Safety and efficacy of catheter-directed therapy versus anticoagulation alone in a higher-risk acute pulmonary embolism population

There is little data comparing safety and efficacy outcomes in patients with pulmonary embolism (PE) receiving catheter directed therapies (CDT) compared to a similar-risk cohort of PE patients receiving anticoagulation alone. 1094 patients with acute PE were studied. CDT and conservatively-managed patients were compared using propensity score matching to assess safety outcomes, which included bleeding and acute kidney injury at 2 and 7 days after PE diagnosis. Efficacy outcomes included change in vital signs over 72 h and in-hospital mortality. PE patients with RV strain who underwent CDT (n = 76) had more bleeding at 2 days (additional 1.04 g/dL loss, 95% CI − 1.48 to − 0.60, p < 0.001) and 7 days (additional 1.36 g/dL loss, 95% CI − 1.88 to − 0.84, p < 0.001) compared to those receiving anticoagulation alone (n = 303). There was a significant increase in creatinine at 2 days (additional 0.22 mg/dL elevation, 95% CI 0.02 to 0.42, p = 0.03), but not at 7 days (additional 0.12 mg/dL elevation, 95% CI − 0.11 to 0.35, p = 0.30). In-hospital mortality for patients receiving CDT versus anticoagulation alone was similar (OR 1.21, 95% CI 0.53 to 2.77; p = 0.65). In patients with baseline abnormal vital signs who received CDT versus anticoagulation alone, heart rate, respiratory rate and oxygen requirement improved significantly faster and to levels closer to normal (p ≤ 0.001). CDT was associated with a small but increased risk of bleeding, but no significant worsening of renal function. CDT may be associated with more rapid improvements in heart rate, respiratory rate, and oxygen requirement.

Patient out-of-pocket savings with lower-dose trastuzumab

e18828

Background: Trastuzumab-related costs of human epidermal growth factor receptor 2 (HER2) positive breast cancer treatment are significant, approaching $70,000 per patient in the adjuvant setting alone. Recently developed population pharmacokinetic models demonstrate that therapeutic levels of trastuzumab are maintained despite decreased maintenance dose or dose frequency, potentially improving trastuzumab’s value proposition. In this observational study, we evaluate the potential patient- and health system-level financial implications of low-dose trastuzumab. Methods: We identified 196 individuals who received treatment for HER2-positive breast cancer with trastuzumab or its biosimilars at a single urban comprehensive cancer center between January 1, 2017 and October 31, 2019. We estimate, using publicly available information, patient-experienced out-of-pocket costs using deterministic modeling under different conditions corresponding to changes in dosing practices, geographic site of trastuzumab administration, patient employment status, intent of treatment, and billing procedures, as well as Medicare cost-savings estimates. Results: Both reducing the maintenance dose of trastuzumab from the FDA-labeled dose of 6 mg/kg to 4 mg/kg Q3W and reducing maintenance dose frequency of trastuzumab from 6 mg/kg Q3W to Q4W were each associated with significant patient savings on drug- and administration-related out-of-pocket costs. Depending on billing practice and clinical scenario, patient savings as a percentage of total costs were 9-20% (neoadjuvant), 17-22% (adjuvant), and 14-19% (metastatic) with 6 mg/kg Q4W dosing. Similarly, 4 mg/kg Q3W dosing had savings of 15-25% (neoadjuvant), 19-28% (adjuvant), and 13-28% (metastatic). Utilization of off-label dosing regimens could result in $2.69 billion USD in annual Medicare spending on trastuzumab, an estimated savings of $670-810 million USD. Conclusions: Meaningful patient-level and system-level financial savings in the treatment of HER2-positive breast cancer can be achieved using pharmacokinetically near-equivalent low-dose or reduced frequency maintenance trastuzumab dosing. Given the significant potential savings, further prospective studies aimed at reducing dose or frequency of monoclonal antibodies in the treatment of cancer are warranted.

Spatial and Temporal Heterogeneity of PD-L1 Expression and Tumor Mutational Burden in Gastroesophageal Adenocarcinoma at Baseline Diagnosis and after Chemotherapy

PURPOSE

Intrapatient heterogeneity of programmed death ligand 1 (PD-L1) expression and tumor mutational burden (TMB) in gastroesophageal adenocarcinoma (GEA) could influence their roles as predictive biomarkers for response to immune checkpoint inhibitors (ICI). In this retrospective analysis, we evaluated the spatiotemporal heterogeneity and prognostic relevance of PD-L1 expression and TMB in GEA.

