Faculty Bibliography

Background: Systemic anticancer therapy (SACT) options for aNSCLC continue to increase each year with approvals of more effective therapies that improve long-term outcomes, as seen with immunotherapies (IO). We aimed to examine real-world trends in SACT distribution and sequence from first- to second-line (1L-2L) for EGFR/ALK-negative nonsquamous aNSCLC from 2011-2018 at US community oncology practices.
Method(s): This study used the nationwide Flatiron Health de-identified, EHR-derived database (cutoff: 31Jan2019). Eligible patients were adults with aNSCLC, nonsquamous histology with known EGFR/ALK-negative status, who initiated 1L SACT from Jan2011-Jun2018. SACT regimens were assigned to mutually exclusive classes in hierarchical order, from highest to lowest: (1) PD1/PD-L1 inhibitor (anti-PD1/L1)-based, (2) EGFR/ALK TKI-based, (3) platinum-based chemotherapy (PBC) combination with vascular endothelial growth factor inhibitor (PBC+VEGF), (4) PBC only, (5) single agent chemotherapy, (6) others. 2L regimens were examined for patients initiating 1L SACT only through 2017 to enable sufficient follow-up. Results were stratified by year and by pre-IO/post-IO years of 1L initiation, defined as 2011-2014/2015-2018, respectively, based on earliest IO approval for 2L therapy in Mar2015.
Result(s): For 1L, in the pre-IO period, most patients were prescribed PBC (53%) or PBC+VEGF (30%), and post-IO, most were prescribed PBC (43%), anti-PD1/L1 (25%), or PBC+VEGF (23%). Among patients prescribed 1L therapy, the percentages who received 2L therapy were 50%-62% post-1L PBC; 56%-62% post-1L PBC+VEGF; and 28%-38% post-1L anti-PD1/L1 (table). A substantial percentage (25%) of those initiating 1L anti-PD1/L1 in 2017 were still on therapy at cutoff. [Figure presented]
Conclusion(s): Changing trends in real-world prescribing of 1L-2L SACT for EGFR/ALK-negative nonsquamous aNSCLC from Jan2011-Jun2018 include decreasing use of PBC in 1L and 2L, decreasing use of PBC+VEGF in 1L, and, increasing use of PD1/PD-L1 inhibitors in 1L and 2L. Slightly more than one-half of patients with nonsquamous aNSCLC prescribed 1L PBC are subsequently treated with 2L therapy. Keywords: nonsquamous NSCLC, real-world therapy trends, Immunotherapy
Copyright

Background: The SACT options for aNSCLC continue to increase each year with approvals of more effective therapies that improve long-term outcomes, as seen with immunotherapies (IO). Our aim was to examine real-world trends in SACT distribution and sequence from first- to second-line (1L-2L) for squamous aNSCLC from 2011-2018 at US community oncology practices.
Method(s): This study used the nationwide Flatiron Health de-identified, EHR-derived database (cutoff: 31Jan2019). Eligible patients were adults who initiated 1L SACT from Jan2011-Jun2018 for aNSCLC with squamous histology, excluding patients with known EGFR/ALK-positive tumors. Descriptive analyses included patients receiving >=1 SACT dose, assigning all SACT regimens to mutually exclusive classes in hierarchical order (combination regimens assigned by highest component), from highest to lowest: (1) PD1/PD-L1 inhibitor (anti-PD1/L1)-based, (2) EGFR/ALK TKI-based, (3) platinum-based chemotherapy combination with vascular endothelial growth factor inhibitor (PBC+VEGF), (4) PBC only, (5) single agent chemotherapy, (6) others. The 2L regimens were examined for patients with 1L SACT initiation only through 2017 to enable sufficient follow-up. Results were stratified by years and by pre-IO and post-IO years of 1L initiation, defined as 2011-2014 and 2015-2018, respectively, based on the earliest IO approval for 2L therapy in Mar2015.
Result(s): For 1L therapy, in the pre-IO period, most patients were prescribed PBC (80%), and post-IO, most patients were prescribed PBC (68%) or anti-PD1/L1 (21%). Among patients prescribed 1L therapy, the percentages who received 2L therapy were 44%-53% following 1L PBC and 25%-37% following 1L anti-PD1/L1, with 19% of 2017 1L anti-PD1/L1 starts still on 1L therapy (table). [Figure presented]
Conclusion(s): PBC remain the most common 1L SACT prescribed through mid-2018 for patients with squamous aNSCLC at US community oncology practices. Prescribing of PD1/PD-L1 inhibitors as 1L has gradually increased since regulatory approval starting in 2015. Approximately one-half of patients with squamous aNSCLC prescribed 1L PBC are treated with 2L therapy. Keywords: squamous NSCLC, real-world therapy trends, Immunotherapy
Copyright

