A Cohort Study to Evaluate Genetic Predictors for Aromatase Inhibitor Musculoskeletal Symptoms (AIMSS): Results from ECOG-ACRIN E1Z11

PURPOSE

Aromatase Inhibitor-Associated Musculoskeletal Symptoms (AIMSS) are common and frequently lead to AI discontinuation. Single nucleotide polymorphisms (SNPs) in candidate genes have been associated with AIMSS and AI discontinuation. E1Z11 is a prospective cohort study designed to validate associations between 10 SNPs and AI discontinuation due to AIMSS.

PATIENTS AND METHODS

Postmenopausal women with stage I-III hormone receptor-positive breast cancer received anastrozole 1 mg daily and completed patient-reported outcomes (PRO) to assess AIMSS (Stanford Health Assessment Questionnaire; HAQ) at baseline, 3, 6, 9, and 12 months. We estimated that 40% of participants would develop AIMSS, and 25% would discontinue AI treatment within 12 months. Enrollment of 1,000 women with a fixed number per racial strata provided 80% power to detect an effect size of 1.5-4. SNPs were in ESR1 (rs2234693, rs2347868, rs9340835), CYP19A1 (rs1062033, rs4646), TCL1A (rs11849538, rs2369049, rs7158782, rs7159713), and HTR2A (rs2296972).

RESULTS

Of 970 evaluable women, 43% developed AIMSS and 12% discontinued AI therapy within 12 months. While more Black and Asian women developed AIMSS compared to White women (49% vs 39%, p=0.017; 50% vs 39%, p=0.004, respectively), AI discontinuation rates were similar across groups. None of the SNPs were significantly associated with AIMSS or AI discontinuation in the overall population, or in distinct cohorts. The odds ratio for rs2296972 (HTR2A) approached significance for developing AIMSS.

CONCLUSION

We were unable to prospectively validate candidate SNPs previously associated with AI discontinuation due to AIMSS. Future analyses will explore additional genetic markers, PRO predictors of AIMSS, and differences by race.

Socioeconomic deprivation and patient-reported outcomes in symptom management trials for patients with breast cancer

158

Background: Patient-reported outcomes (PROs) provide important metrics to guide treatment decision making and have been shown to be predictive of clinical outcomes. Neighborhood socioeconomic deprivation, characterized by lower levels of income, education, and housing quality, has been shown to be associated with worse outcomes in clinical trials. However, the relationship between deprivation and PROs in clinical trials has not been systematically examined. Methods: We examined the relationship between multiple PRO domains and deprivation using data from 3 randomized trials (S0927, S1200, S1202) of interventions for aromatase inhibitor (AI)-associated musculoskeletal symptoms (AIMSS) among female breast cancer patients. The studies were conducted by the SWOG Cancer Research Network. Deprivation was measured using patients’ residential zip code linked to the area deprivation index (ADI), measured on a 0-100 scale and categorized into tertiles based on national ADI distribution. Multivariable linear regression was used with adjustment for patient age, race, ethnicity, cancer stage, days on AI, and body mass index (BMI). Secondary models adjusted for rurality and insurance (Medicaid/no insurance v. private/Medicare). Average joint pain and pain interference (each on 0-10 scales) based on the Brief Pain Inventory, and physical and functional wellbeing (PWB, FWB) based on the FACT-ES (on 0-28 scales), at baseline were analyzed. Results: Overall, 761 patients were examined. Median age was 60 years, 8% of patients were Black, 5.5% were Hispanic, 87% had Stage I or II disease, and median duration on AI was 365 days. 51% of patients were from least deprived areas (bottom tertile of ADI), while 15% were from the most deprived areas (top tertile). Compared with patients from the least deprived areas, patients in the most deprived areas had worse FWB (β = -1.53, 95% CI: -2.7, -0.4; p =.01) and average pain scores (β = 0.51, 95% CI: 02, 0.8; p = 0.002) at baseline. Patients from more deprived areas (middle tertile of ADI) had worse FWB (β = -1.3, 95% CI: -2.1, -0.4; p =.005) and pain interference (β = 0.5, 95% CI: 0.2, 0.9; p =.002) compared to those from least deprived areas. For pain outcomes, the ADI coefficient was attenuated but statistically significant with adjustment for insurance. The association of FWB with ADI was not statistically significant after adjusting for insurance. No statistically significant differences were noted for PWB. Conclusions: Patients with AIMSS who live in neighborhoods with higher social needs had slightly worse FWB joint pain, and pain interference at baseline, but similar PWB. The effect of insurance on the PRO-ADI association indicates that individual access to healthcare explains some of the area-level differences in PROs. Future work will examine differences in PRO trajectory by ADI, adjusted for insurance, over the course of patient participation in these trials.

Association between oncology clinical pathway utilization and quality and cost outcomes in patients with metastatic solid tumors

430

Background: Anticancer drug regimens that are approved by accepted drug compendia and also considered high value based on their efficacy, toxicity, and costs are designated as “on-pathway” for a national commercial payer. This study compared quality and cost of cancer care among patients with metastatic solid tumors treated in the first line setting who were prescribed on- vs. off-pathway regimens. Methods: Using administrative claims data and prior authorization data from a national commercial payer, we identified 8,357 commercially insured or Medicare Advantage adult patients with solid tumor cancers including breast, lung, colorectal, pancreatic, melanoma, kidney, bladder, gastric, or uterine cancer, who were prescribed first-line anti-cancer regimens in the metastatic setting from 2018 to 2021. Patients were classified into on- vs. off pathway group based on the initial anticancer regimen that was prescribed. On-pathway status was prospectively defined by a panel of practicing oncologists based on review of curated evidence and general application of relative clinical value frameworks accepted in the field. We compared post–6-month quality-of-care outcomes including chemotherapy-related avoidable hospitalizations, emergency room (ER) visits, immune-related adverse events (IRAEs) such as endocrinopathies owing to immune checkpoint inhibitors, need for supportive drugs such as granulocyte colony stimulating factor, and cost outcomes between groups. Generalized linear models were used to assess the association between on-pathway regimens and outcomes adjusting for key patient demographics, clinical and provider characteristics. Results: A total of 5,453 (65.3%) patients were prescribed on-pathway regimens. Both groups had similar age (60.1 vs. 59.6, p = 0.06) and ECOG performance status (0.63 vs. 0.62, p = 0.40), with more females in the off-pathway group (54.6% vs. 57.3%, p = 0.02). There was no statistically significant difference in chemotherapy-related avoidable hospitalizations, IRAEs and need for supportive drugs between the two groups after modeling adjustment. However, patients treated on-pathway had higher rates of chemotherapy-related avoidable ER visits (18% vs. 15%, adjusted odds ratio: 1.16, 95% confidence interval (CI): 1.01 to 1.33, p = 0.03). Patients in the on-pathway group had significantly lower 6-month anticancer treatment cost (adjusted cost difference: -$10410, 95% CI: -$14935 to -$5886, p < 0.01), resulting in an overall lower total healthcare costs (adjusted cost difference: -$12826, 95% CI: -$18879 to -$6773, p < 0.01). Conclusions: Pathway regimens for metastatic solid tumors were associated with reduced total healthcare costs and similar quality of care compared with off-pathway regimens. These findings support the use of high value, evidence-based regimens for metastatic cancer patients.

Treatment patterns, clinical outcomes, and costs following the approval of new systemic therapy medications for hepatocellular carcinoma

398

Background: Between March 2017 and May 2020, ten new systemic medication therapies for hepatocellular carcinoma (HCC) were approved for the US market. However, clinical trials for medications only showed marginal improvement in health outcomes compared to placebo or sorafenib, the only medication previously approved for HCC. The objective of this study was to examine the change in treatment patterns and outcomes of incident patients with HCC before and after multiple medication approvals. Methods: This observational cohort study used administrative medical and pharmacy claims data for adult commercially insured and Medicare Advantage patients with at least two diagnoses of HCC during July 2013 to December 2020. Patients were categorized into two groups: the pre-approval group (identified between July 2013 and March 2017) and the post-approval group (identified between April 2017 and December 2020). We compared systemic therapy treatment (e.g., sorafenib and lenvatinib) patterns, survival, hospitalization for serious infections, other treatments such as G-CSF use and liver transplants, and cost (e.g, total all-cause costs and anti-HCC systemic therapy). Generalized linear model and Cox proportional hazards regression were used to assess the effect of new approvals on outcomes while controlling for baseline characteristics. Results: We identified 2,730 patients in the pre-approval group and 2,290 patients in the post-approval group. After risk adjustment, the latter was more likely to use systemic therapy medications (20.4% vs. 13.7%, adjusted odds ratio [aOR] 1.41, 95% confidence interval [CI] 1.22-1.62). Overall survival remained similar between the two groups (adjusted hazard ratio [HR] 0.96, 95% CI 0.87-1.06) and among the subsets who used systemic therapy (aHR 1.10, 95% CI 0.91-1.32). Odds of liver transplant (aOR 1.03, 95% CI 0.80-1.34) and hospitalization for serious infections (aOR 0.95, 95% CI 0.87-1.09) were similar, while the post-approval group had lower odds of G-CSF use (aOR 0.59, 95% CI 0.48-0.73). No differences in total per-member-per-month (PMPM) costs were seen, but the post-approval group had lower PMPM pharmacy-related costs (adj cost diff -$1,188, 95% CI -$1,575 to -$755) and higher PMPM systemic therapy costs (adj cost diff $971, 95% CI $584 to $1446). Conclusions: As expected, use of newer systemic therapies in HCC patients increased following approval of new medications. However, there was no associated change in clinical outcomes such as survival or serious infections, providing additional real-world evidence of marginal improvement, if any. Pharmacy costs decreased in the post-approval group, which may reflect a shift from oral sorafenib to newer injectable systemic therapies.

