Page 1 - CUA2018 Abstracts - Oncology-Testis

Poster session 5: Other Oncology I

UP–5.6 hazards model was used to generated hazard ratios (HR) for identifying
Factors associated with early mortality and survival outcomes factors associated with OS.
among patients with testicular neoplasms: A population–level Results: The mean age was 35.4 years (standard deviation [SD] 12.4)
analysis and 92% of patients were white. More than 95% of patients had germ
1
1,3
1
Zachary Klaassen , Thenappan Chandrasekar , Hanan Goldberg , Karan cell tumours (42% non–seminomatous germ cell tumours [NSGCT], 54%
1
1
1
2
3
Arora , Rashid Sayyid , Charles Victor , Neil Fleshner , Robert Hamilton , seminoma), 35% had radiation (rads), and 30% had chemo. Compared to
Girish Kulkarni 1,3 white patients, black men had a higher likelihood of early mortality (PR
1 Urologic Oncology, Princess Margaret Cancer Centre/University Health 1.96). Men with NSGCT, lymphoma, and variant histology were 1.59,
Network, Toronto, ON, Canada; St. George’s University School of 4.01, and 6.36 times, respectively, more likely to suffer early mortality
2
3
Medicine, St. George’s, Grenada; Institute of Health Policy, Management compared to patients with seminoma. Protective factors for early mortal-
and Evaluation, Toronto, ON, Canada ity included: marriage, later year of diagnosis, and surgical management.
Introduction: Despite remarkable survival outcomes for testicular cancer The model area under the curve was 0.875. Significant predictors for
(TCa) patients in the post–chemotherapy (chemo) era, there is a paucity OS were distant disease (HR 4.49), testicular lymphoma (HR 2.32), and
of data assessing factors associated with ‘early’ mortality after diagnosis. variant histology (HR 2.18). Black men also had inferior OS compared to
The objectives of this study were to assess predictors of early mortality white men (HR 1.55). Rads and chemo as time–varying covariates were
and factors associated with overall survival (OS). not significant in the final model.
Methods: Patients with a testicular neoplasm were identified in the Conclusions: Although patients with TCa have excellent long–term sur-
Surveillance, Epidemiology, and End Results (SEER) database from 1973– vival, there are a subset of patients that suffer mortality within two years
2011 (n=42 004). All patients had at ≥2 years of followup. The primary of diagnosis. Unmarried and black patients are more likely to suffer early
outcome was mortality within two years of diagnosis. A multivariable log mortality, raising the question of appropriate access to initial and fol-
binomial regression model was used to generate prevalence ratios (PR) for lowup care.
predictors of early mortality (≤2 years). A multivariable Cox–proportional

CUAJ • June 2018 • Volume 12(6Suppl2) S95