F. Zhong

A comparison of patients with liver cancer receiving palliative care at two Veterans Affairs (VA) Medical Centers.

75 Background: Liver cancer is a leading cause of death. Lack of data exists on palliative care in this group, and care varies by location. We aim to determine if there are differences in palliative care for patients (pts) with liver cancer by VA site. METHODS In an IRB approved protocol, we reviewed medical records of pts at 2 VA medical centers (S1, S2) with liver cancer, who were seen by Palliative Care between 2006 and 2012. Veterans were compared by 1) demographics: Karnofsky performance status > 50 (KPS), marital status, DNR/DNI, 2) palliative intervention: goals of care discussion, referrals to psychology and PT/OT, and 3) outcomes: ED visit within 1 month (mo) of death, ED visit within 6 mo of death, ICU within 30 days of death, and hospice as site of death, time from diagnosis to palliative care, and time from DNR to death. Chi-square, Fishers Exact, and T-test were performed with Stata. RESULTS here were 82 men, 30 from S1 and 52 from S2, median age 61 (35-87) and 63 (54-88) yrs, respectively. Pts were Caucasian 30 (36%), African American 47 (57%), and other 5 (0.1%). Cancer stages were locoregional 15 (50%), metastatic 13 (43%), and unknown 2 (0.1%) at S1 and 28 (54%), 20 (38%), 5 (0.1%) at S2. Differences were seen in KPS > 50: (S1, S2) [10 (33%), 15(29%) p = 0.010], married: (S1, S2) [7 (23%), 24(46%) p = 0.024], DNR/DNI: (S1, S2) [19 (63%), 13(25%), p = 0.001]. Regarding palliative intervention: goals of care discussion (S1, S2) [26 (87%), 44 (85%), p = 0.044], referral to psych (S1, S2) [16 (53%), 42 (81%) p = 0.014], referral to PT/OT (S1, S2) [17 (57%), 15 (29%), p = 0.007]. Finally in regards to outcome: ED visit within 1 mo of death (S1, S2) [7 (23%), 31 (60%), p = 0.001], ED visit within 6 mos of death (S1, S2) [18 (56%), 35 (67%) p = 0.031], ICU within 30 days of death (S1, S2) [0 (0%), 9 (17%) p = 0.019}, and hospice as site of death (S1, S2) [19 (63%), 21 (40%), p < .0001]. Other key differences included median number of days from diagnosis to initial palliative care consult (S1, S2) [97, 39, p = 0.001] and median number of days from DNR to death (S1, S2) [33, 14, p = 0.001]. CONCLUSIONS Results show there are site-specific differences between VA medical centers. These data may be important in the comparison of other centers and provide a baseline for future studies.

Palliative care interventions and end-of-life care outcomes for hepatocellular patients (pts) at two Veterans Affairs (VA) medical centers.

164 Background: Palliative care interventions and their effect on EOL outcomes for liver cancer pts have not been described. We investigated the association between palliative care intervention and EOL care outcomes. METHODS We reviewed the charts of pts with hepatocellular carcinoma and who were seen by palliative care at 2 VA medical centers from 2006 to 2012. We investigated the association between EOL outcomes (number of ER visits, ICU visits, chemotherapy, place of death, number of hospitalizations during their last 30 days), and interventions such as early referral (within 30 days of diagnosis), defined goals of care, holding a family meeting, and symptom assessment and management. The protocol was reviewed by the IRB of both VAs and analyses were done with SAS Studio v3.4. RESULTS 82 charts were reviewed (30 from EOVA and 52 from BVA). All Pts were men; 30 (39%) were white and 47(61%) African American. Interval from diagnosis to palliative care consult was 68 days (median). 72 (90%) pts were not aware of their diagnosis. 48 (65%) were not aware of prognosis. 65 (88%) had decision making capacity. 32 (39%) pts were DNR/DNI. 60 (86%) pts chose symptom management and 10 (14%) life prolongation. 47 pts (50%) had family meeting. In the last month, 35 pts (48%) had ER visit. 9 pts (12%) had ICU visits, 4 pts (5%) had chemotherapy and 48 pts (71%) had at least 1 hospitalization. 68 (85%) underwent symptom assessment. Most frequently treated symptoms were pain (78%), nausea (27%), constipation (46%), dyspnea (34%). Significant associations were found for early referral and ER admissions (OR = 3.2, p = 0.0178), family meeting and site of death (OR = 0.136 p = 0.0037), family meeting and number of hospitalizations (OR = 0.2652, p = 0.0169). Pts choosing symptom management and who had family meetings associated with hospice/home as place of death (chisq = 5.7368, p = 0.016). Knowledge of prognosis was not associated with site of death. CONCLUSIONS In this population, many palliative care interventions were performed. Early referral, family meetings were significantly associated with site of death, ER admissions, and number of hospitalizations. These results should be replicated in larger studies.

