The OHI index predicts early mortality from organ dysfunction and survival benefit from etoposide in patients with lymphoma

Abstract

Hemophagocytic lymphohistiocytosis (HLH) is a hyperinflammatory syndrome that complicates hematologic malignancies. The Optimized HLH Inflammatory (OHI) index, based solely on the combined elevation of soluble CD25 (sCD25; >3900 U/mL) and ferritin (1000 ng/mL) levels, predicts mortality more effectively than the conventional HLH criteria. This study aimed to determine whether mortality in OHI-positive (OHI+) patients is primarily caused by lymphoma or inflammation-related causes. In a multicenter, retrospective study of patients with newly diagnosed lymphoma with sCD25 and ferritin measurements available, patients were classified as OHI+ or OHI negative (OHI-). The 1-year and 3-year overall survival (OS), event-free survival, and causes of death were analyzed, along with predicted vs observed mortality based on other lymphoma-relevant prognostic indexes. Among 135 patients with lymphoma (70 OHI- and 65 OHI+), OHI+ patients had significantly lower OS at 1 year (33% vs 81%; P < .0001) with a median survival of 190 days. OHI+ patients had a 13-fold increased mortality risk (odds ratio, 13.3; 95% confidence interval, 6.0-28.5) despite similar predicted OS based on conventional indexes (72% vs 65%; P = .46). Mortality causes differed significantly. A total of 58% of OHI+ patients died from multiorgan failure, whereas only 6% died from progressive lymphoma. In contrast, 43% of the OHI- patients died from lymphoma progression. The incorporation of etoposide into lymphoma-directed treatment was associated with improved OS in OHI+ T-cell lymphomas (P = .007). These findings underscore the clinical significance of the OHI index as a prognostic tool in lymphoma, to elucidate the mechanisms of mortality, and to identify a high-risk subgroup for which tailored treatments may lead to improved outcomes.

Graphical abstract

Figure 1.

Patients with OHI⁺ lymphoma have decreased 3-year OS and EFS. Kaplan-Meier curves comparing the 3-year OS (A) and EFS (B) between OHI⁻ (70 patients) and OHI⁺ patients (65 patients). Statistical significance was determined using the log-rank (Mantel-Cox) test (∗∗∗∗P < .0001). An event was defined as death, refractoriness, progression, or relapse of lymphoma. (C) Predicted survival of OHI⁻ and OHI⁺ patients based on the lymphoma-relevant prognostic indexes (left) and the observed survival (right). Statistics were calculated using the χ² test. ∗∗∗∗P < .0001. (D) Forest plot of the OR and 95% CI of the predicted OHI^+/− and the observed survival of patients with OHI^+/− lymphoma. The number of patients at risk at each time point is shown beneath the survival curves. ns, nonsignificant.

Figure 2.

Some disease features differ between OHI⁺ and OHI⁻ patients. Bar plots comparing OHI⁺ (sCD25 >3900 U/mL and ferritin >1000 ng/mL) and OHI⁻ patients in terms of Ann Arbor stage (A); ECOG scores (B); extranodal sites (C); the lymphoma-relevant prognostic indexes (D); extranodal sites location as defined by positron emission tomography-computed tomography scan or bone marrow biopsy (E); bone marrow involvement (F); COO (G); genetic alterations (H); and B symptoms (I). Disease features were compared using the χ² test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CNS, central nervous system; DH/TH, double hit/triple hit lymphoma; NA, not available; ns, nonsignificant.

Figure 3.

Excess mortality among OHI⁺ patients is predominantly attributed to early-onset MOF. Pie charts illustrating the proximate causes of mortality in OHI⁻ (A) and OHI⁺ (B) patients (sCD25 >3900 U/mL and ferritin >1000 ng/mL). (C) Time to death from lymphoma diagnosis stratified by the proximate causes of mortality. CNS, central nervous system; DIC, disseminated intravascular coagulation.

Figure 4.

Incorporation of etoposide into the therapeutic regimens is associated with improved survival in patients with OHI⁺ lymphoma. Kaplan-Meier curves of OHI⁺ patients (sCD25 >3900 U/mL and ferritin >1000 ng/mL) in the cohort stratified by use of HLH-directed treatments (A). (B) Bar plots of the percentage of lymphoma responses as defined by the Lugano criteria after the first line of therapy. Statistical analyses were conducted using the Fisher exact test; Kaplan-Meier curves of patients with lymphoma in the cohort stratified by inclusion of etoposide in the malignancy-directed therapeutic regimen, in OHI⁺ (C) and OHI⁻ (D) B-NHL, and OHI⁺ (E) and OHI⁻ (F) T-cell lymphoma. The number at risk is presented for each group. Statistical analysis was performed using the log-rank (Mantel-Cox) test, and the P values are presented on the plots. CR, complete response; PR, partial response; SD, stable disease; TCL, T-cell lymphoma.