Development of a novel prognostic scoring system for risk stratification of patients with angioimmunoblastic T-cell lymphoma

Angioimmunoblastic T-cell lymphoma (AITL) represents a subgroup of peripheral T-cell lymphoma arising from the T follicular helper (TFH) cell phenotype and accounting for about 2% of non-Hodgkin lymphomas. The median age at diagnosis is ~ 65 years. Peripheral T-cell lymphoma is characterized by an aggressive course and often presents with advanced-stage disease, B symptoms, elevated lactate dehydrogenase (LDH), and high-risk disease.¹

The anthracycline-containing chemotherapy regimen cyclophosphamide, doxorubicin, vincristine, and prednisone is commonly used to treat AITL, with or without the addition of etoposide, and autologous stem cell transplantation (ASCT) may improve outcomes in some patients. However, optimal treatment of AITL remains an unmet need. Therefore, Advani, et al.¹ retrospectively analyzed the prognostic impact of clinical covariates and progression of disease within 24 months (POD24) in patients with AITL and developed a novel prognostic score that has the potential to guide new therapeutic approaches.

Study design

Baseline characteristics

A subset analysis of 282 patients with AITL registered in the T-Cell Project (NCT01142674) was performed (Figure 1).

AITL, angioimmunoblastic T-cell lymphoma.
*Data from Advani, et al.¹
Eligible patients were adults (age ≥18 years) with AITL, and the median age at diagnosis was 64 years (range, 22–88 years). Of the 282 patients in the AITL subset, 74% had lymphadenopathy, 31% had splenomegaly, 22% had hepatomegaly, 30% had polyclonal hypergammaglobulinemia, and 12% reported hemolytic anemia. Additional demographic characteristics are provided in Table 1.

 Table 1. Patient characteristics*

ANC, absolute neutrophil count; β2M, β2microglobulin; CRP, C-creative protein; ECOG, Eastern Cooperative Oncology Group; IPI, International Prognostic Index; LDH, lactate dehydrogenase; PIT, prognostic index for T-cell lymphoma; PIAI, prognostic index for angioimmunoblastic T-cell lymphoma; ULN, upper limit of normal. *Adapted from Advani, et al.1
Parameters	%
Age ≥70, years (n = 282)	38
Male, (n = 282)	60
Stage III–IV, (n = 282)	90
ECOG Performance Status >2, (n = 262)	31
B symptoms, (n = 268)	64
Bulky disease >5 cm, (n = 282)	11
Extra-nodal sites ≥2, (n = 268)	33
Bone marrow involvement, (n = 268)	13
LDH > ULN, (n = 240)	58
Hemoglobin <12g/dl, (n = 261)	61
Platelets <150,000/mm3, (n = 262)	28
Monocytes <800/mm3, (n = 237)	76
ANC > 6500/mm3, (n = 251)	35
β2M > ULN, (n = 125)	79
CRP > ULN, (n = 151)	82
IPI ≥3, (n = 183)	56
PIT ≥2, (n = 183)	62
PIAI ≥2, (n = 183)	63

Study end points

• The primary end point was overall survival (OS) at 5 years.

• The key secondary end point was progression free survival (PFS) at 5 years.
• A comparison of outcomes between patients with or without POD24 after diagnosis was also made.

• Exploratory end points included comparison of outcomes of patients ≤65 years old in first complete remission (CR1) who did or did not undergo ASCT.

Results

Treatment regimens

Complete treatment data was available for 216 patients and is presented in Table 2. Most patients were treated with anthracycline-based regimens. There was no significant difference in outcomes with or without etoposide.

Table 2. Previous treatment regimens*

*Adapted from Advani, et al.1
Treatment regimens (n= 216)	%
Anthracycline + etoposide	16
Anthracycline	65
Other chemotherapy regimens	11
Supportive care	8

Efficacy

Of the patients who were treated with curative intent (n = 106), 51% achieved complete remission and 18% achieved partial response (PR), with an overall response rate (ORR) of 69%. In total, 13% of patients underwent consolidative ASCT in CR1 and 3% of patients received consolidative radiotherapy. ASCT in CR1 was associated with superior outcomes, including favorable 5-year OS and PFS. Survival outcomes are summarized in Table 3.

