ASCO 2016 - Three anti-PD1 reports

A little surprisingly...not much has been said about anti-PD1 in the ASCO abstracts this year.  Perhaps more will be presented at the meeting, or perhaps it is all old news at the moment.  I certainly covered these reports already.  But, here are three anti-PD1 stories out of ASCO this year.

1.  When I started my nivo trial in 2010....you weren't even allowed to take anti-PD1 if you had already taken ipi....then you could, but you couldn't have had an immune related side effect.  All this was based on the thought that if you reacted badly to one...you were bound to have a worse reaction to the other...and you certainly couldn't take it if you already had an autoimmune condition.  Well, wonderful brave ratties have proven that that is NOT TRUE!!!  Here is a post covering some of that:  Anti-PD1 success in melanoma despite preexisting autoimmune disease with links to other reports about previous side effects, yet tolerating treatment and gleaning a response

Here's what came out of ASCO:

Anti-PD-1 therapy in patients with advanced melanoma and preexisting autoimmune disorders (AD) or major toxicity with ipilimumab (IPI).  ASCO 2016. #9515.  J Clin Oncol 2016.  Menzies, Johnson, Ramanujam, et al.

Background: Anti-PD1-antibodies (PD1) have activity in many cancers, and are standard care for melanoma, lung and renal cancer. All trials excluded patients (pts) with significant preexisting AD or major immune-related adverse events (irAEs) with IPI. We sought to explore the safety and efficacy of PD1 in such pts. Methods: Pts with advanced melanoma and preexisting AD and/or major irAEs with prior IPI (requiring systemic immunosuppression, [IS]) treated with PD1 were retrospectively identified. Data regarding AD, IPI and PD1 treatments, toxicity and outcome were examined. Results: 119 pts were included, 95 with prior IPI. 109 received pembrolizumab, 10 nivolumab. 86 (72%) had greater or = to, 3 months follow-up, median (med) 4.6 mo, with med PFS 6.8 mo. 31 (26%) had died. Of 52 pts with preexisting AD, 15 (29%) had active symptoms at PD1 start and 16 (31%) were on IS. The ORR was 33%. 20 (38%) flared with PD1 after a med 1.3 mo, including 7/13 with rheumatoid arthritis, 3/3 with polymyalgia rheumatica, 2/2 with Sjogren’s syndrome, 1/2 with scleroderma, 2/2 with immune thrombocytopaenic purpura, 3/8 with psoriasis, 1/4 with Graves’ disease, 0/6 with gastrointestinal (GI) (including 3 Crohn’s disease) and 0/5 with neurological disorders. 3 (6%) had grade (G) 3 flare, and 2 (4%) discontinued PD1 for flare. 15 (29%) developed other irAEs (5 G3), 3 (6%) discontinued PD1. 67 pts had irAEs requiring IS with prior IPI (9 G2, 51 G3, 7 G4), including 47 with greater than or = to, G3 colitis (15 had infliximab), 2 with G4 hepatitis (1 had antithymocyte globulin), and 9 with hypophysitis. All irAEs except hypophysitis had resolved at PD1 start except in 1 pt (arthritis), 5 were on IS at PD1 start. The ORR was 40%. 2 (3%) had recurrence of IPI irAEs with PD1 (arthritis, colitis), but 23 (34%) developed new irAEs (13, 19% greater than or = to, G3), and 11 (16%) discontinued PD1. There were no treatment related deaths. Conclusions: PD1 have efficacy in pts with preexisting AD and/or major irAEs with IPI. PD1 may flare preexisting AD, particularly rheumatologic, but GI and neurological disorders may flare less. In pts with prior major irAEs with IPI recurrence of the same irAE is rare, but new irAEs occur. The rate of irAEs in these pts appears higher than in clinical trial pts.

Not bad, not at all bad....