Predictors of compliance with payer-led oncology clinical pathways

7

Background: Rising healthcare costs have garnered interest from payers in shifting oncology care towards a value-based practice model. Pathways are a subset of evidence-based guidelines designed to standardize cancer drug prescribing by clarifying decisions along three priorities: efficacy, safety, and cost. However, the major predictors of compliance with clinical pathway recommendations are unknown. Methods: We conducted a retrospective cohort study using administrative claims linked with prior authorization data of Anthem commercial and Medicare Advantage members. We identified members aged 18 or older with a diagnosis of breast, lung, colorectal, bladder, kidney, uterine, pancreatic cancer or melanoma being treated with a first-line treatment regimen in the metastatic setting between July 2018 and October 2021. The primary outcome was pathway compliance (PC), defined as whether a patient’s anti-cancer drug regimen is also an Anthem pathway-endorsed regimen. We built a logistic regression model with stepwise backward selection to identify patient, physician, practice, geographic, and temporal factors that were associated with PC. Results: The cohort comprised 17,584 patients in total. The treatment period yielded 11,277 (64%) observations of patients with PC (48.1% male, mean [SD] age 60.7 years [11]) and 6,307 (36%) without PC (45.1% male, mean [SD] age 60.4 years [11.6]). In adjusted analyses, we find that the odds of PC decrease over time and with variation by malignancy. Any inpatient (OR = 1.32, 95% CI 1.22 – 1.43) or emergency room utilization (OR = 1.21, 95% CI 1.12 – 1.31) in the baseline 6-month period was associated with a higher odds of PC. Conversely, doubling total medical costs in the preceding 6 months was associated with lower odds of PC (OR 0.86, 95% CI 0.83 – 0.88). The odds of PC increased with a higher share (median or above vs below median) of Anthem patients per physician (OR 1.12, 95% CI 1.04 – 1.20). Finally, Oncology Care Model participation was associated with higher odds of PC (OR = 1.13 95% CI 1.04 – 1.23). Conclusions: Among patients treated in the first-line setting for metastatic disease, PC was observed in slightly under two-thirds of cases and was associated with the penetration of Anthem patients in the practice. Participation in OCM was also associated with increased PC. These findings would be of interest to policymakers focused on value-based cancer care.

Predictive value of baseline patient-rated treatment bother for early anastrozole discontinuation in a racially diverse cohort: Results from ECOG-ACRIN E1Z11

12094

Background: The Functional Assessment of Cancer Therapy patient-reported outcomes (PRO) item GP5 (“I am bothered by side effects of treatment”) estimates treatment tolerability. We aimed to extend our previous finding that GP5 predicts early aromatase inhibitor (AI) discontinuation (E1Z03, 96% White) in the racially diverse E1Z11 trial cohort. Methods: E1Z11 was coordinated by the ECOG-ACRIN NCI Community Oncology Research Program (NCORP) Research Base. Postmenopausal women initiating anastrozole per clinical care for ER+ stage I-III breast cancer with a pain score 0-3/10 and no rheumatologic comorbidities were eligible. Accrual of a racially diverse cohort of 1,000 women, including Black and Asian women, was planned. GP5 was administered prior to initiating anastrozole (trial baseline) and at 3, 6, 9 and 12 months. GP5 was scored on a 5-point Likert scale from 0 (not at all) to 4 (very much) and dichotomized as no/little treatment bother (0/1) or moderate/high treatment bother (2-4), consistent with previous analyses. A univariate Cox proportional hazards model estimated baseline GP5’s association with treatment duration via hazard ratio (HR). Early treatment discontinuation status was defined as treatment duration < 12 months with discontinuation not attributed to disease progression or death (n = 4), consistent with previous analyses. Results: 1,046 women enrolled from 6/2013-10/2018 (640 White, 201 Black, 205 Asian), including 590 (56%) from NCORP Community or Minority/Underserved Sites. Approximately 10% (100/987 with GP5 data) reported moderate/high treatment bother prior to initiating anastrozole. Anastrozole discontinuation rate at 1-year was 26.2% overall; it was lower among women with no/little treatment bother (25.7%, GP5 = 0-1) compared to moderate/high treatment bother prior to initiating anastrozole (34.7%, GP5 = 2-4; HR = 1.50, 95% confidence interval [CI]:1.04-2.15, p = 0.027). Subgroup analyses by racial cohort showed a similar predictive effect of GP5 in the White (n = 606, HR = 1.76, 95% CI: 1.12-2.77, p = 0.014) and Black (n = 184, HR = 1.85, 95% CI: 0.92-3.71, p = 0.079) cohorts, but not in the Asian cohort (n = 197, HR = 0.40, 95% CI: 0.10-1.62, p = 0.20). Conclusions: Moderate/high treatment bother prior to starting anastrozole was observed in 10% of patients and associated with a higher risk of early discontinuation, except in Asian patients. Our findings support the presence of a treatment tolerability threshold which can be compromised by pre-treatment burden. Treatment tolerability may also be influenced by cultural and genetic factors, which will be explored in further analysis of genetic and PRO data. Clinical trial information: NCT01824836.

Cluster-randomized trial to evaluate the implementation of reproductive health care in cancer care delivery in community oncology practices: Results from ECOG-ACRIN E1Q11

1519

Background: Reproductive health (RH) needs of women newly diagnosed with cancer have been poorly addressed. RH management must be aligned with cancer treatment to optimize cancer survivorship. The primary objective of the EROS trial is to evaluate the effectiveness of implementing RH programming to improve RH care among reproductive aged women with cancer. Methods: E1Q11 used a cluster randomized design with 17 NCI Community Oncology Research Program (NCORP) Sites randomized to intervention (n = 8) or usual care (n = 9). Intervention sites received study-specific training delivered via webinar and tools to support RH care implementation. Pre-menopausal women aged 15-55 years newly diagnosed with cancer and pre-initiation of treatment were eligible. The primary endpoint was defined as the delivery of RH goal-concordant management within the first 3 months since enrollment. Data were obtained through patient completed questionnaires and medical record abstraction forms at baseline and 3 months. The management rate was analyzed using generalized estimating equations (GEEs) method. Results: From 7/2016 - 4/2021, 434 women enrolled (156 intervention, 278 usual care) and 392 were analyzable. The median age was 41 years. Patients self-identified as White 67.5%; Black 21.1%; Hispanic 15.9%. Most patients had breast cancer (83.5%) and local/regional disease (69.5%). A higher proportion of patients at intervention sites (77.1%, 108/140, 90% CI: 0.71-0.83) received goal-concordant RH care compared to patients enrolled from usual care sites (61.5%; 155/252, 90% CI: 0.56, 0.67). A total of 263/392 (67.1%) patients received goal-concordant RH care within the first 3 months of enrollment. The GEE analysis demonstrated patients enrolled from intervention sites were approximately twice more likely to receive goal- concordant RH care than patients at usual care sites (odds ratio, OR = 2.11, 95% CI: 1.30, 3.44, p = 0.003). Younger age (< / = 35 years vs. > 35 years) and better ECOG performance status (PS 0 vs. PS 1-3) were statistically associated with the adoption of RH goal-concordant management (OR = 2.85, 95% CI: 1.59, 5.12, p = 0.0004 and OR = 1.94, 95% CI: 1.04, 3.63, p = 0.04, respectively). The intervention effect on the primary endpoint remained after age and PS were adjusted in the model (adjusted OR = 2.23, 95% CI: 1.30, 3.84, p = 0.004). Conclusions: The EROS trial demonstrated significant improvement of goal-concordant reproductive health management amongst racially diverse women newly diagnosed with cancer treated in community oncology practices. Sites randomized to intervention more frequently delivered reproductive care compared to usual care sites. Findings support wider implementation of this intervention to improve reproductive health care delivery, improving cancer care quality for pre-menopausal women diagnosed with cancer. Clinical trial information: NCT01806129.

Real-world comparison of outcomes in the first line treatment of metastatic melanoma according to BRAF mutation status: Nivolumab and ipilimumab therapy vs. nivolumab monotherapy

e21519

Background: In metastatic melanoma (MM) patients, nivolumab can be used as monotherapy or in combinations with ipilimumab. The combination therapy has been shown in a randomized clinical trial to provide survival benefit in BRAF mutant patients with an increased toxicity. Using real-world data, we evaluated the impact on overall survival and post-treatment hospitalizations of 1L nivo alone vs nivo+ipi. Methods: We performed a retrospective cohort study of 1L treatment of patients with MM between 1/2016 and 12/2020, utilizing administrative claims data from the Anthem Cancer Care Quality Program. Real-world overall survival (OS) was defined as time from diagnosis to death. A Cox model with inverse probability of treatment weighting (IPTW) was used to adjust for demographic and clinical features. Adjusted hazard ratios (aHR) were estimated using weighted Cox proportional hazards models. Results: Our cohort included 708 1L MM patients, of them 466 (66%) treated with nivo+ipi. There was no difference in BRAF status between the treatment groups. Patients treated with nivo+ipi were younger and had more evidence of brain and lung metastasis. Following adjustment, nivo treated had significant longer OS compared to the nivo+ipi treated (aHR: 1.73 (1.25-2.35), p = 8e-4). This difference mostly stems from increased mortality of the combo treated in the first 6 months. At 24-months, 58% of the combination therapy treated patients were alive compared to 74% in the monotherapy group (p = 0.0006). Survival benefit was only observed in BRAF WT patients. In V600E patients no survival difference was observed (aHR: 0.8 (0.4-1.6), p = 0.5). Patients treated with the combination therapy had 5-fold more hospitalizations during the first 90 days after the treatment start (46% vs 10%, p = 3e-22). Conclusions: Our retrospective cohort study demonstrated improved survival for the monotherapy in BRAF WT patients. Combination therapy was associated with significantly more treatment-related hospitalizations and deaths, mainly in the first 6 months. The increased early morbidity of the combo should be considered in treatment decision.[Table: see text]