EXPERIMENTAL DESIGN

A cohort of 211 patients with stage II-IV GEA was retrospectively reviewed for a total of 407 tumor samples with PD-L1 expression data and 319 tumor samples with TMB data. PD-L1 status was defined as positive if combined positive score (CPS) ≥1 using the 22C3 pharmDx assay. TMB levels were categorized as low, intermediate, or high (≤5, 5-15, or >15 mutations/Mb), or using a single threshold (<10 or ≥10 mutation/Mb), determined by next-generation sequencing using a targeted gene panel.

RESULTS

Of 407 tumors, 56% were PD-L1 negative and 44% PD-L1 positive. Of 319 tumors, 50% were TMB-low, 45% TMB-intermediate, and 5% TMB-high; 86% had <10 and 14% ≥10 mutations/Mb. TMB level was significantly associated with MSI-status. PD-L1 expression and TMB exhibited marked spatial heterogeneity between baseline primary and metastatic tumors (61% and 69% concordance), and temporal heterogeneity between tumors before and after chemotherapy (57%-63% and 73%-75% concordance). PD-L1 expression and TMB were not significantly associated with overall survival.

CONCLUSIONS

PD-L1 expression and TMB exhibit marked spatial and temporal heterogeneity in GEA. This heterogeneity should be considered when obtaining tumor samples for molecular testing and when deciding whether ICI therapy is appropriate.See related commentary by Klempner et al., p. 6401.

Abstract 5451: Longitudinal proteomic assessment of patient with metastatic apocrine adenocarcinoma reveals evolutionary selection for androgen-receptor-dependence and therapeutic response

Combined use of new molecular diagnostic and imaging methods could improve care of individual patients with rare cancers.

A 57-year-old developed apocrine adenocarcinoma of the left axilla with regional lymph node metastasis detected at initial surgery. After adjuvant regional radiotherapy, metastases developed and over the subsequent 7 years he received serial systemic therapy with 11 different regimens. Expression of 128 proteins/phosphoproteins was evaluated in 8 tissue samples from diagnosis, progressive adenopathy at 45 months, brain metastases at 60 months, and abdominal metastases at 63 months by CLIA-standardized reverse phase phosphoprotein arrays (RPPA). Tumor burden was estimated by quantifying the volume of individual lesions with semi-automated segmentation algorithms on serially collected CT images.

Next-generation sequencing (NGS) revealed a PIK3CA K111T mutation, prompting treatment with sirolimus 28 months after diagnosis. Seventeen months later, new neoplastic adenopathy emerged. NGS revealed a new PIK3CA M610V mutation. Semiquantitative scoring of RPPA demonstrated relative increase in AKT (from 0.72 to 0.9) and mTOR (from 0.55 to 0.8) pathway activation. The relative estimated tumor burden (RETB) was 21,145 mm3. Over 6 months of docetaxel therapy the RETB decreased to 12,077 mm3. The patient subsequently received pembrolizumab with increased RETB (71,308 mm3). Gemcitabine appeared to stabilize the RETB (73,168 mm3), but brain metastases emerged. The patient underwent craniotomy, began paclitaxel therapy, and then RETB declined to 35,623 mm3. But new symptomatic abdominal metastases prompted surgical resection. The patient then received the PI3K inhibitor taselisib; however, after an initial decrease, RETB rose to 31,208 mm3.

At this point, the patient began bicalutamide therapy. NGS of the primary mass and the initial metastatic adenopathy revealed no amplification of the androgen receptor (AR). But RPPA demonstrated inversion of the ratio of measured phosphorylated AR S81 to AR S650, suggesting primarily cytoplasmic AR in tumor through 45 months but nuclear AR in samples resected at 60 and 63 months. Immunohistochemistry revealed 3+ nuclear AR expression in the latter metastatic tissue. After 9 months of bicalutamide, enzalutamide and leuprolide were administered for 7 months (final RETB = 1,555 mm3). Treatment continued beyond the study period.

Tumor burden assessment by new CT imaging measurement methods, combined with RPPA could improve adaptive, responsive, therapy for patients with rare tumors. We identified the phenotypic evolution of androgen-driven growth of apocrine adenocarcinoma after cytotoxic and PI3K-mTOR inhibitor therapy. Protein-based diagnostics revealed an effective treatment strategy in late metastatic disease that was not indicated by NGS testing.

Citation Format: Erik Dvergsten, Anthony Serritella, Danielle Kimble, Malcom McIver, Sanja Karovic, Vasiliki Thomeas-McEwing, Hao Yang, Manish R. Sharma, Thomas P. Conrads, Mariaelena Pierobon, Peter Pytel, Binsheng Zhao, Lawrence H. Schwartz, Emanuel F. Petricoin, Russell Szmulewitz, Michael L. Maitland. Longitudinal proteomic assessment of patient with metastatic apocrine adenocarcinoma reveals evolutionary selection for androgen-receptor-dependence and therapeutic response [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5451.

