Personalizing Cancer Supportive Care: Predicting Neutropenic Complications in Patients Receiving Intermediate Risk Cancer Chemotherapy

Neutropenic complications represent important dose-limiting toxicities of cancer chemotherapy. We recently developed and validated a risk model for neutropenic complications in patients with solid tumors or lymphoma receiving cancer chemotherapy (Lyman et al. Cancer 2011). While practice guidelines recommend primary colony-stimulating factor (CSF) prophylaxis for patients at >20% risk of febrile neutropenia (FN), many patients receive chemotherapy regimens associated with an intermediate risk (10–20%) of FN. For these patients, the decision to give or withhold primary CSF prophylaxis is based on clinical judgment. We report here the ability of the risk model to identify patients with individual characteristics placing them at high risk for neutropenic complications among patients receiving intermediate risk chemotherapy regimens.

A prospective cohort study was conducted of consenting patients initiating a new chemotherapy regimen at 115 randomly selected US oncology practices between 2002–2006. The risk of cycle 1 severe or febrile neutropenia was estimated [95% CI] utilizing logistic regression analysis adjusting for key clinical factors including: planned chemotherapy, prior chemotherapy, age, abnormal hepatic or renal function, low pretreatment white blood count, and immunosuppressive medications. The cumulative incidence of severe neutropenia and FN across 4 cycles was estimated by the product limit method of Kaplan and Meier.

Among 3,760 patients with cancers of breast, lung, ovary, colon, or lymphoma, 2,270 received an intermediate risk chemotherapy regimen based on NCCN guidelines. Overall, in the subpopulation receiving intermediate risk regimens, severe or febrile neutropenia occurred in cycle 1 in 21.4% while FN over 4 cycles was observed in 11%, and primary CSF prophylaxis was utilized in 16.4%. The performance of the risk model was good in this subgroup with a c-statistic of 0.82 [0.80–0.84]. Among the half of patients classified as high risk based on the model despite receiving an intermediate risk chemotherapy regimen, cycle 1 severe or febrile neutropenia occurred in 38% [35%–41%] compared to 5% [4%–6%] of patients classified as low risk based on the model receiving such regimens. Model sensitivity and specificity were 89% and 61%, respectively. The cumulative risk of FN over 4 cycles of chemotherapy was 20% in predicted high risk group versus 5% in the low risk group (Figure). The majority of severe or febrile neutropenia events (67%) and FN events (55%) were observed in cycle 1. One-half of high risk patients who did not receive primary CSF prophylaxis in cycle 1 received CSF during subsequent cycles following a neutropenic event.

Our model for predicting neutropenic complications can identify patients at high individual risk for severe neutropenia in cycle 1 or FN in the first 4 cycles of chemotherapy when receiving intermediate risk chemotherapy. This analysis emphasizes the potential value of determining an individual patient's risk of chemotherapy complications based on a validated risk model.

View largeDownload slide

View largeDownload slide

Close modal

Lyman:Amgen: Research Funding. Crawford:Amgen: Consultancy, Honoraria, Research Funding. Dale:Amgen: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Kuderer:Amgen: Research Funding.