Trends in low-value oncology care during the COVID-19 pandemic

6594

Background: Low-value services, which provide minimal patient benefit while entailing costs and risks, are prevalent in cancer care. Shifts in cancer care delivery during the COVID-19 pandemic to minimize exposure provided opportunities for health systems and clinicians to prioritize higher-value over low-value oncology services. Methods: In this retrospective cohort study, we investigated the association between the COVID-19 pandemic period and low-value cancer care practices using administrative claims from the HealthCore Integrated Research Environment, consisting of ̃65 million members managed by 14 health plans across the US. We identified commercial or Medicare Advantage members diagnosed with breast, colorectal, or lung cancer between January 2015 and March 2021. Low-value cancer care practices were identified from peer-reviewed medical literature, including ASCO and ASTRO Choosing Wisely campaigns and evidence-based pathways. Five low-value practices were studied: (1) conventional fractionation instead of hypofractionation for early-stage breast cancer; (2) off-pathway systemic therapy; (3) non-guideline-based antiemetic use for minimal-, low-, or moderate-to-high-risk chemotherapies; (4) Positron Emission Tomography/Computed Tomography (PET/CT) instead of conventional CT for staging; and (5) aggressive end-of-life care (chemotherapy ≤14 days, multiple emergency department visits ≤30 days, ICU utilization ≤30 days, hospice initiation ≤3 days, and/or no hospice before death). We used linear probability models to evaluate the association between the COVID-19 period (March to December 2020) and the 5 outcomes, adjusting for patient, facility, geographic and temporal characteristics. Results: Among 204,581 members (mean age 63.1, 139,488 [68.1%] female), 83,593 (40.8%) had breast cancer, 56,373 (27.5%) had colon cancer, and 64,615 (31.5%) had lung cancer. Rates of low-value care were similar in pre-COVID vs. COVID periods: conventional radiotherapy: 22.1% vs. 9.4%; off-pathway systemic therapy: 36.7% vs. 43.2%; non-guideline-based antiemetics: 61.2% vs. 58.1%; PET/CT imaging: 39.9% vs. 41.3%; aggressive end-of-life care: 75.8% vs. 73.3%. In adjusted analyses, the COVID-19 period was associated with no changes in off-pathway therapy (adjusted percentage point difference [aPPD] 0.82, SD 0.08, p = 0.33), PET/CT imaging (aPPD 0.10, SD 0.005, p = 0.83), and aggressive end-of-life care (aPPD 2.71, SD 0.02, p = 0.16). Small changes in conventional radiotherapy (aPPD 3.93, SD 0.01, p < 0.01) and non-guideline-based antiemetics (aPPD -3.62, SD 0.006, p < 0.01), were noted. Conclusions: The shock of the COVID-19 pandemic did not meaningfully change several metrics of low-value cancer care. Broader changes to payment and incentive design should be considered to turn the tide toward higher-value cancer care.

Real-world data comparing atezolizumab plus taxane therapy versus taxane alone as first-line treatment of metastatic triple-negative breast cancer patients in the United States

e18767

Background: A recent clinical trial (Impassion131) showed that addition of atezolizumab to paclitaxel does not provide survival benefit for first-line (1L) treatment of metastatic triple-negative breast cancer (mTNBC) patients. The study was in contradiction to a previous study (Impassion130) that did show benefit using atezolizumab plus nab-paclitaxel. Using real-world data, we evaluated the impact on overall survival and post-treatment hospitalizations of 1L Atezolizumab + Taxane (Nab-Paclitaxel) (A+T) vs. Taxane agents alone (Paclitaxel or Nab-Paclitaxel) (T) in individuals with mTNBC. Methods: We performed a retrospective cohort study of 1L mTNBC patients between 1/2016 and 6/2021, utilizing administrative claims data from the Anthem Cancer Care Quality Program. Real-world overall survival (rwOS) was defined as time from diagnosis to death. Cox model with inverse probability of treatment weighting was used to adjust for age, ECOG, Socioeconomic deprivation index (SDI), comorbidity, and liver and bone metastatic sites. Median rwOS was estimated using the weighted Kaplan-Meier method. Adjusted hazard ratios (aHR) were estimated using weighted Cox proportional hazards models. We do not adjust for PD-L1 levels shown to have association with improved OS regardless of treatment, therefore, unlikely to confound our results. Results: The A+T combination therapy became a preferred 1L type of treatment for mTNBC patients since its introduction in 2019. We identified 155 (55%) individuals with 1L mTNBC who received A+T and compared this cohort to 128 (45%) individuals who received T alone. The A+T group were younger and had lower proportion of bone metastatic sites (Table). Percentage of patients with previous chemotherapy in the last year was similar between the groups (p = 0.9). Median rwOS was 13.6 and 17.5 months, in A+T and T group, respectively (p = 0.2). In adjusted survival analysis, the addition of atezolizumab showed insignificant increase in the risk of mortality (aHR: 1.09 (0.74-1.61), p = 0.66). Our post-treatment analysis did not identify statistically significant differences between the groups, in terms of post-treatment hospitalizations, adverse events classes, or length of stay of hospitalized patients. Conclusions: Our retrospective cohort study demonstrated that the addition of atezolizumab to 1L treatment, was not associated with survival benefit or reduce post-treatment adverse events in mTNBC patients.[Table: see text]

Real-world data comparing FOLFIRINOX versus gemcitabine nab-paclitaxel as first-line treatment of metastatic pancreatic ductal adenocarcinoma patients in the United States

e16271

Background: The common first-line therapies for metastatic adenocarcinoma of the pancreas (mPC) are FOLFIRINOX and gemcitabine plus nab-paclitaxel (Gem-Nab-P). However, these treatments have not been directly compared in a clinical trial and comparative real-world data analyses on their effectiveness are limited. Using real-world data, we evaluated the impact on overall survival and post-treatment hospitalizations of 1L FOLFIRINOX vs. Gem-Nab-P in individuals with mPC. Methods: We performed a retrospective cohort study of 1L treatment of patients with mPC between 1/2015 and 6/2020, utilizing administrative claims data from the Anthem Cancer Care Quality Program. Real-world overall survival (rwOS) was defined as time from diagnosis to death. Inverse probability of treatment weighting (IPTW) was used to adjust for age, ECOG, Socioeconomic index (SDI), comorbidity, metastatic sites and pre-treatment. Median rwOS was estimated using the weighted Kaplan-Meier method. Results: Our cohort included 1,102 1L mPC patients, 566 (51.4%) treated with FOLFIRINOX (F), and 536 (48.6%) treated with Gem-Nab-P (GNP). F-treated patients were generally younger, with better performance status (ECOG PS), fewer comorbidities and living in regions with higher socioeconomic index. Following adjustments, the Median rwOS was 9.28 and 6.82 months for F-initiated patients and GNP, respectively (p-value = 2.5e-07). This survival benefit of F was observed among all sub-groups, including different ECOG PS, ages, socioeconomic index and metastatic sites. F-treated patients also had fewer post-treatment hospitalizations (p-value=0.027) and lower post-treatment costs (p-value=0.00004). Conclusions: Our retrospective cohort study demonstrated that FOLFIRINOX is associated with improved survival of approximately 2 months over Gem-Nab-P and is also associated with fewer post-treatments complications. A randomized controlled trial comparing these first line treatments is warranted to test the survival and post-treatment complications benefit of FOLFIRINOX over Gem-Nab-P.[Table: see text]

Real-world data comparing third and subsequent line regimens for treatment of metastatic triple-negative breast cancer in the United States

e18765

Background: Eribulin is a commonly prescribed regimen for metastatic triple negative breast cancer (mTNBC) patients pre-treated with one or more chemotherapy regimens, although no clear survival benefit has been shown when compared to treatment of physician's choice (TPC) [1]. Here, using real-world oncology data we compare the impact of treatment with Eribulin vs. other common chemotherapies as third and subsequent lines of therapy on overall survival and post-treatment hospitalizations. Methods: We performed a retrospective cohort study of third and subsequent lines (3L+) treatment of patients with mTNBC between 1/2016 and 6/2021 using administrative claims data from the Anthem Cancer Care Quality Program. Real-world overall survival (rwOS) was defined as time from diagnosis to death. Real-world overall survival (rwOS) was defined as time from diagnosis to death. A Cox model with inverse probability of treatment weighting was used to adjust for age, ECOG, socioeconomic status, comorbidity, and liver and bone metastatic sites. Median rwOS was estimated using the weighted Kaplan-Meier method. Adjusted hazard ratios (aHR) were estimated using weighted Cox proportional hazards models. Results: We identified 210 (52%) individuals with mTNBC who received Eribulin (E) for 3L+, compared to 189 patients treated with other chemotherapies (O), which include carboplatin+gemcitabine, nab-paclitaxel, vinorelbine, paclitaxel, carboplatin and nab-paclitaxel+atezolizumab, ordered by frequency. We found no significant difference in any of the demographic and clinical features tested between the groups (Table). Median rwOS in the E group was 6.4 months, compared to 7.6 months in the O group (p-value = 0.4). In adjusted survival analyses to known confounders, E showed no significant difference in mortality (aHR: 1.09, 95% Confidence Interval [CI]: 0.87-1.37, p-value = 0.46). Furthermore, we found no significant difference in hospitalizations following the treatment between the E to the O group (36% vs. 39%, p-value = 0.6), length of stay (7 vs. 8 days, p-value = 0.8) and classes of associated severe adverse events. Conclusions: Our analysis of real-world oncology data aligns with National Institute for Health and Care Excellence (NICE) recent recommendation in showing no survival benefit or reduction in post-treatment adverse events when treating with Eribulin at 3L+ in metastatic TNBC patients.[Table: see text]

Differences in clinician and patient assessment of baseline neuropathy in patients receiving taxane-based chemotherapy enrolled to SWOG S1714 (NCT# 03939481)

12024

Background: Chemotherapy induced peripheral neuropathy (CIPN) can lead to treatment dose reduction or discontinuation and significantly impact quality of life and functional status. Clinical trials have historically excluded patients with pre-existing neuropathy. Thus, it is unknown how many patients start treatment with baseline neuropathy as well as the impact on the trajectory of neuropathy symptoms. There are several patient- and clinician-based methods to assess CIPN; however, there is no consensus on the best method to evaluate CIPN or whether clinician versus patient assessment differs at baseline. Methods: In SWOG 1714, we enrolled patients ≥ 18 years of age with Stage I-III non-small cell lung, breast, or ovarian/fallopian tube/peritoneal cancer starting treatment with a taxane. Patients with baseline neuropathy were eligible. Neuropathy was assessed with patient-reported outcomes (PROs), including the European Organization for Research and Treatment of Cancer QLQ-CIPN20 (CIPN-20) and the PRO version of the Common Terminology criteria for Adverse Events (PRO-CTCAE) for severity of and interference caused by numbness and tingling, and the clinician-assessed National Cancer Institute (NCI)-CTCAE Grading Scale Version 5.0 for nervous system disorders. Results: Of 1336 patients enrolled on S1714, 1322 (99.0%) were eligible. The median age was 55.7 years (range 23.9-85.5) and 98.9% were female. The cohort was racially/ethnically diverse with 73.6% White, 11.3% Black, 4.6% Asian, and 10.5% Other and 10.5% Hispanic/Latino. Most of the patients enrolled had breast cancer (91%) and 67 patients (5.1%) reported having a neurological condition. Paclitaxel was administered to 60.2% and docetaxel to 39.8% and 98.5% planned to start treatment with full dose of taxane. Based on clinician assessment with NCI-CTCAE, 87.6% of patients at baseline had Grade 0 peripheral sensory neuropathy, 10.2% Grade 1, 2.0% Grade 2, and 0.2% Grade 3. The mean baseline CIPN-20 sensory subscore (range 0-100, higher number indicating greater severity) was 5.68 (standard deviation 10.41). Using the PRO-CTCAE for severity of numbness and tingling, 75.4% reported no baseline symptoms, 18.2% "mild", 4.8% "moderate", 1.1% "severe", and 0.5% "very severe" symptoms. With respect to interference of numbness and tingling with daily activities, 88.5% reported “not at all”, 8.3% “a little bit”, 2.0% “somewhat”, 0.9% “quite a bit”, and 0.3% “very much”. Conclusions: In this diverse cohort of predominantly breast cancer patients, there was limited evidence of significant pre-existing neuropathy. Clinician assessments of neuropathy may underestimate the symptoms of patients, emphasizing the importance of PROs in evaluating symptoms, particularly when baseline symptom is an exclusion criterion for clinical trials. Funding: NIH/NCI/NCORP grant UG1CA189974

