P-177: Predictive factors for severe infections and early death during novel agent-based induction therapy in newly diagnosed, transplant-eligible myeloma: a multicohort analysis from phase III trials

Abstract

Background At initial diagnosis of multiple myeloma (MM), about 25% of patients develop severe infections and up to 10% decease during induction therapy (IT) due to immunosuppression from MM and antineoplastic therapy3. The present analysis aims at characterization and identification of predictive factors for severe infections (SI) and deaths during IT. Methods Between 07/2005 and 01/2018, 1333 patients from three subsequent multicenter phase III trials, HD4, MM5 and HD6 from the German-speaking Myeloma Multicenter Group (GMMG), received a novel agent-based IT with either bortezomib (BTZ) / doxorubicine / dexamethasone (DEX; PAD: n=192, HD4; PAd: n=296, MM5), BTZ / cyclophosphamide / DEX (VCD: n=300, MM5), BTZ / lenalidomide / DEX (VRD: n=272, HD6) or elotuzumab / VRD (ELO-VRD: n=273, HD6). SI were defined as any infection ≥ grade 3 according to the Common Terminology Criteria for Adverse Events (CTCAE). Uni- and multivariate logistic regression models were used to assess predictive factors accounting for trial effects. Results SI occurred in HD4-PAD: 27.1%, MM5-PAd: 10.8%, MM5-VCD: 9.3%, HD6-VRD: 7.3% and HD6-ELO-VRD: 9.9% of patients (overall rate of severe infections of 11.9%). Death from any cause occurred overall in 1.8% (n=24) of patients during IT with 62.5% (n=15) of deaths being infection-related. Multivariate analyses identified three major factors, besides trial effects, to predict an increasing risk for SI during IT: age >60 years (odds ratio, OR=1.81,p<0.001), International Staging System (ISS) stage III (OR=1.79,p<0.016) and low platelets (<150/nl; OR=2.07,p<0.001) at initial diagnosis. An additive score based on these three factors (one risk factor=one point) was built and included n=559, n=559 and n=203 patients with a score of 0, 1 and ≥2 points, respectively. A higher score gradually predicted an increasing risk for both, SI and early death during IT: 0 points: 7.9% / 0.9%, 1 point: 11.8% / 1.8% and ≥2 points: 22.2% / 4.4%. Conclusions Our present analysis highlights the association between SI and early deaths in newly diagnosed MM treated with novel agent-based IT. Our proposed easy-to-use scoring system allows the identification of a subgroup of patients at high risk for SI and early death during IT in clinical routine. This enables close monitoring of patients at risk and might guide preventive anti-infective strategies in clinical routine and future prospective trials. Validation of the score is being planned. At initial diagnosis of multiple myeloma (MM), about 25% of patients develop severe infections and up to 10% decease during induction therapy (IT) due to immunosuppression from MM and antineoplastic therapy3. The present analysis aims at characterization and identification of predictive factors for severe infections (SI) and deaths during IT. Between 07/2005 and 01/2018, 1333 patients from three subsequent multicenter phase III trials, HD4, MM5 and HD6 from the German-speaking Myeloma Multicenter Group (GMMG), received a novel agent-based IT with either bortezomib (BTZ) / doxorubicine / dexamethasone (DEX; PAD: n=192, HD4; PAd: n=296, MM5), BTZ / cyclophosphamide / DEX (VCD: n=300, MM5), BTZ / lenalidomide / DEX (VRD: n=272, HD6) or elotuzumab / VRD (ELO-VRD: n=273, HD6). SI were defined as any infection ≥ grade 3 according to the Common Terminology Criteria for Adverse Events (CTCAE). Uni- and multivariate logistic regression models were used to assess predictive factors accounting for trial effects. SI occurred in HD4-PAD: 27.1%, MM5-PAd: 10.8%, MM5-VCD: 9.3%, HD6-VRD: 7.3% and HD6-ELO-VRD: 9.9% of patients (overall rate of severe infections of 11.9%). Death from any cause occurred overall in 1.8% (n=24) of patients during IT with 62.5% (n=15) of deaths being infection-related. Multivariate analyses identified three major factors, besides trial effects, to predict an increasing risk for SI during IT: age >60 years (odds ratio, OR=1.81,p<0.001), International Staging System (ISS) stage III (OR=1.79,p<0.016) and low platelets (<150/nl; OR=2.07,p<0.001) at initial diagnosis. An additive score based on these three factors (one risk factor=one point) was built and included n=559, n=559 and n=203 patients with a score of 0, 1 and ≥2 points, respectively. A higher score gradually predicted an increasing risk for both, SI and early death during IT: 0 points: 7.9% / 0.9%, 1 point: 11.8% / 1.8% and ≥2 points: 22.2% / 4.4%. Our present analysis highlights the association between SI and early deaths in newly diagnosed MM treated with novel agent-based IT. Our proposed easy-to-use scoring system allows the identification of a subgroup of patients at high risk for SI and early death during IT in clinical routine. This enables close monitoring of patients at risk and might guide preventive anti-infective strategies in clinical routine and future prospective trials. Validation of the score is being planned.