Assessment of FcgRIIIA single nucleotide polymorphisms (SNPs) on the efficacy of IgG1 monoclonal antibodies (mAbs) in PANGEA study patients (pts) with advanced gastroesophageal adenocarcinoma (aGEA)

4526

Background: Targeted therapies (Ttx) have had limited efficacy in aGEA. The phase IIa PANGEA trial assessed the outcomes of pts treated with IgG1 mAbs targeting receptor tyrosine kinases (RTKs) or PD-1 based on predefined molecular groups. The fragment C (Fc) portion of mAbs binds to IgG receptors (FcgR) of immunologic effector cells such as natural killer (NK) cells, leading to antibody-dependent cell-mediated cytotoxicity (ADCC). The FcgR subclass, FcgRIIIA, has genetic variants with different Fc binding affinities. A single nucleotide polymorphism (SNP) in FcgRIIIA substitutes phenylalanine (F) with valine (V) at amino acid position 158, enhancing FcgR’s affinity for the IgG1 Fc domain. Pts with V/V or V/F FcgRIIIA allotypes have enhanced NK cell binding affinity compared to the homozygous F/F allotype. We evaluated the association of FcgRIIIA SNPs on Ttx outcomes amongst PANGEA pts and another cohort of aGEA pts treated with IgG1 mAbs. Methods: Whole-blood samples were identified from aGEA pts (N = 104), including 70/80 available PANGEA pts, who were treated with an IgG1 mAb (trastuzumab 24, anti-EGFR 21, anti-PD1 30, ramucirumab 48) in at least 1 Ttx line. After lymphocyte DNA extraction, FcgRIIIA genotyping was performed. The Cox proportional hazard model and log-rank tests, adjusted for pt age, were used to assess for an association of genotype with overall survival (OS). Results: Of 104 genotyped pts, the F/F, F/V & V/V genotypes were observed at a frequency of 32%, 51% and 17% respectively. There was no significant difference in median OS (mOS) between the F/F, F/V or V/V or comparing F/F vs V/F+V/V overall, nor in the PANGEA-only cohort. A trend of increased mOS was seen in 20 non-PANGEA pts harboring F/V or V/V compared to 14 F/F pts (mOS 43.4 vs 23.1 months, HR 0.41 [0.15-1.14] p = 0.09). However, 3-year OS rates trended higher in V/F+V/V pts (22%, 16/71) compared to F/F pts (7%, 2/33) (p = 0.09). At 3 years, 50% of V/V+V/F non-PANGEA pts were alive versus 13% of F/F pts (p = 0.04), while 13% of V/V+V/F PANGEA pts were alive versus 0% of F/F pts (p = 0.32). Conclusions: Amongst pts receiving IgG1 mAbs, the high affinity V FcgRIIIA SNP enriched for a subgroup of ‘extreme responders’ alive 3 years from diagnosis. Multivariate analyses accounting for baseline characteristics in a larger number of pts are ongoing to further elucidate the role of FcgRIIIA SNPs as predictive biomarkers. These findings may have implications on IgG1 mAb ADCC optimization.

Evaluation of spatiotemporal heterogeneity of PDL1 expression in gastroesophageal adenocarcinoma (GEA) at baseline diagnosis and after chemotherapy