Long-term results from a randomized blinded sham- and waitlist-controlled trial of acupuncture for joint symptoms related to aromatase inhibitors in early stage breast cancer (S1200)

12018

Background: Musculoskeletal symptoms are the most common side effect of aromatase inhibitors (AIs) among breast cancer (BC) survivors. We previously reported that true acupuncture (TA) resulted in better pain outcomes than either sham acupuncture (SA) or wait-list controls (WC) at 6 weeks with durable effects through 24 weeks, with minimal toxicity. We now report the 52-week outcomes. Methods: We conducted a SWOG multicenter randomized controlled trial among postmenopausal women with early-stage BC. Patients taking an AI for ≥30 days and reporting a worst pain score of ≥3 out of 10 using the Brief Pain Inventory-Worst Pain (BPI-WP) were eligible. Subjects were randomized 2:1:1 to TA vs. SA vs. WC. Both the TA and SA protocols consisted of a 12-week intervention, with 2 sessions per week for 6 weeks, followed by 1 session per week for 6 additional weeks. At 24 weeks, all subjects remained blinded to intervention arm but were offered 10 sessions of true acupuncture. Endpoints included BPI scores, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) for hips and knees, the Modified Score for the Assessment of Chronic Rheumatoid Affections of the Hands (M-SACRAH), PROMIS Pain Inventory Short Form (PI-SF), and Functional Assessment of Cancer therapy Endocrine Symptoms (FACT-ES). Results: Among 226 patients registered, 110 were randomized to TA, 59 to SA and 57 to WC. Baseline characteristics were similar among the arms. At 52 weeks, follow-up assessments were available for 91 (82.7%) TA, 53 (89.8%) SA, and 47 (82.5%) WC patients. In a linear regression adjusting for the baseline score and stratification factors, 52-week mean BPI-WP scores were 1.08 points lower (correlating with less pain) in the TA compared to SA arm (95% CI: 0.24-1.91, p =.01), and were 0.99 points lower in the TA compared to WC arm (95% CI: 0.12-1.86, p =.03). The proportion of patients experiencing a clinically meaningful (>2) reduction (i.e. improvement) in BPI-WP was 64% for TA compared to 45% on SA and 53% on WC. Patients randomized to TA had reduced BPI pain interference at 52 weeks compared to SA (adjusted difference = 0.58, 95% CI: 0.00-1.16, p =.05) but not compared to WC (adjusted difference = 0.33, 95% CI: -0.28-0.93, p =.29). Also, at 52 weeks, patients randomized to TA had improved PROMIS PI-SF T-scores compared to SA (adjusted difference = 2.35, 95% CI: 0.07-4.63, p =.04) but not compared to WC (adjusted difference = 1.28, 95% CI: -1.09-3.66, p =.29). No statistically significant differences were observed in other measures. Conclusions: Women with breast cancer receiving AI therapy and treated with 12 weeks of TA for joint symptoms had reduced levels of worst pain compared to control patients, an effect that was durable through one year despite completion of protocol acupuncture at 12 weeks, and the offering of acupuncture to all participants at 24 weeks. Clinical trial information: NCT01535066.

Randomized, phase III study of early intervention with venetoclax and obinutuzumab versus delayed therapy with venetoclax and obinutuzumab in newly diagnosed asymptomatic high-risk patients with chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL): EVOLVE CLL/SLL study (SWOG S1925, NCT#04269902)

TPS7567

Background: Currently, asymptomatic patients with CLL/SLL are observed without treatment until development of symptoms or cytopenias. Historically, early intervention studies with chemoimmunotherapy have not resulted in an overall survival (OS) benefit and have resulted in toxicity. The introduction of targeted therapies, such as venetoclax and obinutuzumab (VO), have provided tolerable/efficacious options for CLL patients. In the CLL14 study, symptomatic CLL patients receiving frontline therapy with VO had longer progression-free survival (PFS) and deeper remissions [more minimal residual disease-undetectable (MRDu)] compared with those receiving chlorambucil and obinutuzumab (Fischer 2019). The CLL-International Prognostic Index (CLL-IPI; Table) is a validated prognostic model to predict which patients are highest risk for a shorter time to first therapy and shorter OS. We aim to use VO as early intervention in asymptomatic, high-risk patients with CLL to potentially lengthen OS and thus alter the natural history of the disease. Methods: On 12/14/20, we activated the S1925 study for adult patients with CLL or SLL, who were diagnosed within 12 months of enrollment. Eligible patients have a CLL-IPI score ≥ 4 (Table) or complex cytogenetics (≥3 cytogenetic abnormalities) and do not meet any criteria for initiation of treatment by the International Working Group for CLL (IWCLL; Hallek 2018) guidelines. Enrolled patients are randomized in a 2:1 manner to early versus delayed (at the time IWCLL indication for treatment is met) therapy with VO. VO is administered for a fixed duration of 12 months as previously described (Fischer 2019). The primary endpoint is OS. We hypothesize that early intervention with VO will improve the rate of 6-year OS from 60% to 80%. This design requires 222 eligible patients for 88% power (2-sided α=0.05) for the primary comparison. To allow for 10% ineligibility, we will enroll 247 patients. Estimated accrual time is 4 years. Secondary endpoints include: rates of response, PFS, and relapse-free survival; safety; time to 2nd CLL-directed therapy; and quality of life (FACT-Leukemia total score). The primary translational objective is to evaluate the prognostic association between OS and peripheral blood MRD status at 15 months after treatment initiation by flow cytometry. Additional exploratory objectives include the association of other clinical outcomes, baseline prognostic factors, and IWCLL-defined response with MRD status at multiple timepoints. Currently, enrollment is open. Clinical trial information: NCT04269902. [Table: see text]

Real world experience with standalone immunotherapy regimens: Immune-related adverse events, healthcare utilization and cost among patients with commercial or Medicare Advantage insurance

2625

Background: Immunotherapy is a fast growing class of cancer therapy. We evaluated the rates of immune-related adverse events (irAEs), healthcare utilization, and costs up to 1 year post-index among patients using monotherapy (PD-1/PD-L1 inhibitor) and combination therapy (PD-1/PD-L1 with CTLA-4 inhibitors). Methods: We reviewed claims from the HealthCore Integrated Research Database (HIRD), which contains commercially insured/Medicare Advantage members and captures clinical, utilization, and cost measures. We analyzed both the monotherapy (M) and combination therapy (C) cohorts focusing on members with ≥ 6 months of baseline continuous medical and pharmacy coverage. Descriptive and multi-variate analyses were performed. Results: The C cohort had 904 and M had 9,084 patients, with mean ages of 58 and 64 years, respectively. Prominent cancer types were melanoma for C and lung for M. The most common incident irAEs (%) for C vs. M were: endocrinopathies (27.7, 14.7), hepatitis (17.1, 7.7), nephritis (21.0, 14.0), neuropathy (6.6, 7.0), followed by colitis, dermatitis, and myocarditis. After adjustment, C therapy showed greater risk of all-cause inpatient admissions (OR 2.27, 95% CI 1.93, 2.66), all-cause emergency department (ED) visits (OR 1.55, 95% CI 1.33, 1.81) and irAE-related visits (See table). Mean adjusted all-cause cost difference for C vs M was +$43,747 (95% CI $38,440, $49,427). In age ≥65 subset, 222 received C and 4,208 received M. C therapy patients had more irAE-related hospitalizations (45.3% vs. 57.7%, p=0.0004). Costs were similar to the main cohort. Conclusions: C therapy showed greater incident irAE rates, increased utilization and medical costs compared to M therapy. Limitations include less precise ascertainment of irAEs in claims data and generalizability only to those with commercial or Medicare Advantage insurance. Our study highlights the increased toxicity and cost tradeoffs involved in choosing combination immunotherapy over monotherapy.[Table: see text]

Changes in prescribing of oral capecitabine versus intravenous (IV) 5-fluorouracil (5-FU) in gastrointestinal (GI) cancers during the COVID-19 pandemic

e18596

Background: Several oncology guidelines recommend using oral drugs vs. IV to minimize COVID-19 risk for patients with cancer. We examined the association between prescribing patterns of oral capecitabine vs. IV 5FU for GI cancers and social distancing, measured by the change in population mobility patterns in response to shelter-in place policies, during the pandemic. Methods: Using claims data for commercially insured members, we included patients 18 years of age or older with colorectal, gastroesophageal, or pancreatic cancer, who had continuous health plan coverage for at least 2 months before and 1 month after initiating chemotherapy with capecitabine or 5-FU from January 2017 to August 2020. We analyzed unadjusted trends in proportion of chemotherapy that was oral during pandemic (March 1st to August 31st, 2020) compared to previous years. Then, we conducted difference-in-differences analysis using COVID-19 Community Mobility Reports, by Google, and utilizing different levels of changes in mobility trends across states over time. In our main model, we used a 20% decrease in retail and recreation visits as our threshold and compared the prescribing rates in states below and above the threshold as well as before and after the pandemic began. We also used different thresholds and categories of places to check the sensitivity of our findings. Models are adjusted for age, gender, month of year, urban status, comorbidities, and state of residence at chemotherapy start date. Results: A total of 17,414 nationally distributed patients (69% colorectal, 13% gastroesophageal, 18% pancreatic) were included (mean age, 58.8 years; 41% female). During the pandemic, 1,875 patients (65% colorectal, 15% gastroesophageal, 20% pancreatic) were identified. The proportion of oral regimens did not change significantly for colorectal and gastroesophageal patients and decreased by 7.4 percentage points (pp) (p < 0.01) for pancreatic patients. In regression modelling with mobility data, oral prescribing rates for colorectal patients increased by 3.1 pp (p < 0.01), largely driven by increases for female patients (9.2 pp, p = 0.02). We observed a decrease in oral prescribing rates among pancreatic patients (-1.20 pp, p = 0.04) and did not observe a significant change for gastroesophageal patients. Our results are not sensitive to different social distancing specifications. Conclusions: We observed differential impact of the pandemic on oral prescribing rates by GI cancer type and gender. Oral prescribing increased among colorectal cancer patients driven mostly by higher oral prescribing in females. For pancreatic and gastroesophageal patients, oral prescriptions either remained unchanged or decreased. This observation may reflect a variable impact of the pandemic on women as compared to men and might involve heightened caregiving responsibilities for women.

