A. Lopezguillermo et al., LARGE-CELL LYMPHOMA - A STUDY OF PROGNOSTIC FACTORS AND ASSESSMENT OF5 RECENTLY PROPOSED PREDICTIVE SYSTEMS, Leukemia & lymphoma, 10(1-2), 1993, pp. 101-109
The main initial and evolutive variables of 133 patients with large-ce
ll lymphoma treated with adriamycin-containing regimens were evaluated
for prognostic significance. At the time of analysis, 66 patients had
died with the median survival of the series being 48.9 months. Variab
les associated with poor prognosis in the univariate study were: lymph
oma of immunoblastic subtype, advanced Ann Arbor stage, presence of B-
symptoms, poor performance status, bulky disease (greater-than-or-equa
l-to 10 cm), involvement of two or more extranodal sites, bone marrow
infiltration, and high serum LDH levels. In the multivariate analysis,
Ann Arbor stage (p < 0.001), bulky disease (p = 0.004), performance s
tatus (p = 0.018), and histologic subtype (p = 0.021) retained their p
rognostic value. After excluding those patients with localized disease
(stage 1), the Ann Arbor staging system lost prognostic significance
in favor of bone marrow infiltration (p = 0.009) and serum LDH (p < 0.
001). However, when response to treatment was included in the regressi
on model, it proved to be the most important prognostic factor (p < 0.
001), followed by serum LDH (p = 0.004). On the other hand, when the a
nalysis was restricted to complete responders, serum LDH at diagnosis
was the only parameter useful to predict survival (p = 0.008). Finally
, five recently proposed prognostic classifications were useful to sep
arate different risk-groups of patients when applied to the series.