Toxicity was prospectively evaluated using the MAX2 index, previously developed by Dr. Extermann and her coinvestigators. Patients with dementia or planned concomitant radiotherapy were excluded.
A total of 518 patients were included in the analysis. The median age was 76 years. In all, 23 tumor types, and 111 chemotherapy regimens were included. Almost a third of patients (32%) had grade 4 hematologic toxicities, 56% had grade 3/4 nonhematologic toxicity, and 68% had a combination. The median time to first toxicity was 22 days. Development of the model was based on two-thirds of the patient population, and validation on the remaining third.
The study followed patients through the course of chemotherapy up to 1 month after the last cycle. If chemotherapy continued beyond 6 months, assessment was ended at 6 months. The primary end point was the first occurrence of a grade 4 hematologic toxicity and/or a grade 3/4 nonhematologic toxicity.
The researchers assessed several independent variables, including age, sex, body mass index, diastolic blood pressure, comorbidities (Cumulative Illness Rating Scale for Geriatrics [CIRS-G]), CBC count, liver test results, creatinine clearance, albumin level, self-rated health, Eastern Cooperative Oncology Group (ECOG) Performance Status, Instrumental Activities of Daily Living, Geriatric Depression Scale (GDS) score, Mini-Mental State Examination (MMSE) results, Mini Nutritional Assessment (MNA) results, cancer stage, marrow invasion, pretreatment with chemotherapy, tumor response, and chemotoxicity (MAX2 index).
Dr. Extermann noted that other research has suggested that grade 4 hematologic toxicities and grade 3/4 nonhematologic toxicities have different predictors. And her group found:
• On univariate analysis for predictors of hematologic toxicities, diastolic blood pressure, albumin, lactate dehydrogenase, and IADL were significant (after adjustment for chemotoxicity).
• ECOG Performance Status, hemoglobin, creatinine clearance, albumin, MMSE, self-rated health, MNA, and the Cumulative Illness Rating Scale for Geriatrics severity index were significant predictors of nonhematologic toxicities.
The researchers then developed a scoring system for predicting hematologic and nonhematologic chemotoxicity. Scores from both subscales are totaled, and patients with a score of 0-3 are considered low risk while those with a score of 4-6 are grouped as intermediate low risk and those with 7-9 are intermediate high risk. Patients with a score greater than 9 are considered high risk.
A Young Field Looks Ahead
Dr. Martine Extermann
Dr. Extermann is planning to extend her scoring tool to include chemoradiation therapy. The score will also need to be validated for specific tumor types, and she would like to assess what happens once a patient has a severe toxicity. What predicts that they are going to have another one? Another question is whether a score can help identify a starting chemotherapy dose.
Dr. Hurria plans to validate her group’s model in a larger group of patients. It also needs to be validated for patients with specific cancers, such as breast or prostate. This could provide an evidence base to identify specific interventions to reduce or prevent severe chemotoxicities. Work also needs to be done on the pharmacology of chemotherapy drugs and how the pharmacology changes with the aging process.
No less important will be gaining insight into the decision making process as more choices become available to the elderly, she added. How can an oncologist understand what’s meaningful to the patient choosing among options? What’s the best way to communicate with this growing patient group?
Dr. Extermann reported that she has received honoraria from Amgen Inc. Dr. Hurria reported that she has significant financial relationships with Amgen, Genentech Inc., Abraxis BioScience Inc., and Pfizer Inc.