Abstract only
397
Background: Stereotactic body radiation therapy (SBRT) may benefit patients with hepatocellular carcinoma (HCC) who are not candidates for other liver-directed therapies due to poor ...liver function (plf). However, optimal patient selection, safety, and efficacy of SBRT in plf-HCC patients are not known. We studied dosimetric, baseline liver function, and radiographic features at the time of treatment that may help identify HCC patients with liver dysfunction who would benefit most from SBRT. Methods: Medical records of plf-HCC patients with cirrhosis treated with SBRT at a single institution between 2013-2016 were reviewed. Prescription doses were tailored to liver function based on INR and total bilirubin (tBili) and uninvolved liver volume. Time to local progression was evaluated using cumulative incidence analysis (Gray’s test) and competing risks regression analysis. Local progression was defined using RECIST criteria. Overall survival was estimated using the Kaplan-Meier method and Cox proportional hazards model. Results: 26 plf-HCC patients with median baseline MELD 11 (range 1-35), tBili 1.6 (0.5-6.5), INR 1.2 (1.0-9.0), tumor size of 4.1 cm (1.7-8.5 cm), and liver volume 1251 cc (596-2322 cc) were treated with SBRT. 54% received SBRT for retreatment of the same tumor. Patients were heavily pretreated with 50%, 19%, and 8% previously receiving TACE, ethanol ablation, and RFA, respectively. With a median prescription of 30 Gy (8-50 Gy) in 5 fractions (2-5), the median liver dose was 8 Gy (3-14 Gy). Median increase in MELD 90 days post-SBRT was 1.1 (0-19), with 35% experiencing no increase and 25% increasing >2.5. Local control (LC) was 55% at 6 months with one patient bridged to liver transplant. LC was not associated with improved survival (p = 0.39) and median overall survival (MS) was 8.6 months (0.9-31.7 mo). However, longer MS was seen for patients without ascites (MS=16.5 vs. 4.1 mo, p = 0.005), tBili < 2 (MS=16.5 vs. 5.2 mo, p = 0.015) and MS was not reached in patients treated to > 30 Gy (p = 0.04) or with MELD < 11 (p = 0.02). Conclusions: In this cohort of HCC patients with poor liver function, the absence of ascites, tbili < 2, and lower MELD scores were associated with improved survival following SBRT.
Abstract only
375
Background: SBRT is emerging as a treatment option for patients with unresectable liver tumors. However, generating high quality liver SBRT plans is technically challenging and ...user-dependent. Aiming for high tumor doses and low normal tissue doses can demand lengthy planning times and result in plans of inconsistent quality. This study investigates knowledge-based planning (KBP) as a standardized method to ensure efficacy, safety, and efficiency for SBRT plans and allow for broader use of this therapy. Methods: SBRT treatment plans were manually optimized for 55 liver cancer cases by an expert liver dosimetrist in a commercial treatment planning system (Varian Eclipse v13.6). Each volumetric modulated arc therapy plan was approved by a physician using strict criteria for target coverage and normal tissue sparing. The plans were used to create a custom model in a KBP system (RapidPlan, Varian Eclipse v13.6) designed to improve both efficiency and standardization of quality. To validate the model, 15 new cases were optimized manually by an expert liver dosimetrist and then semi-automatically using the KBP model. The validation plans were compared based on target coverage, normal tissue sparing, and planning time. Results: Compared to manual plans created by an expert liver dosimetrist, KBP-generated plans showed similar target coverage and improved normal tissue sparing with similar planning times. Mean and minimum target doses were similar, as was D98, p > 0.2 for all. Normal tissue complication probability for liver damage was marginally lower with KBP, mean 3 vs. 1%, p = 0.07. Doses to adjacent organs including stomach, heart, and bowel were similar. Manual planning required a median and mean time of 15 and 20 minutes, respectively, range 8-55 min. KBP required similar times of 12 and 19 min, range 8-50 min. 9 of 15 KBP cases were automated, while 6 plans required dosimetrist improvement. Conclusions: Using KBP, high quality plans for liver SBRT can be created automatically or semi-automatically. These plans are comparable to those generated by an expert liver dosimetrist. KBP could be used to standardize treatments between institutions, particularly when experience with liver SBRT is limited.
