by the current WHO classification. Of the eight cases, one tumor with the pro-T-cell phenotype was diagnosed as ETP-LBL (12.5% of all cases). The patient experienced early disease progression during induction therapy. As is the case with ALL, ETP-LBL may represent a high-risk disease subtype in adult patients with LBL. Although our study included a relatively small number of patients, our findings are in line with a recent report from the MD Anderson Cancer Center analyzing clinical outcomes of adults with T-ALL/T-LBL [20].

Table 3. Treatment results according to immunophenotype data on T-ALL/LBL.

Authors and year	Diseases	Immunophenotypes	Patients		Median age	Treatment (protocol)	CR		Induction failure		OS		DFS/EFS		Reference No.
			n	%	years (range)
Children and adolescents
Reiter et al. (2000)	LBL		105		8.8 (1.1–16.4)	NHL-BFM-90									[8]
		Pro-/pre-T-cell		10									82% (EFS at 5 years)	NS
		Intermediate T-cell		40									93% (EFS at 5 years)
		Mature T-cell		6									100% (EFS at 5 years)
		T-cell, not further specified		44									91% (EFS at 5 years)
van Grotel et al. (2008)	ALL		72			DCOG ALL-7, ALL-8, or ALL-9									[9]
		Pro-/pre-T		33	6.8 (1.1–13.4)								79% (DFS at 5 years)	NS
		Cortical T		42	7.1 (1.5–15.9)								68% (DFS at 5 years)
		Mature T		25	7.3 (1.8–16.7)								65% (DFS at 5 years)
Coustan-Smith et al. (2009)	ALL		139		8.8 (0.5–18.9)	Total therapy studies XIII, XIV, and XV									[3]
		ETP		12							19% (at 10 years)	p < 0.0001	22% (EFS at 10 years)	p < 0.0001
		Non-ETP		88							84% (at 10 years)		69% (EFS at 10 years)
Patel et al. (2012)	LBL		142		Children and adolescents	COG A5971									[7]
		Early T		11									Median EFS: NR	p = 0.073
		Non-early T		89									Median EFS: NR
Inukai et al. (2012)	ALL		91		(1–18)	TCCSG L99–15									[11]
		ETP		5									40.0% (EFS at 4 years)	p = 0.014
		Typical T		95									70.9% (EFS at 4 years)
Patrick et al. (2014)	ALL		222		(1–24)	MRC UKALL2003									[12]
		ETP		16							82% (at 5 years)	p = 0.1	77% (EFS at 5 years)	p = 0.2
		Typical T		84							91% (at 5 years)		85% (EFS at 5 years)
Kobayashi et al. (2014)	ALL and LBL		77		(2–15)	JACLS NHL T-98 or ALL T-97									[10]
		Immature*		52							86% (at 10 years)	p = 0.0547	73% (EFS at 10 years)	p = 0.2525
		Mature*		48							68% (at 10 years)		60% (EFS at 10 years)
Madanat et al. (2016)	ALL		63			A modified St. Jude total XV									[13]
		ETP		21	10.8 (3.3–14.4)						77% (at 5 years)	p = 0.823	80% (EFS at 5 years)	p = 0.3812
		non-ETP		79	9.25 (0.58–17.7)						79% (at 5 years)		65% (EFS at 5 years)
Children, adolescents, and adults
Vitale et al. (2006)	ALL		90		26.6 (14.0–51.8)	GIMEMA LAL 0496									[14]
		Pre-/pro-T		51			56%	p = 0.002
		Cortical/mature		49			91%
Neumann et al. (2012)	ALL		178		(15–65)	GMALL 05/93, 06/99, 07/03 trials									[15]
		ETP		32			79%	p = 0.65			35% (at 10 years)	p = 0.60	46% (CR duration at 9 years)	p = 0.16
		Non-ETP early T		68			82%				38% (at 10 years)		57% (CR duration at 9 years)
Van Vlierberghe et al. (2013)	ALL		53		Adults	ECOG E2993									[16]
		Early immature		53							34% (at 5 years)	p = 0.0112
		Cortical/mature		47							62% (at 5 years)
Allen et al. (2013)	ALL and LBL		48		13 (0.7–81)	DFCI, COG, CALGB, NY-I, NY-II, or others									[17]
					[18 or under]
		ETP		11	9.8				25%	NS
		Non-ETP		89	8.9				21%
					[over 18]
		ETP		30	40				33%	p = 0.3
		Non-ETP		70	52				0%
Shimizu et al. (2013)	ALL		36			JALSG ALL90, ALL93, ALL97, or ALL202									[18]
		Pro-/pre-T		61	32 (16–72)		45%	p = 0.04			27% (at 3 years)	p = 0.66	19% (DFS at 3 years)	p = 0.49
		Cortical/medullary		39	33 (17–66)		79%				38% (at 3 years)		32% (DFS at 3 years)
Chopra et al. (2014)	ALL		69			INCTR, BFM-90, or hyper-CVAD									[19]
		ETP		13	22 (12–53)						77.5% (at 60 days)	p = 0.005	37.6% (EFS at 60 days)	p = 0.002
		Non-ETP		87	11 (1.5–48)						96.6% (at 60 days)		80.8% (EFS at 60 days)
Jain et al. (2016)	ALL and LBL		111		30 (13–79)	Hyper-CVAD and others									[20]
		ETP		17			73%	p = 0.03			Median OS: 1.7 years	p = 0.008	Median EFS: 1.2 years	p = 0.17
		Non-ETP		83			91%				Median OS: NR		Median EFS: NR
Brammer et al. (2017)	ALL		88		31 (2–72)	Hyper-CVAD and others									[21]
		ETP		18							29% (at 3 years)	NS
		Pro-/pre-T		32							41% (at 3 years)
		Cortical T		31							27% (at 3 years)
		Mature T		19							31% (at 3 years)
The current study	LBL		8		31 (18–68)	Hyper-CVAD
		ETP		12.5					100%	p = 0.064
		Non-ETP		87.5					14%