Patient-reported outcomes for determining prognostic groups in veterans with stage IV solid tumors starting systemic therapy.

48 Background: A Recursive Partitioning Analysis (RPA) prognostic algorithm based on quality of life and symptoms predicted 4 groups with distinct median survivals in patients with metastatic solid tumors receiving chemotherapy (ASCO 2013, Abst 9567). We update our findings. METHODS The RPA algorithm is based upon Karnofsky performance status (KPS), Functional Assessment of Cancer Therapy (FACT) physical well-being (PWB) subscale, and Memorial Symptom Assessment Scale Short Form (MSAS-SF) physical symptom distress (PHYS) subscale. Starting in 2007, a convenience sample of Veterans who were prescribed systemic treatment for their cancer was enrolled in an IRB approved protocol, and completed quality of life (FACT- G) and symptom (MSAS SF) questionnaires at the first cycle of treatment. We analyzed records of patients with stage IV metastatic solid tumors enrolled through June 2013, and determined survival as of June 15, 2014. Analyses were performed with STATA 11.0. RESULTS There were 97 patients(pts). The median age was 64 yrs, range 27-88. Males comprised 95 (98%) pts. First line chemotherapy was given to 78 (80%) pts. The most common primary sites were lung cancer 33 (35%), prostate 17 (17%) and colon 11 (11%) pts. Median KPS was 90% range 40-100%, PWB median 23 (range 6-28), and MSAS SF median PHYS 0.76 (range 0-3.2). Overall median survival was 285 days (range 6-2,358) and 80 pts (82%) had died. There was 1 pt in group 1, 58 in group 2, 12 in group 3, and 23 in group 4. The patient in group 1 had uterine sarcoma. Median survival (days) by RPA group was 155 for group 1, 177 for group 2, 292 for group 3, and 674 for group 4 (p=.011). CONCLUSIONS These preliminary findings suggest that this algorithm is capable of dividing patients with metastatic solid tumor who are starting systemic therapy into prognostic groups. Further development is indicated.

Comparison of outcomes of patients with hepatocellular carcinoma (HCC) over 2 consecutive decades for a VA population.

e14536 Background: Hepatocellular carcinoma (HCC) incidence is increasing in the United States. We compared the outcome of HCC patients (pts) over two consecutive decades for a VA population. METHODS In an IRB-approved protocol, we reviewed the records of pts with diagnosis (dx) of HCC seen at a VA Medical Center from 1/1/1990 to 12/31/2009. Demographics, Vietnam era (V) status, stage, alpha-fetoprotein (AFP) level at dx, hepatitis B (HBV) and C (HCV), heavy alcohol use (E), tobacco use (T), BMI, diabetes (DM), ferritin (F) level, and treatment modalities were reviewed. Cox survival regression analysis and tests of proportions were performed. RESULTS All were men. There were 42 pts were from the first decade (d1) (1990 to 1999) and 72 pts from the 2nd decade (d1) (2000 to2009). The median (M) age at dx was 62.25 yo (35-87) for the whole population (W). Comparison of two decades is shown in the table. For the second decade, more variables were analyzed: M BMI of 70 pts was 25.3 (16.9-42.4). 34 pts (47.2%) were Caucasian, 33 pts (45.8%) African American, and 5 pts (7%) others. 25pts (34.7%) has DM, 57 (79%) E use, and 61 (84.7%) T use. Two pts (2.78%) received transplant, 4 (5.56%) resection, 16 (22%) local therapy (L), and 11 pts (15.3%) systemic therapy. In univariate survival analysis, age (p<0.004), stage (p<0.0001), AFP (p<0.0017), F (p<0.002), and V (p<0.0001) were significant predictors of survival. In multivariate survival analyses that included stage, AFP, F, and V for W, stage (p<0.019), AFP (p<0.0008), and F (p<0.035) were significant predictors. CONCLUSIONS The incidence of hepatocellular carcinoma increased at this medical center. Pts were diagnosed at an earlier stage and survival improved in 2nd decade. Stage, AFP, and ferritin levels were independent predictors of survival. [Table: see text].