4528

Background: PDL1 expression is a predictive marker for response to anti-PD1/PDL1 agents (IO) for GEA. As a prognostic biomarker, data are conflicting. Molecular heterogeneity of various biomarkers for GEA has been established. To characterize heterogeneity of PDL1 expression and its clinical relevance, we compared PDL1 expression in primary (10) and metastatic (met) tumors of newly diagnosed stage IV advanced GEA (aGEA), and before and after chemotherapy treatment (tx) for stage II–IV GEA. We assessed the prognostic relevance of PDL1 expression in aGEA. Methods: We retrospectively reviewed a cohort of 130 patients (pts) diagnosed with GEA in 2013–2019, with a total of 328 tumor samples with PDL1 expression data. PDL1 was defined as positive if combined positive score (CPS) was ≥1 using the 22C3 pharmDx assay. Analysis was performed by McNemar’s test for paired PDL1 and univariate Cox proportional-hazards model for overall survival (OS). Results: Of 328 tumors, 45% were PDL1+ and 55% PDL1-. CPS ranged 0–100 (median 1, IQR 0–5), and CPS was ≥10 for 19% of tumors. Concordance between PDL1 status of paired baseline 10 and baseline met tumors was 63% (32/51) (Table). Of 31 PDL1+ baseline 10 tumors, 52% (16/31) had PDL1- paired baseline met tumors, while of 20 PDL1- baseline 10 tumors, only 15% (3/20) had PDL1+ paired baseline met tumors. Only 35% (18/51) of met tumors were PDL1+, compared to 61% (31/51) PDL1+ 10 tumors ( p< 0.003). Post-tx tumors exhibited 62% (46/74) concordance of PDL1 status compared to pre-tx 10 tumors. Of 43 PDL1+ baseline tumors, 35% (15/43) were PDL1- post-tx; of 31 PDL1- baseline tumors, 42% (13/31) were PDL1+ post-tx ( p= 0.71). In pts with aGEA at diagnosis, OS did not significantly differ depending on baseline 10 tumor PDL1 status (median OS of 17.9 [95% CI 14.6–26.5] months for PDL1- and 16.7 [12.0–26.3] months for PDL1+; p= 0.6), nor depending on baseline met PDL1 status. Conclusions: PDL1 expression demonstrated notable baseline discordance between 10 and met tumors, particularly directional from PDL1+ 10 tumor to PDL1- met. Discordance before and after chemotherapy was also observed, but with similar proportions of PDL1+ pre-tx and post-tx tumors. These findings may have predictive IO therapeutic implications if confirmed in larger independent analyses. [Table: see text]

Personalized antibodies for gastroesophageal adenocarcinoma (PANGEA): Secondary and final primary efficacy analyses

4561

Background: Targeted therapies (tx) have had limited benefit in advanced (aGEA) due to baseline spatial (primary vs metastatic tumor PT/MT) & temporal molecular heterogeneity (BMH/TMH). We previously reported PANGEA methods & results: 35% BMH rate & 10 OS results achieving 1yr OS of 66% & mOS of 16.4 months (m) using the personalized tx strategy (Catenacci et al. GI ASCO 2020 Abstr356). Here we will report the TMH rates at progressive disease points (PD1 & PD2), ORR/PFS/DCR in each of 3 tx lines, time to strategy failure (TTF), & updated OS/safety. Methods: PANGEA enrolled newly diagnosed aGEA pts who then received up to 3 cytotoxic (cx) tx lines (L). Baseline tissue biomarker profiling (BP) was mandated on PT/MT & PD1/PD2, & ctDNA analysis throughout. After initiating 1L cx & upon learning MT BP results, antibody (AN) was added by a predefined prioritized tx algorithm incorporating tissue & blood BP (Table). At PD1, pts went to 2L cx + initial AN. Upon results of PD1 BP, pts changed AN only if BP evolved per tx algorithm. The same was done at PD2. The 10 endpoint was 1yr OS; enrolling 68 pts provided 80% power to detect a 63% 1yr OS compared to historical 50% 1yr OS (HR 0.67), using a 1-sided test (0.10 alpha). Results: 80 pts were enrolled, & 68 tx’d per protocol. At data cut-off 2/1/20, 15 pts were still on trial with only 2 of these pts on tx <12m (8 pts in 1L, 5 in 2L, 2 in 3L). All 68 pts had at least 1 dose of 1L tx, 87% 2L tx, & 36% 3L tx. AN assigned by the tx algorithm at 1L, OS, TTF, & ORR1/PFS1/DCR1 of 1L tx are shown in Table; 2L & 3L ORR/DCR outcomes will be shown. The 3yr & 4yr OS rates were 12% & 8%. TMH leading to molecular subgroup change by tx algorithm was 51% after 1L & 36% after 2L; details & results by subgroup will be provided. Any grade >3 non-heme tox thru all 3 tx lines was seen in 25% of pts. Conclusions: PANGEA showed superior 10 & 20 endpoint efficacy, even when excluding HER2- pts, compared to historical outcomes. Clinical trial information: NCT02213289 . [Table: see text]

Evaluation of spatiotemporal heterogeneity of tumor mutational burden (TMB) in gastroesophageal adenocarcinoma (GEA) at baseline diagnosis and after chemotherapy