Background: Information from real-world clinical settings remains limited regarding outcomes of pembrolizumab therapy for advanced NSCLC. Our aim was to examine real-world time on treatment (rwToT) and overall survival (OS) for patients prescribed pembrolizumab monotherapy for previously treated, PD-L1-expressing advanced NSCLC, thus clinically similar to patients in the KEYNOTE-10 (KN010) trial.
Method(s): This retrospective study used Flatiron Health's nationally representative EHR-derived database to identify adult patients with histologically confirmed advanced NSCLC and PD-L1 tumor proportion score (TPS) >=1% previously treated with platinum-containing chemotherapy (and appropriate TKI if nonsquamous NSCLC with EGFR/ALK aberration). Eligible patients initiated pembrolizumab monotherapy from January 1, 2016, to May 31, 2018; those with <6 months of follow-up were excluded. Kaplan-Meier (KM) rwToT and OS were calculated.
Result(s): Median follow-up was 15.6 months (range 6.0-32.8 months). Of 281 eligible patients (56% male), median age was 68 years; 36% had squamous NSCLC; 10% brain metastases; and 57%, 18%, and 25% ECOG performance status 0-1, >=2, and unknown, respectively. Baseline characteristics were similar across PD-L1 TPS distributions. The table summarizes rwTOT and OS by PD-L1 TPS categories. [Figure presented]
Conclusion(s): Real-world patients treated with pembrolizumab monotherapy after platinum-containing chemotherapy for PD-L1-expressing advanced NSCLC experienced rwToT and OS benefits similar to findings in a clinical trial setting, validating KN010 findings and approved indication in second line. Keywords: Pembrolizumab, overall survival, real-world time on treatment
Copyright

Background: The SACT options for aNSCLC continue to increase each year with approvals of more effective therapies that improve long-term outcomes, such as immunotherapies (IO). Our aim was to examine trends in SACT rates from 2011-2018 for patients with aNSCLC and no known EGFR/ALK aberrations at US community oncology practices.
Method(s): We used the nationwide Flatiron Health EHR-derived database (data cutoff: 31Jan2019), which incorporates oncologist-defined, rule-based lines of therapy. Adults with aNSCLC diagnosis date from Jan2011-Jun2018, inclusive, with recorded EHR activity <=90 days after diagnosis, were eligible. Patients with known EGFR/ALK-positive tumors, or nonsquamous histology (NSQ) with unknown or untested EGFR/ALK status, were excluded. We summarized no-treatment and first-line (1L) SACT rates as a proportion of aNSCLC diagnosed from 2011-2018, while limiting 2L and 3L SACT rates to aNSCLC diagnoses from 2011-2016 in order to have sufficient follow-up. Results were stratified by NSQ and squamous (SQ) histology, as well as by pre-IO and post-IO years of aNSCLC diagnosis, defined broadly as 2011-2014 and 2015+, respectively, based on the earliest IO approval for 2L therapy in Mar2015.
Result(s): The figure depicts 1L, 2L, and 3L SACT rates by year of aNSCLC diagnosis for EGFR/ALK-negative NSQ and for SQ. For NSQ, the pre-IO and post-IO no-treatment rates were 28% (2286/8246) and 22% (2520/11,639), respectively. For SQ, the pre-IO and post-IO no-treatment rates were 34% (1781/5200) and 26% (1549/6040). [Figure presented]
Conclusion(s): For both EGFR/ALK-negative NSQ and SQ aNSCLC, 1L SACT rates are trending upward, with no-treatment rates showing substantial drops in the post-IO (versus pre-IO) period. The 2L and 3L SACT rates are variable for both NSQ and SQ; and SACT rates for NSQ tend to be substantially greater than for SQ across all lines of therapy and years of diagnosis. Keywords: real-world treatment rates, Immunotherapy, advanced NSCLC
Copyright