Prospective validation of genetic predictors of aromatase inhibitor-associated musculoskeletal symptoms (AIMSS) in a racially diverse cohort: Results from ECOG-ACRIN E1Z11

12003

Background: AIMSS are common and frequently lead to early discontinuation of adjuvant AI therapy. Single nucleotide polymorphisms (SNPs) in candidate genes have been associated with AIMSS and AI discontinuation. The primary objective of E1Z11 was to validate previously identified associations between 10 specific SNPs in candidate genes and AI discontinuation due to AIMSS in a community-based, racially diverse cohort. Methods: Postmenopausal women with hormone receptor-positive stage I-III breast cancer enrolled onto a prospective multi-site cohort study, the majority through the NCI Community Oncology Research Program (NCORP). Participants received anastrozole 1 mg oral daily, and completed patient-reported outcomes (PROs) at baseline, 3, 6, 9, and 12 months. AIMSS was defined as >20% increase in Stanford Health Assessment Questionnaire (HAQ) score over baseline occurring within 1 year of AI therapy. We projected 40% would develop AIMSS and 25% would discontinue AI treatment within 1 year, informing a planned enrollment of 1000 women with a fixed number per strata (600 Caucasian, 200 African-American [AA] & 200 Asian) to provide 80% power to detect an effect size of 1.5-4. SNPs include ESR1 (rs2234693, rs2347868, rs9340835), CYP19A1 (rs1062033, rs4646), TCL1A (rs11849538, rs2369049, rs7158782, rs7159713), and HTR2A (rs2296972). Hardy-Weinberg equilibrium (HWE) was evaluated within each racial subset. SNP genotypes were coded as additive effects on the log odds ratio by coding as 0, 1 or 2 for the count of the minor allele. A Cochran-Armitage trend test was used with a 1-sided alpha of 0.0025 (Bonferroni correction for 10 tests). Results: We enrolled 999 evaluable women (616 Caucasian, 184 AA, 199 Asian). Genotyping was successful in 974 (98%). AIMSS developed in 43%, and AI therapy was discontinued in 12% within 1 year. While more AA and Asians developed AIMSS compared to Caucasians (48% vs 38%, p=0.017; 50% vs 38%, p=0.004), AI discontinuation rates were similar across racial groups. HWE was satisfied for all SNPs at the 5% alpha level, except for TCL1A/rs11849538 (p=0.002) in the AA cohort. None of the 10 SNPs were significantly associated with AI discontinuation or development of AIMSS in the overall population, or in any of the 3 cohorts. Conclusions: Although AIMSS were more common in AA and Asians, AI discontinuation rates were similar in the 3 cohorts. We were unable to prospectively validate 10 SNPs in 4 genes previously associated with AI discontinuation due to AIMSS. Future analyses will include other predictors of AIMSS, PROs, and additional genetic markers for the entire cohort and by race. Support: NCI UG1CA189828, UG1CA233196, UG1CA233277, UG1CA233320, UG1CA233178, UG1CA233160, UG1CA232760, UG1CA233341, UG1CA233329, U10CA180821, UG1CA189821, UG1CA189830, U10CA180888, UG1CA189859, UG1CA189863, UG1CA189971. Clinical trial information: NCT01824836.

Association between a national insurer’s pay-for-performance program for oncology and changes in prescribing of evidence-based cancer drugs and spending

2016

Background: Efforts to standardize quality and control cost growth for cancer care have focused heavily on promoting evidence-based cancer drug prescribing. We evaluated the association between a national commercial insurer’s ongoing pay-for-performance (P4P) program for oncology and changes in prescribing of evidence-based cancer drugs and spending. Methods: Retrospective difference-in-differences quasi-experimental study utilizing administrative claims data from the insurer’s commercial health plans in 14 states covering 6.7% of US adults. We included patients 18 years of age or older with breast, colon, or lung cancer who were prescribed cancer drug regimens by 1,867 participating oncology physicians between 2013 and 2017. We leveraged the geographically staggered, time-varying rollout of the P4P program to simulate a stepped-wedge study design. Specifically, we estimated a patient-level model clustered by physician and used physician fixed-effects to examine pre- to post-intervention changes in evidence-based prescribing and spending for patients of participating physicians eligible earlier versus later in the period of P4P program rollout. We evaluated four categories of spending over a 6-month episode period: cancer drug spending; other (non-cancer drug) health care spending; total episode spending; and patient out-of-pocket spending. Results: The P4P program was associated with an increase in evidence-based regimen prescribing from 57.1% of patients in the pre-intervention periods to 62.2% in the post-intervention periods for a difference of +5.1 percentage points (pp) (95% CI 3.0 pp to 7.2 pp, P< 0.001). The P4P program was also associated with a differential $3,235 (95% CI $1,004 to $5,466, P= 0.005) increase in cancer drug spending, a differential $253 (95% CI $101 to $406, P= 0.001) increase in patient out-of-pocket spending, but no significant changes in other health care spending or total health care spending over the 6-month episode period. Conclusions: A national insurer’s oncology P4P program was associated with a 5.1 percentage point increase in prescribing of evidence-based cancer drug regimens. Our findings suggest that P4P programs may be effective in increasing evidence-based cancer drug prescribing at national scale -- enhancing cancer care quality. However, they may also increase out-of-pocket expenses and may not lead to savings in total health care spending during the 6-month episode.

Incidence of and risk factors for hospitalizations from chemotherapy among patients with stage III and stage IV colorectal cancer

e16090

Background: Despite the decreasing colorectal cancer (CRC) mortality rate over the past decade, complications from CRC treatment remain a challenge. Prior research has shown that a majority of patients with stage III CRC in the adjuvant setting experience hospitalizations due to chemotherapy-related toxicity. Minimal research, however, has examined risk factors of these events and the prevalence of hospitalization among stage IV CRC patients. Methods: We used claims data from a geographically-diverse private health insurer—including both commercially-insured and Medicare Advantage patients—to estimate and characterize risk factors of hospitalizations among Stage III or IV CRC patients. We compared sociodemographic, clinical, as well as provider characteristics and cancer treatment regimens between patients with and without hospitalizations from the initiation of chemotherapy to 60 days after the end of chemotherapy. Results: Incidence rates for hospitalization from chemotherapy were 49% and 70% for stage III and IV CRC patients, respectively. Although the oldest stage III CRC patients (age 75+) were the most likely to experience hospitalizations, the youngest age group (age 18-49) of stage IV patients experienced the highest incidence (74%) of hospitalizations (p < 0.05). Higher values of the Elixhauser comorbidity index was associated with a higher risk for hospitalizations among patients with stage III CRC (p < 0.001). Both stage III and stage IV patients with diabetes were more likely (p < 0.05) to have hospitalizations from chemotherapy (55% and 73%, respectively). Conclusions: Hospitalization from chemotherapy is very common among stage III and IV CRC patients. These data identify subgroups at higher risk. Study findings may inform choice of cancer treatment regimen and focus on key underlying medical needs

The use of optimal evidence-based chemotherapy (chemo) regimens in physician offices versus hospital outpatient facilities

6514

Background: The proportion of infused chemo administered in hospital outpatient facilities (HOF) increased from 6% in 2004 to 43% in 2014. The average annual cost for patients receiving chemo was significantly higher in HOFs than in physician offices (POs). One option to explore differences in the quality of care between these two settings is to examine the use of chemo regimens, which, based on their efficacy, toxicity, and costs, have been designated as “on-pathway.” This study compared on-pathway rates among patients receiving infused chemo administered in POs vs. those in HOFs. Methods: Using administrative claims data, we identified 61,496 breast, lung, or colorectal cancer patients receiving chemo from 2013 to 2018. Chemo regimens were considered “on-pathway” when they were on payer's program list of optimal regimens when administered. Generalized linear models examined the association between site of service and on-pathway prescribing rates, and costs of care. Models adjusted for age, sex, year, rural status, cancer type and setting, and comorbidities, with fixed effects for providers. Results: Percentage of infused chemo administered in HOFs increased from 44.2% in 2013 to 54.7% in 2018. After adjustment, on-pathway prescribing rate did not differ significantly between HOFs and POs (50.1%, 95% CI: 48.6%-51.5% vs. 49.8%, 95%CI: 48.3%-51.3%, p = 0.65). 6-month chemo cost ($56,885, 95% CI: $54,364-$59,524 vs $32,240, 95% CI: $30,929-$33,605, p < 0.001) and 6-month medical cost ($114,280, 95% CI: $110,716-$117,960 vs $79,455, 95% CI: $77,089-$81,893, p < 0.001) were significantly higher in HOFs vs. POs. Conclusions: Quality of care as measured by use of optimal chemo regimens was similar in hospital and office setting. Cost continues to be significantly higher in hospital setting. These findings provide a strong basis for site-neutral reimbursement policies.