4546

Background: Tumor mutational burden (TMB) may be a predictive marker for response to anti-PD1/PDL1 agents (IO). Molecular heterogeneity of various biomarkers for GEA has been established. To characterize heterogeneity of TMB and its clinical relevance, we compared TMB in primary (10) & metastatic (met) tumors at baseline newly diagnosed stage IV advanced GEA (aGEA), and before & after chemotherapy treatment (tx) for stage II–IV GEA. We assessed the prognostic relevance of TMB in aGEA. Methods: We retrospectively reviewed a cohort of 127 patients (pts) diagnosed with GEA in 2012–2019, for a total of 280 tumor samples with TMB data. TMB level was defined as low (≤5/Mb), intermediate (int) ( > 5/Mb, ≤15/Mb), or high (hi) (≥15/Mb), determined by Foundation One. Analysis was performed by Fisher’s exact test for PDL1/TMB, McNemar’s test for paired TMB, and univariate Cox proportional-hazards model for overall survival (OS). Results: Of 280 tumors, 50% (140/280) had low TMB, 45% (125/280) int TMB, & 5% (15/280) hi TMB. TMB ranged 0–58.6/Mb (median 5.3/Mb). Of tumors with hi TMB, 53% (8/15) were MSI-Hi, while of MSI-Hi tumors, 100% (8/8) were TMB hi. TMB level did not correlate with PDL1 status ( p= 0.83). Concordance between TMB levels of paired baseline 10 and baseline met tumors was 66% (29/44) (Table). TMB level was lower in the met than in the 10 in 23% (10/44) of cases, and higher in the met in 11% (5/44). Of 4 TMB hi baseline 10 tumors, 2 were not TMB hi in the met; of 40 TMB low/int baseline 10 tumors, 0 were TMB hi in the met ( p= 0.16). Post-tx tumors exhibited 71% (42/59) concordance of TMB levels compared to pre-tx 10 tumors. Of 2 TMB hi baseline tumors, 1 was not TMB hi in the post-tx tumor; of 57 TMB low/int baseline tumors, 0 were TMB hi in the post-tx tumor ( p= 0.32). In pts with aGEA at diagnosis, OS did not significantly differ depending on baseline 10 tumor TMB level (median OS of 21.4 [95% CI 15.4–27.9] months for TMB low, 14.6 [10.9–23.5] months for TMB int, and 9.6 [3.9–NA] for TMB hi; p= 0.3), nor depending on baseline met TMB level. Conclusions: Notable baseline spatial discordance of TMB was observed, particularly TMB hi 10 to low/int met. Discordance was also observed before & after tx, without significant increase towards TMB hi temporally. Spatiotemporal heterogeneity may impact the role of TMB as a predictive biomarker & warrants further study. [Table: see text]

Phase I/II trial of enzalutamide (Enz) plus mifepristone (Mif) for metastatic castration-resistant prostate cancer (mCRPC)

91

Background: Resistance to androgen receptor (AR) targeted therapies is common in mCRPC. Glucocorticoid receptor (GR) expression increases with AR inhibition in patients (pts) and blockade of GR signaling inhibits CRPC growth in preclinical models when combined with AR blockade. We thus conducted a phase I/II open label trial of Enz combined with Mif, a GR, AR, and progesterone receptor antagonist for pts with mCRPC to assess the feasibility and impact on disease progression with dual AR/GR antagonism. Methods: The phase I dose escalation portion assessed the safety of the two-drug combination and a recommended phase II dose (R2PD) was determined based on safety, pharmacokinetic and endocrine assessments. In the phase II portion, patients (pts) received 12 weeks of Enz (160mg/day) followed by randomization to Enz alone or Enz plus Mif with PSA-progression free survival (PFS) as the primary endpoint. 42 pts were to randomize to each arm to provide 80% power to detect a hazard ratio of 0.6, with a one-sided alpha of 0.1; there was a planned interim futility analysis after 50% of progression events. Results: 106 pts (18 phase I/88 phase II) were enrolled. Pts had a median age of 70 (range 53-89) and baseline PSA of 12.8 (range 0.1-755). 34% of pts received prior docetaxel. The RP2D was 120mg/day Enz and 300mg/day Mif. In phase II, 33 patients were randomized to each arm, with well-balanced baseline demographics. 22 pts were not randomized (15 due to disease progression, 2 due to toxicity, and 5 due to the interim study analysis). The interim analysis showed no difference between arms in PSA-PFS (hazard ratio = 1.34, p=0.395), 12-month PSA-PFS of 31% in both arms, and per-protocol, the trial was stopped. Toxicities were similar in the arms, e.g. fatigue (12% vs. 14%), hot flashes (6% vs. 5%), and pain (4% vs. 4%). Conclusions: The addition of Mif to Enz following a 12-week Enz lead-in did not delay time to PSA progression. Further analyses of secondary endpoints, including translational biomarkers such as hormone levels, GR/AR-v7 expression in circulating tumor cells and cell free DNA analyses are ongoing. Clinical trial information: NCT02012296.