Objective:
To determine whether age is a prognostic factor of breast cancer and should be used to make treatment recommendations, because younger patients are considered to have a poorer prognosis compared with that of older patients and, thus, often receive more aggressive therapy.
Design:
A large group of patients with operable breast cancer, all of whom were followed up prospectively as part of two multicenter trials.
Setting:
Case Western Reserve University, Cleveland, Ohio, was the primary hospital and study center, with 12 participating regional institutions.
Patients:
All 1353 patients underwent uniform local-regional therapy that consisted of a modified radical mastectomy. Patients who were node negative were followed up, and patients who were node positive received systemic chemoendocrine therapy.
Main Outcome Measures:
Patients were followed up at regular intervals for either recurrence or death.
Results:
Patients ranged in age from 22 to 75 years with a median age of 56 years. Younger patients had more estrogen receptor—negative tumors (P<.0001) and a greater number of positive lymph nodes (P<.0001). Of the 241 black patients in the study, a greater percentage were younger compared with white patients (P<.0001). Age was considered in a Cox's multivariate model, together with nodes, tumor diameter, estrogen receptor content, and race. Age was not a significant predictor of either disease-free (P=.33) or overall (P=.30) survival. Using mixture models with covariates, the estimated average hazards (where λ indicates the force of mortality) of breast cancer deaths per year were similar (P, not significant) for patients 45 years old or younger (λ=0.061), older than 45 years but 65 years old or younger (λ=0.052), and older than 65 years (λ=0.061).
Conclusions:
In conclusion, younger patients as a group have more aggressive and advanced breast cancer at presentation compared with older patients. Considered in a multivariate model, together with other variables, age does not provide independent prognostic information and should not be used alone for management decisions.(Arch Surg. 1994;129:483-488)