Abstract only
350
Background: To report outcomes of pooled data from patients with early stage hepatocellular carcinoma (HCC) treated with stereotactic body radiation therapy (SBRT) at two North ...American Institutions. Methods: An IRB approved collaborative review of patients with HCC treated with radical intent SBRT was conducted. Inclusion criteria included patients with Stage I-IIIA HCC (UICC/AJCC 7th Ed.) treated with SBRT (≥ 4.5 Gy/ fraction) from June 2003 until Dec 2016. Patients who were treated with SBRT were ineligible for resection, percutaneous ablative or hepatic intravascular therapies. Patients with vascular invasion and those treated with palliative intent (e.g. HCC rupture) were excluded. Overall survival, local control and toxicity of treatment were reviewed retrospectively. Results: Of 310 eligible patients, 23% were Child-Pugh (CP) class B/C (21%/2%), and 40% had failed prior liver directed therapies. The median HCC diameter was 2.4 cm (range 0.5-18.1 cm), and the median prescribed dose was 39 Gray (Gy) in 5 fractions (range: 14 - 60 Gy in 2-6 fractions). Median BED was 78.75 Gy (Range: 23.8-180.0 Gy). 8.4% of patients underwent liver transplant after SBRT. Local control at 1, 3 and 5 years was 91.5%, 82.6% and 82.6%. On multivariable analysis (MVA), the use of breath-hold motion management, but not T stage, size or dose, was significantly associated with local control (p = 0.0098). The 1, 3, and 5 year overall survival (OS) was 77.3%, 37.9% and 23.5%. Factors associated with improved OS on MVA included baseline CP A score (HR = 0.58, p < 0.0045), AFP < 10 µg/L (HR = 0.66, p = 0.0094), and transplant post SBRT (HR = 0.05, p < 0.0001). The median survival of CP A vs. B/C patients was 30.3 and 17.6 months respectively. CTCAE (v4.0) grade 3 or higher luminal gastrointestinal organ toxicity occurred in 2.5% of patients, while a decline in CP score ≥ 2 points was seen in 16.7% of patients at 3 months post SBRT. Grade 3 and above elevated liver enzymes were seen in 12.6% and 8.1% of patients at baseline and at 3 months post SBRT. Conclusions: Similar to Asian series, this North American pooled analysis found high sustained local control and excellent survival in patients with early stage HCC treated with SBRT.
Abstract only
365
Background: Radiation-induced liver disease (RILD) has been tracked when assessing the safety of fractionated liver radiotherapy. However, due to the rarity and severity of events, ...a more sensitive measure of liver damage is necessary. This study characterizes the time course of changes in generalizable measures of liver function as potential toxicity metrics. Methods: In this IRB-approved retrospective study, the records of 63 patients with 87 HCC tumors treated with SBRT at the University of Michigan between 2006 and 2012 were reviewed. Changes in Child-Pugh (CP), MELD, and MELD-Na were analyzed using the Student t test and chi-squared test. Results: 83% of the patients had cirrhosis, 62% hepatitis (hep) C, 7% hep B, 25% alcoholic, and 13% other. 83% had prior liver-directed therapy. 24% had >1 tumor concurrently treated with SBRT. Median tumor size was 2.3 cm (0.7-10), gross tumor volume was 9.2 cc (0.6-469), and mean liver dose-GTV was 4.4 Gy (0-17.6 Gy). Prior to SBRT, 73% were CP A, 25% CP B, and 2% CP C. Median baseline CP, MELD, and MELD-Na were 5, 9, and 10. Mean CP increases after 3, 6, 9, and 12 months were 0.84, 1.79, 1.72, and 1.33; increases in MELD 1.47, 2.88, 5.38, and 3.91; increases in MELD-Na 1.45, 2.03, 4.24, and 2.48. All changes were significantly increased from baseline. On univariate analysis, >1 tumor and cirrhosis predicted for a 1+ point increase in CP and >1 tumor and higher mean liver dose-GTV for a 2+ point increase in CP. Older age, >1 tumor, smaller GTV, and smaller max tumor dimension predicted for a 10+ point increase in MELD and >1 tumor, hep C cirrhosis, and higher baseline MELD-Na for MELD-Na. Patients treated concurrently to >1 tumor had greater increases in CP (3.60 vs. 0.81), MELD (9.33 vs. 1.97), and MELD-Na (8.47 vs. 2.19), p < 0.0001 for all, compared to those with 1 tumor. No differences were seen with gender, portal vein thrombosis, number of prior treatments and baseline Child-Pugh classifications. Conclusions: We describe a time course and predictive factors for change in CP, MELD, and MELD-Na scores after SBRT for HCCs. These should be investigated further for potential use in toxicity modeling after incorporating dosimetric parameters.