Sex differences in adverse event reporting in SWOG chemotherapy, biologic/immunotherapy, and targeted agent cancer clinical trials

11588

Background: Women have more adverse events (AEs) from chemotherapy than men, but few studies have explored sex differences in biologic/immunotherapies (BIs) or targeted therapies. We examined subjective (symptomatic) and objective AEs by sex across different treatments. Methods: We analyzed drug-related severe (grade 3) or worse AEs by sex in SWOG phase II and III clinical trials conducted between 1980-2018, excluding sex-specific cancers. AE codes and grade were categorized using the Common Terminology Criteria for Adverse Events (CTCAE). Subjective or symptomatic toxicities were defined as those aligned with the NCI’s new Patient-Reported Outcome (PRO) CTCAE; lab-based or physician-determined AEs were designated as objective. Multivariable logistic regression was used, adjusting for age, race, and disease prognosis. Thirteen symptomatic and 19 objective AE categories were examined. Results: In total, 36,397 patients (women, 13,907 [38.2%]; men, 22,490 [61.8%]) experiencing 522,835 AEs on 297 trials with 385 treatment arms were analyzed. Overall, 29.1% (n = 10.860) had severe or worse toxicity. Women experienced an increased risk of severe symptomatic AEs for BIs (OR = 1.53, 95% CI: 1.32-1.78, p < .0001), chemotherapy (OR = 1.31, 95% CI: 1.24-1.39, p < .0001), and targeted therapies (OR = 1.23, 95% CI: 1.06-1.43, p = .008). Women also had an increased risk of severe objective AEs for BIs (OR = 1.53, 95% CI: 1.32-1.78, p < .0001), chemotherapy (OR = 1.35, 95% CI: 1.28-1.43, p < .0001), but not targeted therapies (OR = 1.08, 95% CI: 0.94-1.25, p = .28). Across all treatments, sex differences were greater for hematologic (OR = 1.29, 95% CI: 1.24-1.35, p < .0001) v. non-hematologic (OR = 1.13, 95% CI: 1.08-1.18, p < .0001) objective AEs. Conclusions: The greater severity of both symptomatic and objective – especially hematologic – AEs in women across multiple treatment paradigms indicates broad-based sex-differences exist. This could be due to AE reporting, pharmacogenomics of drug metabolism and disposition, total dose received, and/or adherence to therapy. Particularly large sex differences were observed for patients receiving BIs, suggesting studying AEs from these agents is a priority.

Reproductive health in cancer

TPS6649

Background: Reproductive health needs of females ages 15 to 55 with cancer are poorly understood and overlooked, despite their importance to patients. Uncertainties regarding fertility and pregnancy are intricate and challenging for a patient and a cancer care team. The goals of reproductive health care may come in conflict with the primary objective of cancer care. However, with appropriate counseling and preventive measures, this conflict can be absolved to unify objectives. The primary objective of this trial is to evaluate the success of the implementation of reproductive health programming among reproductive aged females with cancer. To expand knowledge on this subject, the NCI approved the ECOG-ACRIN EROS trial: Engendering Reproductive Health within Cancer Survivorship, with two ancillary studies: Endocrine Disruption in Cancer Care and Sexuality in the first 5 years after Cancer Diagnosis. Methods: The primary study is a multicenter, cluster randomized control trial, with NCI Community Oncology Research Programs randomized to either the non-intervention arm (usual standard practice related to reproductive health) or the intervention arm (using study-specific training and tools). The accrual goal is 668 patients based on the expectation that the intervention can increase the adoption of appropriate reproductive health management within 3 months from the baseline visit from 50% to 80%, with the first 200 to consent to the endocrine disruption substudy. All patients participate in the sexuality part. Pre-menopausal female patients ages 15 to 55 with an initial diagnosis of any type of cancer who have not initiated treatment of any type are eligible to participate. Patients are asked to complete 2 questionnaires at 8 timepoints regarding their reproductive health interests. Providers are also asked to complete questionnaires regarding their healthcare practice in general as well as specific to patients enrolled in this study. The reproductive health management rate at each time point will be summarized, by arm, with frequency and percentage along with its 95% confidence interval. The comparison between the two arms, using a GEE model, can evaluate the intervention effect. EROS was activated in September 2015. To date, 264 patients have been enrolled to the main study and half of these patients are participating in the endocrine disruption correlative. Outcomes pertaining to management and treatment implementation and modification are the cornerstones of this study. It should inform organizations in cancer care to improve guidelines and to include a reproductive health assessment for all young females with cancer. Clinical trial information: NCT01806129.

Osteonecrosis of the jaw in patients with cancer receiving zoledronic acid for bone metastases: SWOG S0702, NCT00874211

11502

Background: Osteonecrosis of the jaw (ONJ) may occur in cancer patients (pts) with metastatic bone disease (MBD) treated with bone modifying agents. No large prospective studies have precisely determined the incidence of ONJ. A better understanding of the true incidence and predictors of ONJ is needed. Methods: SWOG S0702 was a prospective observational study that assessed the cumulative incidence (CI) of ONJ at 3 years in pts with MBD from any malignancy receiving zoledronic acid (Zol). Participants must have had either limited or no prior exposure to bone modifying agents and a clinical care plan that included use of Zol within 30 days of registration. Cancer treatments, bone modifying agents (including Zol), and dental care were administered as clinically indicated and were not directed by S0702. Baseline and every 6 m followup dental exams were recommended. Report forms (medical, dental and pt reported outcomes) were submitted every 6 m but if ONJ was diagnosed, follow up interval became every 3 m. Protocol defined ONJ required exposed bone in the maxillofacial region present 8 weeks or more in a pt who was receiving or had been exposed to a bisphosphonate, and had not had radiation therapy to the craniofacial region. Results: The study enrolled 3,491 evaluable pts (breast 1,120; myeloma 580; prostate 702, lung 666, other 423) between 2009-2013. About 2/3 of pts had a baseline dental exam. Overall, 87 pts had confirmed ONJ. The cumulative incidence of ONJ was 0.8% at year 1 (95% CI: 0.5%-1.1%), 2.0% at year 2 (95% CI: 1.5%-2.5%), and 2.8% at year 3 (95% CI: 2.3-3.5%). Rates of 3-year confirmed ONJ were highest in myeloma pts (4.3%; 95% CI, 2.8%-6.4%). Pts with planned Zol dosing intervals of every 3-4 weeks (n = 3,032, 87.2%) were much more likely to experience ONJ than pts with planned dosing intervals of 5 weeks or greater (n = 447, 12.8%; 3.2% vs 0.7%; HR = 4.80, 95% CI, 1.52-15.18, p = .008). Fewer total number of teeth, the presence of dentures and any oral surgery at baseline were all associated with a higher rate of ONJ. Conclusions: About 1 in 40 patients receiving Zol for MBD developed ONJ. S0702 provides information to guide stratification of risk for developing ONJ in pts with MBD receiving Zol. Cancer type, oral health and frequency of Zol dosing affect risk of ONJ. Clinical trial information: NCT00874211.

Abstract P1-11-04: Association between body mass index (BMI) and response to duloxetine for aromatase inhibitor (AI)-associated musculoskeletal symptoms (AIMSS)

Background: AIMSS occurs often in women treated with AI therapy for early stage breast cancer and can negatively impact adherence and persistence with therapy. Duloxetine is a serotonin norepinephrine reuptake inhibitor used to treat mood disorders and chronic pain. In SWOG S1202, patients with AIMSS treated with duloxetine reported statistically significant improvement in pain by 12 weeks compared to placebo. Obesity is a predictor of increased likelihood of developing AIMSS, and a prior study of omega 3 fatty acid versus placebo for AIMSS showed a potential differential response to therapy by BMI. In this exploratory analysis of S1202, we investigated the association between baseline BMI and response to therapy.

Methods: In S1202, 299 postmenopausal women with stage I-III hormone receptor-positive breast cancer on AI therapy who developed new or worsening average pain of 4-10 on a numerical rating scale were enrolled, randomized 1:1 to duloxetine or placebo with randomization stratified by baseline pain (4-6 vs. 7-10) and prior taxane therapy (yes vs. no). Patients were treated for 12 weeks. Patient-reported outcomes including Brief Pain Inventory (BPI) were obtained at baseline and weeks 2, 6, 12, and 24. Patients were categorized into BMI&lt;30 kg/m2 (non-obese) or BMI≥30 kg/m2 (obese). The pre-specified aim of this secondary analysis was to examine whether the effect of intervention on BPI average pain at 12 weeks differed between obese and non-obese patients. Multiple linear regression was used, adjusting for the stratification factors and the baseline score. We tested whether the interaction of BMI status and intervention effect was statistically significant at α=.05.

Results: 289 patients were eligible for the analysis, 54% of whom were obese. The cohorts were well balanced other than by race. The difference by intervention arm in the 12-week mean BPI scores between baseline and follow-up scores was substantially different for the obese versus non-obese cohorts. In the patients with BMI&lt;30, the reduction in observed mean average pain score was similar in the duloxetine- and placebo-treated patients (-2.46 points vs. -2.34 points, p=.75). In contrast, in the patients with BMI≥30 the reduction in pain score was statistically significantly greater for the duloxetine-treated compared to the placebo-treated patients (-2.73 points vs. -1.64 points, p=.003; interaction p-value=.02). Differences in intervention effects between obese and non-obese groups were even stronger at 2-weeks (interaction p-value=.001) and 6-weeks (interaction p-value&lt;.0001). Similar findings were evident for other pain-related patient-reported outcomes.

Conclusions: In the placebo-controlled S1202 trial, obese patients with AIMSS obtained more analgesic benefit from duloxetine. Additional studies are warranted to determine the biologic basis for these findings, such as a different mechanism underlying development of AIMSS or pain expression in patients with obesity, or other confounding variables related to analgesic response to duloxetine relative to placebo.

Support: NIH/NCI grants CA189974, CA189821, CA180820; and in part by Damon Runyon-Lilly Clinical Investigator Award #CI-53-10 [to NLH], and in part by Lilly USA, LLC.

Citation Format: Henry NL, Unger JM, Till C, Schott AF, Crew KD, Lew DL, Fisch MJ, Moinpour CM, Wade JL, Hershman DL. Association between body mass index (BMI) and response to duloxetine for aromatase inhibitor (AI)-associated musculoskeletal symptoms (AIMSS) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-11-04.