Abbreviations: ALL, acute lymphoblastic leukemia; BFM, Berlin-Frankfurt-Munster; CALGB, Children’s Oncology Group (COG), Cancer and Leukemia Group regimen B; CD, cluster of differentiation; CI, cumulative incidence; COG, the Children’s Oncology Group; CR, complete remission; DCOG, Dutch Childhood Oncology Group; DFCI, Dana Farber Cancer Institute; DFS, disease-free survival; ECOG, the Eastern Cooperative Oncology Group; EFS, event-free survival; ETP,early T-precursor; GIMEMA, the Gruppo Italiano Malattie Ematologiche dell’Adulto; GMALL, the German Multicenter Study Group for Adult ALL; INCTR, International Network for Cancer Treatment & Research; JALSG, the Japan Adult Leukemia Study Group; LAL, leucemia acuta limfoide; LBL, lymphoblastic lymphoma; MRC, the Medical Research Council; NHL, non-Hodgkin’s lymphoma; NR, not reached; NS, not significant; NY-I, NY-II, Memorial Sloan Kettering; OS, overall survival; TCCSG, the Tokyo Children’s Cancer Study Group.

*Immature: CD7+(CD2+ or CD5+)CD3−.

*Mature: CD7+CD2+CD5+CD3+.

Although ETP-ALL responds poorly to chemotherapy and has a very high risk of relapse in children and adults [3,11,19], there are some contrary reports [12,13,15]. There are remarkable little data published on whether or how the immunophenotype of adult LBL differs from childhood T-LBL/ALL. Table 3 shows treatment results according to immunophenotype data on T-ALL/LBL. In children and adolescents, pro/pre-T-cell and immature immunophenotypes were not associated with worse survival [8–10]; on the other hand, the phenotypes of ETP-ALL/LBL were associated with worse survival in two of the five studies (Table 3). In patients with adult ETP-ALL/LBL, about half of the study suggested that both immature and ETP phenotype could affect worse outcomes (Table 3). The prognostic value of immunophenotyping is still under investigation, especially in adult patients with T-LBL. Recent advances in molecular studies could lead to selection of optimal treatments for individuals with ETP-ALL [23]. Comprehensive studies including a focus on genetic alterations as well as immunophenotypic analysis are required for better understanding of the characteristics of ETP-LBL.

T-ALL and T-LBL are defined as the same entity, i.e. precursor lymphoid neoplasms, in the current WHO classification. Although there are several studies defining genetic and molecular differences between T-ALL and T-LBL, the two diseases are usually separated by an arbitrary cut point of 25% bone marrow infiltration. Bone marrow infiltration below 25% is considered to be T-LBL. The degree of blast infiltration has, therefore, been used as the sole criterion to distinguish between T-ALL and T-LBL in many clinical trials. In this study, two patients with ≥25% bone marrow blasts were treated as having T-LBL (UPN 1 and 3), both with cortical T-cell phenotypes. One patient (UPN 1) had a mediastinal mass and received consolidative mediastinal irradiation, whereas the other presented with cervical and axillary lymphadenopathy. Both patients have remained in complete remission. Because adult ALL is generally considered to have an unfavorable prognosis, the use of T-LBL-specific therapies may be a rational approach for some patients who are diagnosed with T-ALL.

In summary, our study yielded two suggestive findings: (1) ETP-LBL defined by immunophenotyping had a poor prognosis and (2) T-LBL-specific therapies may be a rational treatment approach for some patients diagnosed with T-ALL in clinical trials. Further studies are needed to validate the results of our analysis in a large cohort of cases.

Acknowledgements

The authors thank Yasutaka Okada, Koichi Koike, and Mako Hagino for their excellent laboratory works; and Daiki Hirano for data collection. This study was presented at the 74th Annual Meeting of the Japanese Cancer Association, 8–10 October 2015.

Disclosure statement

K.Y. has received honoraria from Kyowa Hakko Kirin. T.K. has received honoraria from Eli Lilly and Kyowa Hakko Kirin.

Additional information

Funding

This work was supported in part by the National Cancer Center Research and Development Fund (23-A-16, 23-A-17, 26-A-4), Health and Labour Sciences Research Grant for Clinical Cancer Research (19-27, 22-29) from the Ministry of Health, Labour and Welfare, a Health and Labour Sciences Research Grant (H27-Cancer Control-Ippan-005), a grand from the Japan Agency for Medical Research and Development (AMED) (17ck0106349h0001), and a Grant-in-Aid for Aichi Cancer Center Project Research. K.Y. has received research funding from Kyowa Hakko Kirin.