BACKGROUND: Pediatric HIV has evolved from a pre-antiretroviral era (pre-1989 or Pre-ART) to an antiretroviral (ART) era (1989-1996) & to a highly active antiretroviral therapy (HAART) era (post-1996). As we have passed the 3rd decade following these individuals, we thought it useful to review clinical, laboratory, & social outcomes. METHODS: A retrospective, cross-sectional study of 399 children infected perinatally. They were divided into Pre-ART, ART & HAART groups. A Kaplan-Meier plot was constructed. 179 have been lost to follow-up at an average of 7.6 (0.3-27.6) years. RESULTS: Approximately 40, 80 and 90% of individuals in the Pre-ART, ART and HAART groups have long-term survival. 121 died at an average of 5.1 (0-26.1) years. Pre-ART, ART, & HAART groups had mean most recent CD4% values (+/-SEM) of 16.74(1.09), 22.97(0.96) & 33.07 (2.09), respectively (p<0.001). Pre-ART RNA is limited in that era and present if they survived to another era. In this group the median RNA values in those who died (311,300, n=16) was greater than in survivors (19,402, n=45). 43 % of the individuals in the ART group 77% of individuals in the HAART group had most recent HIV RNA<400 copies/ml.18 individuals >18 years of age have only a grade school or no education. 55 have graduated high school or received an equivalency diploma. 23 more have completed college. Nadir & recent CD4% of those who did & did not complete high school was equivalent to college graduates. 16 survivors (1/2 male) have had 18 uninfected children. CONCLUSIONS: This first long term follow-up study demonstrates remarkable survival and social skills of our patients.

The prevalence of negative life events (NLE) and daily hassles, and their direct and moderated associations with depression, were examined among HIV-infected adolescents. Specifically, we examined whether the negative association with depression of NLE, daily hassles, and/or passive coping were moderated by social support or active coping strategies. Demographic characteristics, depression, coping, social support, NLE, and daily hassles were collected at baseline as part of the Adolescent Impact intervention via face-to-face and computer-assisted interviews. Of 166 HIV-infected adolescents, 53% were female, 72.9% black, 59.6% with perinatally acquired HIV (PIY), the most commonly reported NLE were death in family (81%), violence exposure (68%), school relocation (67%), and hospitalization (61%); and for daily hassles "not having enough money (65%)". Behaviorally infected youth (B

Violence exposure among HIV-infected adolescents was estimated using Poisson regression and adjusted event rate ratios (ERR). Of 166 urban adolescents (99 perinatally infected youth [PIY]; 67 behaviorally infected youth [BIY]) 52.5% (n = 85) experienced violence; 79% (n = 131) witnessed violence. Sexual violence was experienced by 18% (6 PIY, 24 BIY) before age 13 years and by 15% (6 PIY, 19 BIY) during adolescence. BIY were significantly more likely than PIY to have experienced and witnessed violence. Controlling for transmission, ever-bartered sex (ERR = 1.92, CI [1.31 to 2.81], p = .009) and family disruptions (ERR = 1.19, CI [1.03 to 1.39], p = .022) were associated with violence victimization. Family disruptions (ERR = 1.17, CI [1.05 to 1.30], p = .004), female gender (ERR = 1.32, CI [1.05 to 1.66], p = .017), and heterosexual orientation (ERR = 1.48, CI = [1.11 to 1.97], p = .006) were associated with witnessing violence.

We examined the prevalence and predictors of drug use among a diverse group of adolescents living with HIV infection acquired perinatally or through sexual risk behaviors ("behaviorally acquired"). Adolescents ages 13-21 (n = 166) who were receiving care at one of five pediatric/adolescent HIV clinics in three US cities (Baltimore MD, Washington DC, and New York NY) and were enrolled in a behavioral intervention were interviewed at baseline regarding lifetime drug use experiences and depression symptoms. A majority of study participants reported using alcohol (57.2%) and marijuana (51.2%); 48.8% reported tobacco/cigarette use. The mean age of onset of use for each type of drug was 14 years or younger. A larger proportion of participants with behaviorally acquired HIV than adolescents with perinatally acquired HIV reported lifetime use of alcohol (76.1 vs. 44.4%), marijuana (73.1 vs. 36.4%), tobacco (70.2 vs. 34.3%), and club drugs (22.4 vs. 3%) (all p < 0.001).

Disease knowledge is an important component of medication self-management. We examined HIV knowledge and understanding among 166 HIV-infected 13- to 21-year-olds (53% female, 72.9% African American, 59.6% perinatally infected). Behaviorally infected youth outperformed perinatally infected youth on HIV transmission knowledge and were more likely to report provider discussions about getting condoms and how drugs impair sexual decision making. While youth accurately answered most (78%) true/false questions, many could not define resistance (71.0%), viral load (59.0%), or CD4 (43.4%). Only 29.5% reported knowing either their CD4 count or viral load. Provider discussions about biomarkers likely contribute to greater HIV knowledge and understanding, especially for youth with lower CD4 counts, perhaps because illness triggers extended provider discussions.