Real-world comparison of biosimilar filgrastim and reference filgrastim in cancer patients receiving chemotherapy

85

Background: The approval pathway for biosimilars was established in 2010 to facilitate drug competition and generate cost savings. Biosimilar filgrastim products (filgrastim-sndz or tbo-filgrastim) were the first approved biosimilars in the US. This study examined study outcomes (incidence of febrile neutropenia (FN), growth factor related adverse events (AE) and filgrastim treatment cost) between users of biosimilar and reference filgrastim drugs among the commercially-insured US population. Methods: A retrospective observational design was used to analyze administrative claims data on adults with cancer treated with chemotherapy between 09/01/2012 - 4/30/2017. The population was grouped into biosimilar filgrastim users (B) and filgrastim users (F) according to the earliest filgrastim use (index date). FN risk factors were assessed within 6 months prior to the index date. Study outcomes were assessed within 21 days post index date using multivariable regression modelling. Results: Among 1,694 biosimilar users and 10,460 filgrastim users, no meaningful differences in baseline FN risk factors were observed. In adjusted results, no difference in incidence of FN was observed between the two groups (B: 13.3%; F: 11.0%; odds ratio [OR]: 1.17; 95% confidence intervals [CI] 0.99-1.38; p = 0.06). Rates of AE were similar between the two groups (B: 6.9%; F: 6.4%; OR: 1.04; 95% CI: 0.84-1.29, p = 0.72). The mean number of filgrastim doses administered was 3.4 in both groups (Relative difference: 0.01; 95% CI: -0.02-0.04; p = 0.53). The mean filgrastim drug costs were similar between the two groups (B: $2,491; F: $2,510, Cost Ratio [CR]: 0.96; 95% CI: 0.91-1.01; p = 0.12). However, the mean filgrastim drug cost was 6.7% less for filgrastim-sndz when compared to reference filgrastim (sndz: $2,342; CR: 0.91; 95% CI: 0.84-0.99; p = 0.04); but not for tbo ($2,549; CR: 0.97; 95% CI: 0.91-1.04; p = 0.40). Conclusions: Biosimilar filgrastim is similar to the reference filgrastim in drug safety and effectiveness. The early adoption and saving impact of biosimilar filgrastim has been modest, which highlights the structural and regulatory challenges to biosimilar uptake.

A randomized phase III double-blind clinical trial (S1600) evaluating the effect of immune-enhancing nutrition on radical cystectomy outcomes

TPS529

Background: This phase III trial will compare the impact of consuming Specialized IMmunonutrition (SIM) to oral nutritional support on postoperative complications after radical cystectomy. Specialized IMmunonutrition is fortified with nutrients (L-arginine, omega-3 fatty acids, dietary nucleotides, and vitamin A) that have immediate effects on immune and inflammatory responses, muscle sparing, and wound healing after surgery. Methods: Two hundred patients will be randomized in a 1:1 fashion to one of two arms. Subjects will be stratified by diversion type (neobladder vs. ileal conduit) and whether the patient had neoadjuvant chemotherapy (any vs. none) to balance the intervention assignment according to important prognostic factors. Nutrition drinks are given three times a day for 5 days before and 5 days after surgery. The primary endpoint is 30 day overall complication rate. Two hundred patients gives 80% power to detect an absolute reduction of 23% in the 30 day overall complication rate (from 65% down to 42%). Secondary endpoints will assess infections, muscle mass, readmissions, quality of life, recurrence, DFS, OS, immune response, cytokines, amino acids, and fatty acids. The trial is testing a high yield, low-risk, low-cost strategy to improve the outcome of patients with bladder cancer who undergo cystectomy. Funding: NIH/NCI/DCP grant award UG1CA189974. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Abstract GS4-04: Randomized blinded sham- and waitlist-controlled trial of acupuncture for joint symptoms related to aromatase inhibitors in women with early stage breast cancer (S1200)

Background: Musculoskeletal symptoms are the most common side effect of aromatase inhibitors (AIs) and can result in decreased quality of life and discontinuation of therapy. Pilot data from two prior single institution studies showed that acupuncture decreased AI-induced joint symptoms in breast cancer (BC) patients.

Methods: We conducted a SWOG multicenter randomized controlled trial among postmenopausal women with early stage BC. Patients taking an AI for ≥30 days and having a worst pain score of ≥3 out of 10 using the Brief Pain Inventory (BPI-WP) were eligible. Subjects were randomized at a 2:1:1 ratio to true acupuncture (TA) vs. sham acupuncture (SA) vs. waitlist control (WC). The TA protocol used a standardized protocol of body and auricular acupoints tailored to joint symptoms. The similarly standardized SA protocol utilized superficial needling of non-acupoints. Both the TA and SA protocols consisted of a 12 week intervention, with 12 sessions administered over 6 weeks, followed by 1 session per week for 6 additional weeks. The primary endpoint was change in the BPI-WP (worst pain) score at 6 weeks. Secondary outcomes included other BPI scores, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) for the hips and knees, the Modified Score for the Assessment of Chronic Rheumatoid Affections of the Hands (M-SACRAH), and functional testing with grip strength and "Timed Get Up and Go" (TGUG). The design specified alpha=.025 two-sided tests to account for two independent comparisons (TA vs. SA and TA vs. WC).

Results: Among 226 patients registered, 110 were randomized to TA, 59 to SA and 57 to WC. Baseline characteristics were similar between the groups. In a linear regression adjusting for the baseline score and stratification factors, 6-week mean BPI-WP scores were 0.92 points lower (correlating with less pain) in the TA compared to SA arm (95% CI: 0.20-1.65, p=.01), and were 0.96 points lower in the TA compared to WC arm (95% CI: 0.24-1.67, p=.01). The proportion of patients experiencing a clinically meaningful (&gt;2) reduction (i.e. improvement) in BPI-WP was 58% for TA compared to 33% on SA and 31% on WC. Patients randomized to TA had improved symptoms compared to SA at week 6 according to all other BPI pain measures (average pain, p=.04; pain interference, p=.02; pain severity, p=.05; worst stiffness, p=.02). Results were similar compared to WC. Patients randomized to TA compared to SA or WC had statistically significant or marginally statistically significant improvements in BPI pain measures at week 12. Patients randomized to TA had generally improved symptoms compared to SA or WC at week 6 and at week 12 according to the M-SACRAH and WOMAC measures (p&lt;0.05). With regard to adverse events, more patients on the TA arm experienced Grade 1 bruising compared to SA (47% vs. 25%, p=.01). No other differences by arm for selected adverse events were observed.

Conclusions: This study was the first large multicenter trial to investigate the effect of acupuncture in treating AI-induced joint symptoms in BC patients. According to multiple measures, TA generated better outcomes than either SA or WC with minimal toxicity.

Citation Format: Hershman DL, Unger JM, Greenlee H, Capodice J, Lew DL, Kengla AT, Melnik MK, Jorgensen CW, Kreisle WH, Minasian LM, Fisch MJ, Henry L, Crew KD. Randomized blinded sham- and waitlist-controlled trial of acupuncture for joint symptoms related to aromatase inhibitors in women with early stage breast cancer (S1200) [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr GS4-04.

Occurrence and characteristics of hospitalizations during first-line chemotherapy among individuals with metastatic colorectal cancer

691

Background: Choosing chemotherapy for metastatic colorectal cancer (mCRC) requires balancing clinical effectiveness and risk of complications. This study characterized real-world inpatient/ER hospitalizations (HOSP) during first-line chemotherapy among individuals with mCRC. Methods: We conducted a retrospective cohort study of adults with mCRC identified using claims data from the HealthCore Integrated Research Environment as initiating first-line chemotherapy from 12/23/2013 to 06/30/2016 (no minimum follow-up). Cohorts were analyzed in aggregate and for the most frequently observed first-line agents (5 overlapping subcohorts). HOSPs were identified from initiation of first-line chemotherapy to 30 days after the end of first-line chemotherapy or last available data. Results: A total of 717 individuals (mean age 55y; 58% male; 44%/39%/6%/12% with ECOG = 0/1/2+/missing; median follow-up 116 days) met study criteria. Metastasis was most commonly to the liver (51%) and 53% of patients had cancer-attributable morbidities. Chemotherapies included 5-FU (79%), oxaliplatin (67%), bevacizumab (58%), irinotecan (21%), and capecitabine (19%). Overall, 40% of patients had ≥1 HOSP [n = 285; total 415 events], ranging from 38% to 49% across the 5 chemotherapy-based subcohorts; 12% (n = 85) had > 1 HOSP. The median time to first HOSP for patients with an event was 52 days. The median length of inpatient stays was 4 days; Infections/neutropenia (21%), bowel-related complications (17%), cardiac and circulatory disorders (9%), malnutrition (5%), pain (5%) and renal disease (2%) were the most common issues associated with inpatient HOSPs. An increase in HOSPs was observed with worsening ECOG status: 0 (34%), 1 (46%), and 2+ (65%). In regression analyses, ECOG≥1 was associated with a 64%-72% increase (p < 0.01) in the odds of HOSPs compared to patients with ECOG = 0. Conclusions: Approximately 40% of mCRC patients had hospitalizations during the study period. Hospital stays were typically short and associated with infections, neutropenia, or bowel-related complications. Further research is needed to determine how many of these hospitalizations may be avoidable.

Abstract S5-06: Randomized, placebo-controlled trial of duloxetine for aromatase inhibitor (AI)-associated musculoskeletal symptoms (AIMSS) in early stage breast cancer (SWOG S1202)

Background: Adherence to AI therapy for adjuvant treatment of hormone receptor-positive breast cancer is poor, primarily because of AIMSS. Premature discontinuation of AI therapy can lead to increased likelihood of breast cancer recurrence. Duloxetine (dulox) is a serotonin norepinephrine reuptake inhibitor that is FDA-approved for treatment of multiple chronic pain disorders. Phase II data from an open label trial of dulox for treatment of AIMSS demonstrated a 61% improvement in pain. We hypothesized that treatment of AIMSS with dulox would improve average joint pain compared to placebo (plac).

Methods: Postmenopausal women with stage I-III breast cancer who had been taking AI therapy for between 3 wks and 36 mo were enrolled. To be eligible, patients were required to have average pain of ≥4/10 using the Brief Pain Inventory (BPI) that developed or worsened since AI therapy initiation, and not have any contraindications to dulox therapy. Patients were randomized 1:1 to dulox 30 mg daily for 7 d then 60 mg daily for 11 wks then 30 mg daily for 7 d, or to matching plac, stratified by baseline pain (4-6 vs 7-10) and prior taxane use (yes vs no). Pain, depression, and quality of life (QoL) were assessed after 2, 6, and 12 wks of therapy, as well as at the 24 wk time point. The primary analysis used linear mixed models to examine average pain through 12 wks by arm, adjusting for the stratification factors and assessment time. Clinically significant change in average pain was defined as a ≥2-point decrease from baseline.

