A new prognostic model for PTCL-NOS: developed on prospectively collected data

In the recent past, different models have been established in order to assess the prognostic outcome of peripheral T-cell lymphoma not otherwise specified (PTCL-NOS). However, these studies had an important key limitation: their retrospective nature. Thus, Massimo Federico, Founding Chair of the Lymphoma Hub Executive Steering Committee, from University of Modena and Reggio Emilia, Modena, Italy, examined a new model based on data from the prospective T-Cell Project (NCT01142674). The results were published ahead of print in the British Journal of Haematology on 19 April 2018.

In total, 311 patients (median age = 63 years, range 23–83) with PTCL-NOS were enrolled for model development. 246 patients (79%) were treated with curative intent:

• 182 patients (74%) were treated with CHOP (cyclophosphamide, doxorubicin, vincristine, prednisolone)
• 31 patients (13%) received CHOP-like regimen
• 43 patients (18%) were treated with etoposide-containing (CHOEP/CHOEP-like) therapy
• 21 patients (9%) received other regimens

The following factors were examined from previously reported studies: ECOG PS, Ann Arbor stage, B-symptoms, number of extra-nodal sites, LDH, albumin serum level, platelet count, haemoglobin level, lymphocyte/monocyte ratio, neutrophil/lymphocyte ratio, age, and sex. Study endpoints included overall survival (OS) and progression-free survival (PFS). Median follow-up was 46 months. For validation, 98 controls (median age = 61 years, range 24–90) were included in the study. The median follow-up was 18 months in the validation cohort.

Risk groups were identified based on the relative risk of death with all the potential risk factors linking the groups with a similar relative risk.

Key findings:

Survival

• Median OS was 20 months
  • 3-year OS: 41% (95% CI, 34–47)
  • 5-year OS: 32% (95% CI, 26–38)
• Median PFS was 10 months
  • 3-year PFS: 28% (95% CI, 23–34)
  • 5-year PFS: 23% [95% CI, 18–29)

Prognostic model development

• Univariate analysis showed that all variables have significant impact on OS
• Four factors were significantly predictive for OS: disease stage, ECOG-PS, serum albumin level, and absolute neutrophil count (ANC)
• Three risk groups were identified based on OS and PFS:
  • OS:
    • 3-year and 5-year OS in the ‘low risk’ group (LR, n = 48, 15%, score 0): 76% (95% CI, 61–88) and 69% (95% CI, 49–83)
    • 3-year and 5-year OS in the ‘intermediate risk’ group (IR, n = 189, 61%, score 1–2): 43% (95% CI, 35–51) and 31% (95% CI, 23–40)
    • 3-year and 5-year OS in the ‘high risk’ group (HiR, n = 74, 24%, score 3–4): 11% (95% CI, 4–21) and 8% (95% CI, 2–18)
    • P < 0.001
  • PFS:
    • 5-year PFS in the LR group: 52% (95% CI, 33–67)
    • 5-year PFS in the IR group: 22% (95% CI, 16–30)
    • 5-year PFS in the HiR group: 7% (95 CI%, 2–16)
    • P < 0.001
  • Validation patients:
    • LR, n = 18 patients (18%)
    • IR, n = 54 patients (55%)
    • HiR, n = 26 patients (27%)
    • 3-year OS in the LR group: 69% (95% CI, 46–100)
    • 3-year OS in the IR group: 41% (95% CI, 29–57)
    • 3-year OS in the HR group: 31% (95% CI, 17–57)
    • P = 0.02

The authors found that the new T-cell score model, developed from prospectively collected data, showed improved results in terms of stratifying patients and disease outcome. In comparison to other methods, the new model could distinguish better between patients based on their risk, especially for patients with unfavorable outcomes. The study group added that “further studies implementing some of the emerging biological variables to clinical factors, need to be performed to determine if clinical risk can be further refined to allow for better risk stratification.”