Abstract only
380
Background: When tumors are near organs at risk such as bowel or heart, this results in partial underdosing relative to prescribed dose. We hypothesized that equivalent uniform dose ...(EUD) may better capture the biologic effect than other dose metrics. Methods: In this IRB-approved retrospective study, 257 primary and metastatic liver tumors from 168 patients treated with SBRT at the University of Michigan Health System between 2005 and 2014 were identified. Univariate analysis and multivariate Cox regression models were used to correlate patient, tumor, and treatment characteristics with rates of LC. Results: 1 and 2 year LC were 93% and 75% for all. Of the 156 primary liver tumors, 1 and 2 year LC were 94% and 76%, compared to 83% and 61%, for patients with colorectal metastases (CRM). 22 tumors locally progressed in 17 patients of which 9 were CRM, 5 hepatocellular carcinomas (HCC), and 3 other. Tumors that locally progressed received a median prescribed dose of 50 Gy (24 - 60), GTV EUD (a=-10) 49 Gy (23 - 72), and PTV EUD 46 Gy (17 - 69). Median gross tumor volume (GTV) was 21 cc (1-103). Tumors that did not locally progress received a median prescribed dose of 50 Gy (19 - 60), GTV EUD (a=-10) 52 Gy (17 - 86), and PTV EUD 47 Gy (7 - 76). Median GTV was 10 cc (0.2 - 2092). On univariate analysis, CRM (p = 0.005) was correlated with higher progression, while high GTV and PTV EUD (a=-10) (p = 0.01, 0.05) were correlated with lower progression. HCC (p = 0.06) was borderline for low progression. Age, gender, primary or metastatic, prescribed dose, biologically equivalent dose (BED), minimum GTV or PTV dose and GTV did not predict for progression. In a multivariable model including age, gender, histology, prescribed dose, BED, GTV EUD and GTV size, only GTV EUD predicted for LC, HR = 0.947, 95% CI = 0.901 - 0.996, p = 0.036. Every 1 Gy increase in EUD decreased the risk of local failure by 5%. Conclusions: EUD was predictive of local progression even when accounting for histology, size, and dose as measured by prescribed dose and BED. GTV EUD (a=-10) over 47.64 Gy yielded a >95% local control. EUD may be the preferred dose metric in future tumor control studies.
Abstract only
424
Background: Patients (Pts) with pre-existing liver dysfunction are at high risk for further damage after SBRT. We completed a phase 2 study of individualized SBRT, utilizing pre- ...and during-treatment indocyanine green (ICG) clearance to adapt treatment and maximize both safety and efficacy. Methods: From 5/10-10/14, pts with hepatocellular carcinoma (HCC) or metastases (mets) were enrolled and underwent SBRT planning up to a target dose of 50-60 Gy or as limited by a 15% normal tissue complication probability for radiation-induced liver disease (RILD). ICG retention at 15 minutes (ICGR15) was measured prior to and 1 month after 3 of 5 planned treatments. Using a Bayesian adaptive model, RT dose was scaled down as necessary for the final 2 treatments to keep ICGR15 < 44% after the full treatment and thus minimize toxicity. Follow up was every 3 months for 2 years. Results: 90 pts received SBRT to 116 tumors and had at least 1 year of potential follow up. Median age was 62 years, range 34-85. 69 had HCC, 4 intrahepatic cholangiocarcinoma, and 17 mets. 62 had cirrhosis, most commonly HCV and alcoholic. Median Child-Pugh (CP) score was 6, range 5-9. 20 pts were CP B/C. Median pre-RT ICGR15 was 22, range 4-75, normal 4-10. Pts had a median of 1.5 (range 0-6) prior liver-directed therapies, most commonly transarterial chemoembolization (70), prior RT (36), and radiofrequency ablation (13). Median tumor size was 3 cm, and 12 had portal vein involvement. 63 received all 5 fractions (48 full dose, 15 with dose reduction due to elevated ICGR15); 27 received only 3 treatments. Median prescription dose was 47 Gy. Treatment was well tolerated with no classical RILD and a lower complication rate than expected without adaptation. 4 pts had grade 3 ascites. 2 pts had GI bleed after SBRT. 14% and 10% of pts experienced at least a 1 or 2 point increase in CP 6 months post SBRT. Local control (95%CI) at 1 and 2 yrs was 99 (96,100)% and 90 (81,100)%. 4 recurrent tumors were 3 HCC and 1 met, measuring 26, 12, 30, 38mm; treated to 30, 50, 33, 30 Gy. Conclusions: Individualized adaptive SBRT, based on ICG clearance is a promising method of allowing pts to receive the maximally aggressive dose based on each pt’s individual tolerance to RT. Funded by P01 CA59827
Abstract only
245