Results: 299 patients were randomized between June 2013 and October 2015, 10 of whom were ineligible. 127 dulox-treated and 128 plac-treated patients were evaluable for the primary analysis. No sizeable imbalances in baseline factors were noted by arm. Seventeen pts reported grade 3 adverse events (AEs) (dulox: 12/138 (8.7%), plac: 5/141 (3.5%)), and 40 pts discontinued treatment because of AEs (dulox: 21 (52.5%), plac: 19 (47.5%)). Mean observed average pain, the proportion of pts experiencing clinically significant change in average pain from baseline, and percent reduction in average pain all indicated greater improvement for dulox compared with plac through 12 wks, but were similar by arm at wk 24 (12 wks after completion of intervention; see Table). In multivariable linear mixed model analysis, the BPI average pain was reduced on average by 0.82 points more on dulox compared to plac over the first 12 wks (95% CI -1.24 to -0.40, p=0.0002). Similar patterns were observed for worst pain, pain interference, joint pain, stiffness, and functioning, and QoL.

Table: Observed Average Pain Scores by Assessment TimeTime PointBaseline2 weeks6 weeks12 weeks24 weeksduloxplacduloxplacduloxplacduloxplacduloxplacAverage pain5.445.493.514.412.953.962.913.453.373.42Percent reduction--34%20%46%28%46%36%37%37%Patients with clinically significant change--54%44%69%49%69%60%60%59%

Conclusions: Treatment with duloxetine was superior to placebo for the treatment of AIMSS among women with early stage breast cancer, was well tolerated, and was associated with improvements in QoL.

Clinicaltrials.gov NCT01598298.

Citation Format: Henry NL, Unger JM, Schott AF, Fehrenbacher L, Flynn PJ, Prow D, Sharer CW, Lew DL, Moseley A, Fisch MJ, Moinpour C, Hershman DL, Wade III JL. Randomized, placebo-controlled trial of duloxetine for aromatase inhibitor (AI)-associated musculoskeletal symptoms (AIMSS) in early stage breast cancer (SWOG S1202) [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr S5-06.

Use of the aromatase inhibitor anastrozole in the treatment of patients with advanced prostate carcinoma

BACKGROUND

Men with prostate carcinoma initially respond to therapies designed to inhibit androgen secretion or block its action. Later, the tumors in these patients become refractory to androgen-related therapies. Therefore, additional hormonal maneuvers that would benefit these men currently are needed. Reports of androgen receptor mutations and historic clinical observations raised the hypothesis that estrogens might be involved in the proliferation of androgen-refractory prostate carcinoma.

METHODS

To explore this hypothesis, 14 men with advanced prostate carcinoma that was refractory to medical or surgical orchiectomy and antiandrogens were entered into a clinical Phase II trial involving suppression of estrogens. After complete evaluation, each patient received 1 mg daily of the third-generation aromatase inhibitor anastrozole until disease progression. Follow-up included serial determinations of prostate specific antigen (PSA), measurements of evaluable lesions, and assessment of intensity of pain.

RESULTS

No patient experienced an objective response or disease stabilization as measured by PSA level or the greatest dimension of the lesion. Minimal improvement of bone pain was reported in two patients receiving intensive analgesic medication.

CONCLUSIONS

It was concluded that the dependence of androgen-insensitive prostate carcinoma on estrogens for proliferation is uncommon and that aromatase inhibitors may not have a place in the treatment of prostate carcinoma at this stage of the disease.

Relationship between platinum-DNA adducts in leukocytes of patients with advanced germ cell cancer and survival

Platinum-DNA adducts can be measured in peripheral blood leukocytes during platinum-based chemotherapy, and high adduct levels have been correlated with favorable clinical response in patients with germ cell cancer. Twenty-five patients with advanced germ cell cancer were treated with platinum-based chemotherapy regimens using the same dose and schedule of cisplatin. Platinum-DNA adducts were measured by atomic absorption spectrometry on the first and fifth days of the first cycle of cisplatin-based therapy. The patients were followed prospectively for 6-35 months (median, 26 months). Twenty-two patients had adduct levels measured 24 h after the first dose of cisplatin. There was no difference in the mean adduct levels of those who were alive and without progression of disease compared to those who were dead or progressing (P = 0.65). Twenty-three patients had day 5 adduct levels measured. The mean day 5 adduct level in the 15 patients who were alive and without progression was 62.133 fmol/microgram compared to 153.50 fmol/microgram in the patients who were dead or progressing (two-sided P = 0.02). Contrary to previous reports, these data indicate that high platinum-DNA adduct levels do not correlate with favorable outcome in patients with advanced germ cell cancer.

Preventing pneumococcal bacteremia in patients at risk. Results of a matched case-control study

BACKGROUND

Three randomized controlled trials of the effectiveness of pneumococcal vaccine in elderly and chronically ill adults in the United States have failed to show significant protective efficacy during 44,213 person-years of follow-up. Case-control studies have greater statistical power to detect significant prevention of rare diseases such as pneumococcal bacteremia, but they also have a greater susceptibility to bias, necessitating consistent results from multiple studies. Three case-control studies at two different universities have shown prevention of systemic infection, but another study found no benefit.

METHODS

Patients with pneumococcal bacteremia who were at least 2 years old and had chronic illness indicating the need for pneumococcal vaccine, or who were at least 65 years old were compared with matched control subjects for frequency of prior vaccination. Matching variables included date of admission, age, sex, race, type and duration of chronic illness serving as the major vaccine indication, number of vaccine indications and number of medical hospitalizations since licensure of the pneumococcal vaccine in 1978, and type of primary medical care.

RESULTS

Pneumococcal vaccination was documented in the records of six (7%) of 85 cases and 26 (17%) of 152 control subjects, suggesting 81% efficacy in conditional logistical regression analysis (95% confidence interval, 34% to 94%, P = .008).

CONCLUSIONS

Four case-control studies at three universities have now demonstrated significant protective efficacy of pneumococcal vaccine for preventing pneumococcal bacteremia. The development of antibiotic-resistant Streptococcus pneumoniae indicates an urgent need for an increased rate of vaccination among high-risk patients and for the development of more immunogenic conjugate vaccines that may enhance efficacy among elderly and immunocompromised patients as well as infants.

Enterobacter endocarditis

Endocarditis due to Enterobacter species is very rare. We recently cared for a patient who developed E. cloacae endocarditis following mitral valve replacement with a porcine heterograft, and was successfully treated with antibiotic therapy alone. A review of the literature disclosed an additional 17 well-described cases of enterobacter endocarditis. Two-thirds of the patients had underlying cardiac disease. The mitral valve was most frequently involved (10/16 cases) with 4 of the patients having concomitant aortic valve involvement. The overall mortality rate was 44.4%. Antibiotic therapy of enterobacter endocarditis should consist of the combination of a beta-lactam antibiotic and an aminoglycoside with careful monitoring of blood cultures to assure the adequacy of therapy. Resistance of enterobacter to previously susceptible antibiotics may occur during therapy due to induction of a chromosomally-mediated beta-lactamase, necessitating a change in antimicrobial therapy. Valvular surgery is indicated for patients failing medical management.

Predicting death in patients hospitalized for community-acquired pneumonia

OBJECTIVE

To validate a previously reported discriminant rule for predicting mortality in adult patients with primary community-acquired pneumonia and to determine which factors available at hospital admission predict a fatal outcome among such patients.

DESIGN

Historical cohort study.

SETTING

University hospital.

PATIENTS

Adults admitted to the hospital for community-acquired pneumonia.

MEASUREMENTS

Using stepwise logistic regression, we analyzed prognostic factors (data available at admission and recorded in the medical record) that showed a univariate association with mortality. The predictive values of three discriminant rules were measured to validate the results of a previous study.

MAIN RESULTS

Of 245 patients, 20 (8.2%) died. Of 42 prognostic factors identified in previous studies, 8 were associated with mortality, but only a respiratory rate of 30/min or more, a diastolic blood pressure of 60 mm Hg or less, and a blood urea nitrogen of more than 7 mmol/L remained predictive in the multivariate analysis. A discriminant rule composed of these three variables was 70% sensitive and 84% specific in predicting mortality, yielding an overall accuracy of 82%.

CONCLUSION

Tachypnea, diastolic hypotension, and an elevated blood urea nitrogen were independently associated with death from pneumonia in our study, confirming the value of a previously reported discriminant rule from the British Thoracic Society. This rule may be useful in triage decisions because it identifies high-risk patients who may benefit from special medical attention.

The seroprevalence of cytomegalovirus among Virginia State prisoners

A prevalence serosurvey of adult male prisoners entering the Virginia State Prison was conducted to evaluate the epidemiology of cytomegalovirus within this population. Four hundred and forty-five (97%) of 459 male inmates provided serum for analysis and 427 completed a detailed demographic questionnaire. Sera were tested for cytomegalovirus by passive latex agglutination and 64% were reactive. Multivariate discriminant analysis showed an independent association of seropositivity with age, non-white race, and a history of gonorrhea. There was no apparent contribution from admitted homosexual contact though this may have been under-reported. There was no correlation of seropositivity with intravenous drug use or with the length or number of prior incarcerations. Prisoners possess the same correlates for cytomegalovirus seropositivity as the general adult population; past imprisonment did not independently contribute to cytomegalovirus seropositivity.

Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection. A randomized, controlled trial

We performed a randomized controlled trial of an attachable subcutaneous cuff for the prevention of central vascular catheter-related infection among patients receiving intensive care. Catheters were placed percutaneously into new sites with or without a cuff and were dressed with polyantibiotic ointment containing polymyxin, neomycin, and bacitracin. Microbial colonization developed in 34.5% of 29 control and 7.7% of 26 cuffed catheters. Catheter-related bloodstream infection occurred with 13.8% of control vs 0% of cuffed catheters. The cuff was not associated with adverse effects. An unexpectedly large proportion (75%) of catheter infections were due to Candida albicans. This may have been due, in part, to the use of polyantibiotic ointment, as suggested by a pooled analysis of previous trials that demonstrated increased Candida colonization of catheters with the ointment, which is not fungicidal. These data suggest that the cuff can reduce the incidence of catheter-related infection among high-risk patients receiving catheter site care with an antibacterial ointment.