Palliative Care

PV QA 3 - Poster Viewing Q&A 3

TU_30_3039 - Refining the TEACHH model: Towards improved clinical utility in the modern era

Tuesday, October 23
1:00 PM - 2:30 PM
Location: Innovation Hub, Exhibit Hall 3

Refining the TEACHH model: Towards improved clinical utility in the modern era
M. S. Krishnan1, L. M. Hertan2,3, Y. H. Chen1,2, A. Nichipor2, Y. Khouj4, C. Zaslowe-Dude1, and T. A. Balboni1; 1Brigham and Women's Hospital/Dana-Farber Cancer Institute, Boston, MA, 2Harvard Medical School, Boston, MA, 3Beth Israel Deaconess Medical Center, Boston, MA, 4Dana-Farber Cancer Institute, Boston, MA

Purpose/Objective(s): The TEACHH model was previously developed to predict life expectancy (LE) in patients with metastatic cancer receiving palliative radiation and divided patients into three prognostic groups using the following risk factors: (1) Non-breast/non-prostate primary histology (2) Eastern Cooperative Oncology Group Performance Status (ECOG PS) ­> 2 (3) Age ≥ 60 (4) Receipt of ≥ 2 prior palliative chemotherapy courses (5) Hepatic metastases and (6) Prior hospitalizations within 3 months of consultation. However, the model was developed prior to the immunotherapy era and is limited in its clinical utility due to broad ranges of survival in each prognostic group. The aim of this project is to refine the TEACHH model to improve its clinical utility in a modern cohort.

Materials/Methods: We retrospectively reviewed the charts of patients seen in consultation for palliative radiation at two tertiary centers and 4 community satellite practices from 7/1/2015 – 12/31/2016. Patients were randomly selected from this cohort for refinement of the TEACHH model; the remaining patients will be used to validate the model in a future study. When patients had multiple courses of treatment, only their last course was used for analysis. Cox proportional hazard models were used to evaluate the association of TEACHH risk factors and additional established clinical predictors of LE with the probability of surviving to the clinically relevant timepoints of 2 months, 6 months and 1 year.

Results: A total of 833 patients were included in this analysis. On univariate analysis, factors significantly associated with a shorter LE were ECOG PS > 0 (p<0.0001), non-breast/non-prostate primary histology (p<0.0001), the presence of adrenal (p=0.002), spine (p=0.01), lung (p=0.03), or liver metastases (p=0.0008), greater than 2 sites of metastatic disease (p<0.0001), ≥ 2 prior palliative chemotherapy courses (p=0.0003), prior hospital (p=0.0006) or ER admissions (p<0.0001) within 3 months of consultation, inpatient consultation (p<0.0001), consultation in a community setting (p<0.0001), shorter time from primary diagnosis to development of metastases (p=0.0002) and shorter time from metastasis to radiation consultation (p=0.01). Factors not significantly associated with life expectancy included gender, race, age, and prior palliative radiation.

Conclusion: All TEACHH risk factors except for age remained significantly associated with LE on univariate analysis in this modern cohort. However, multiple additional predictors of LE were also found. We will perform multivariate analysis and then develop a refined prognostic model to predict life expectancy at the clinically relevant endpoints of 2 months, 6 months, and 1 year in patients seen for palliative radiation consultation in the modern era.

Author Disclosure: M.S. Krishnan: None. L.M. Hertan: None. Y. Chen: None. Y. Khouj: None. C. Zaslowe-Dude: None. T.A. Balboni: Employee; Dana-Farber Cancer Institute. Research Grant; Templeton Foundation. Steering Committee Member; ASCO Palliative Care Steering Committee Member.

Monica Krishnan, MD

Brigham and Women's Hospital

No relationships to disclose.


Send Email for Monica Krishnan


TU_30_3039 - Refining the TEACHH model: Towards improved clinical utility in the modern era

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Refining the TEACHH model: Towards improved clinical utility in the modern era