Background: Stereotactic body radiation therapy (SBRT) is an effective treatment for patients with hepatocellular carcinoma or liver metastases. Predictive biomarkers for liver ...toxicity would allow for careful patient selection and the adaptation of therapy. Prior studies have shown specific miRNAs are elevated prior to clinical signs of toxin induced liver injury. We hypothesized that miRNAs can be used to predict radiation induced liver toxicity prior to clinical manifestations. Methods: As part of an IRB approved clinical trial (NCT01522937) patients with liver tumors underwent a split course of SBRT. All patients underwent liver function testing using indocyanine green clearance (ICG) followed by three SBRT fractions. One month later ICG was reassessed and two more fractions were given if the patient did not have a decline in liver function. Plasma samples were obtained at baseline and 1 month after finishing three of five SBRT fractions for 89 patients. For this exploratory analysis, we selected the 10 patients that had a 2+ point increase in Child-Turcotte-Pugh score within 6 months of completing therapy and 19 additional patients without toxicity. The levels of 752 miRNAs were quantified for each sample using qPCR. We then used a univariate tree based classifier and the elastic net with 10-fold cross validation to calculate the AUC for predicting liver toxicity in individual and small groups of miRNAs, respectively. Results: Several miRNAs were found to be potentially predictive of toxicity on univariate tree based classifier analysis including the liver specific microRNA miR.122.3p, the epithelial specific miR.141.3p and miR.200b.3p, the neuroendocrine related miR.375 and miR.217, and miR.125a.5p which plays a role in hepatitis and regulates ERBB2 and ERBB3 signaling. The elastic net model achieved a conservative AUC estimate of 0.74 +/- 0.04 using 69 baseline miRNAs and 0.76 +/- 0.06 using 11 mid-treatment miRNAs to predict toxicity. Conclusions: Our preliminary analysis of the miRNAome from 29 patients receiving liver SBRT shows promising results for the ability of these markers to select patients at risk for radiation induced liver toxicity. Further validation in a larger patient cohort is needed. Clinical trial information: NCT01522937.
Abstract only
86
Background: The Common Terminology Criteria for Adverse Events (CTCAE) is frequently used to grade the severity of acute radiation dermatitis (ARD), but has not been validated ...despite decades of clinical use. We sought to develop a photonumeric scale to consistently describe ARD in breast cancer patients undergoing radiation (RT). Methods: Patients enrolled on a prospective study that included photographs and quantitative measurements of erythema and hyperpigmentation using colorimetry. 209 photographs from 35 patients with white skin and 369 photographs from 50 patients with skin of color were used to develop two photonumeric scales. Predominant erythema (in white skin) OR hyperpigmentation (in skin of color) were rated on a 4 point scale, with grading of desquamation on a separate 3 point scale. Four raters used both CTCAE and photonumeric scales to independently score all photographs. Intra- and inter-rater agreements were assessed using weighted kappa scores. Results: Using the CTCAE, 95% of photos were rated as grade 1 or 2. There was a trend toward higher grade in patients with skin of color, with grade 2 toxicity in 43% vs. 24%. Intra-rater agreement for CTCAE ratings was 65—87% (kappa 0.34—0.67), with a wide range of inter-rater agreement (56—81% agreement fraction, kappa 0.04—0.58). Using the photonumeric scale, intra-rater agreement was high for erythema/hyperpigmentation in patients with white skin (74—82%, kappa 0.49—0.70) and skin of color (69—86%, kappa 0.55—0.79), along with desquamation (78—87%, kappa 0.52—0.66). There was moderate inter-rater agreement for erythema/hyperpigmentation (51—82%, kappa 0.15—0.71) and desquamation (63—88%, kappa 0.36—0.58). Colorimetric measurements correlated strongly with photonumeric grade. Conclusions: We report a new photonumeric scale for ARD in breast cancer patients with satisfactory reliability across the spectrum of skin pigmentation. Intra-physician ratings were consistent, with moderate inter-physician agreement. The CTCAE functions as a binary scale, with 95% of ARD rated as grade 1 or 2 toxicity. Future work includes correlation with patient-reported outcomes and physician ratings at the point-of-care. Funded by a Munn Idea Grant (G011480).