Table 3. Survival outcomes*

ASCT, autologous stem cell transplant; CR1, first complete remission; CI, confidence interval; OS, overall survival; PFS, progression free survival; POD, progression of disease within 24 months. *Adapted from Advani, et al.1
Time of data collection	Patients	PFS% (95% CI)	OS% (95% CI)	p value
At 2 years	With POD24 (n = 73)	48	63	<0.001
	Without POD24 (n = 193)	2	6	<0.001
At 3 years	All patients	38 (16–43)	50 (15–64)
At 5 years	All patients	32 (17–39)	44 (15–54)	-
	Chemotherapy with Etoposide	-	50 (18–68)	0.769
	Chemotherapy without Etoposide	-	43 (22–62)
	Consolidative ASCT in CR1, (n = 27)	79	89	0.05
	Transplant-eligible patients without ASCT, (n = 56)	31	52	0.022

POD24 was a powerful prognostic factor, with remarkable differences in the PFS and OS of patients with vs without POD24. Of the patients with POD24 (n = 73), 65 died within 24 months, while only 8 were alive at the last follow-up.

Deaths

In total, 128 deaths were recorded, with 48% and 84% of deaths recorded within the first and second year from diagnosis, respectively.

The most common causes of death were

• progressive disease (69%)
• infection (16%)
• second malignancies (1%)
• other treatment-related toxicities (3%)

Prognostic indices

It was observed through multivariate analysis that four variables had the highest prognostic values. This included age ≥60 years, ECOG performance status >2, elevated β2M, and elevated C-creative protein (Table 4). A novel prognostic score, the AITL score, was developed by combining these four covariates to stratify patients into low- (17%), intermediate- (23%), or high-risk (60%) groups.

Table 4. Multivariate analysis for PFS*

β2M, β2microglobulin; CI, confidence interval; CRP, C-creative protein; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; LDH, lactate dehydrogenase; ULN, upper limit of normal. *Adapted from Advani, et al.1
	Multivariate analysis
	p value	HR (95% CI)
Age ≥60 years	0.003	2.98 (1.7-6.8)
Extra-nodal sites ≥2	0.082	3.2 (1.7-5.1)
ECOG performance status >2	0.001	4.3 (2.1-6.8)
B symptoms	0.092	2.31 (1.8-3.5)
LDH > ULN	0.071	1.73 (1.4-3.1)
Hemoglobin <12g/dl	0.123	0.75 (0.2-2.4)
Monocytes <800/mm3	0.082	1.31 (0.9-3.3)
β2M > ULN	0.002	3.21 (2.2-6.5)
CRP > ULN	0.003	1.9 (1.2-4.6)

All prognostic indices identified high-risk subgroups with inferior OS and PFS compared with low-risk patients (Table 5). In comparison with the existing prognostic indices, the AITL score demonstrated the greatest discriminant power with a lower Akaike’s information criterion and higher Harrell C-statistic than the other prognostic indices.

Table 5. Comparison of AITL score with established prognostic indices*

AITL, angioimmunoblastic T-cell lymphoma; AIC, Akaike’s information criterion; IPI, International Prognostic Index; NR, not reached; OS, overall survival; PFS, progression free survival, PIT, prognostic index for T-cell lymphoma; PIAI, prognostic index for angioimmunoblastic T-cell lymphoma. *Adapted from Advani, et al.1
Prognostic index	IPI		PIT		PIAI		AITL score
Risk group (score)	Low (0–2)	High (3–5)	Low (0–1)	High (2–4)	Low (0–1)	High (2–5)	Low (0–1)	Intermediate (2)	High (3–4)
Patient, %	44	56	38	62	37	63	17	23	60
Median PFS, months	34	8	29	9	46	13	31	12	9
5-year PFS, %	40	25	37	27	44	24	41	37	13
p value	0.0006		0.03		0.006		0.003
Median OS, months	NR	16	NR	17	NR	24	NR	35	20
5-year OS, %	61	32	64	29	61	36	63	54	21
p value	0.0005		0.0002		0.001		0.0003
Harrell C-statistic	0.697		0.648		0.711		0.785
AIC global fit	566.4		575.3		543.4		524

Conclusion

This study proposed a novel prognostic score—the AITL score—that uses four clinical criteria to classify patients into low-, intermediate-, and high-risk groups with greater discriminant power than established prognostic indices. POD24 was identified as an important prognostic factor for the prediction of OS and use of B2M and CRP were suggested as independent prognostic factors for the prediction of PFS. Major limitations of the study included selection bias and lack of a randomized comparison where higher incidences of bulky disease, elevated LDH, and anemia were observed in the non-transplanted group.