Abstract only
288
Background: The impact of nCRT for patients (pts) with BR PDA has been debated, with wide variation in practice. Most studies in pts undergoing nCRT have no or historic controls, ...rather than a direct comparison of BR pts treated in the same era. We aimed to assess the effect of nCRT on operative, pathologic, and survival endpoints in pts with BR PDA. Methods: 241 pts underwent pancreatectomy (177 without and 64 with nCRT) for PDA (2001-2011). CT scans at dx were re-reviewed by a pancreatic radiologist using a standardized template. Pts were categorized as resectable, BR, or unresectable according to 2012 NCCN criteria. BR pts (n=74) were extracted for analysis. Survival was compared using log-rank tests; pathologic features and peri-op factors were compared using Wilcoxon rank sum tests (continuous variables) or Fisher's exact test (binary variables). Results: Median follow up was 26 ±21 months. nCRT was associated with higher 1 and 3 yr OS, lower margin and node positivity, decreased pre-op CA19-9, and lower post-op pancreatic fistula (POPF), with higher operative time, vascular reconstructions, and transfusion (PRBCs). On univariate analysis, POPF negatively impacted survival (HR 2.34, 1.16-4.75, p=0.02) as did pre-op CA19-9>150 U/mL (HR 2.11, 1.16-3.87, p=0.015). Conclusions: Pts receiving nCRT for BR PDA had higher 1 and 3 yr OS, more negative margins, and fewer positive nodes, with minimal additional peri-op co-morbidity, compared with pts going directly to surgery. These data support consideration of enrolling all pts with BR PDA into a neoadjuvant program prior to resection. Table: see text
Abstract only
219
Background: Interest is growing in value in health care, defined as better outcomes at lower costs. A primary driver of cost in radiation oncology is the use of IMRT. We examined ...the patterns and correlates of use of IMRT across Michigan using publicly available data. Methods: As a certificate of need state, Michigan requires every radiation oncology facility to report yearly the number of external beam and IMRT treatments delivered. Data for 2005-2008 were obtained through a Freedom of Information Act request of the Michigan Department of Community Health, while 2009-2010 data were available at its website. Percentage of external beam treatments delivered using IMRT (IMRT%) was examined across centers over time and repeated-measures longitudinal linear regression was used to identify factors associated with use. Results: During 2005-2010, 48 to 65 centers reported data. Median IMRT% (range) rose steadily during the study period: 2005 16% (0-64); 2006 21% (0-57); 2007 27% (0-79); 2008 37% (7-85); 2009 41% (0-87) 2010 45% (7-100). There was also significant between-center variation (see table). Regression modeling demonstrated that IMRT% was associated with year (+6.7% per year, p<0.0001), facility type (+7.1% freestanding versus hospital, p<0.11), facility annual volume (+5.0% high volume: 7,000+ versus low: <7,000, p=0.01) and the interaction between year and volume (low volume +2.4% per year versus high volume p<0.02). The significant interaction between year and volume suggests that the greatest IMRT% growth was in low volume centers (6.7% per year versus 4.3% per year for high volume). Conclusions: IMRT utilization has grown steadily across Michigan between 2005 and 2010. There is significant variation in its use that appears to be related in part to facility characteristics. The newly established Michigan Radiation Oncology Quality Collaborative (MROQC) is beginning to explore the use of IMRT in patients with breast and lung cancer statewide to identify those groups of patients where improved outcomes may justify its higher cost. Table: see text