2.  Last year I posted this:  Folks with melanoma liver mets show less response to Pembro. Why?

On the other hand, President Jimmy Carter did well on Pembro (with very limited dosing compared to many...he's already off treatment) for brain met (radiated) and liver mets (though it was resected to some extent).   This from ASCO:

Correlation between metastatic site and response to anti-Programmed Death-1 (PD-1) agents in melanoma.  ASCO 2016. #9549.  J Clin Oncol 2016. Goldinger, tsai, Tume, Hamid, et al.

Background: Antibodies against PD-1 have shown remarkable activity in the treatment of advanced melanoma. Clinical characteristics correlated with response have not yet been fully characterized. Previous work has shown that site of metastasis can correlate with treatment outcome. We sought to confirm these findings in a larger cohort of metastatic melanoma patients (pts). Methods: We conducted a multicenter retrospective study in a cohort of metastatic melanoma pts treated with anti-PD-1 agents. Relevant clinical data were retrieved from electronic medical records. Tumor responses were evaluated by RECIST v1.1. A multivariate regression model was used to investigate demographic and clinical variables associated with response. Results: We identified 337 pts who received pembrolizumab (2mg/kg or 10mg/kg Q3W) (N = 326) or nivolumab (3mg/kg Q2W) (N = 11) on clinical trials (N = 273) and early access programs (N = 64) at UCSF and USZ. In the multivariate model, pts who had liver metastasis and prior ipilimumab were less likely to respond to treatment. Data for subgroup variables are shown. Pts with lung metastasis were more likely to respond. Conclusions: This multivariate model supports a significant correlation between liver metastasis and prior ipilimumab with poor response to anti-PD-1 therapy in melanoma pts. Presence of lung metastasis was correlated with improved response, though this just approached statistical significance. Efforts are ongoing to better identify clinical characteristics and biomarkers of response, and to further investigate the tumor microenvironment.

Variable	Odds ratio	95% CI	p value
Liver metastasis	0.35	0.21-0.59	< 0.0001
Prior ipilimumab	0.39	0.24-0.64	0.0002
Elevated LDH	0.64	0.40-1.04	0.07
ECOG	0.71	0.43-1.16	0.17
WBC	0.99	0.92-1.07	0.86
Brain metastasis	1	0.56-1.8	1
Lung metastasis	1.56	0.99-2.51	0.06

For what it's worth....

3.  I've actually posted these results before.  In Weber's report here:  Immunology update webinar for melanoma...  where this was presented (red in this case are HIS words):

   CheckMate 064: Study Design:  Cohort A - nivo 3mg/kg q2wk X6, then ipi 3mg/kg 1 3wk X4.  Cohort B - ipi 3mg/kg 1 3wk X4, then nivo 3mg/kg q2wk X6. Both cohorts were followed with nivo, 3mg/kg q2wk until progression, toxicity, or withdrawal of consent. Toxicities were no different between arms.  Significant difference in response rates due to sequence of drugs.  Nivo followed by ipi gave better results. "A real surprise."
I already had a rant on this one:  Sequential nivo then ipi orr of 41%!

Efficacy Summary Week 13				Week 25
NIVOàIPI (n = 68)		IPIàNIVO (n = 770)		NIVOàIPI (n = 68)		IPIàNIVO (n = 70)
Confirmed ORR, %	--		--		41.2		20.0
Complete response, n	0		0		0		0
Partial response, n	24		7		28		14
Conventional ORR, %	35.3		10.0		47.7		22.6
Progression rate, %	38.2 (26.7–50.8)		61.4 (49.0–72.8)		38.2 (26.7–50.8)		60.0 (47.6–71.5)

And you can see the link to my earlier rant above as well!!  At any rate...here it is...out of ASCO:  

Survival outcomes of nivolumab (NIVO) given sequentially with ipilimumab (IPI) in patients with advanced melanoma (CheckMate 064).  ASCO 2016. # 9517.  J Clin Oncol 34, 2016.  Weber, Gibney, Sullivan, Sosman, Slinghuff, Hodi, et al.

Background: Combining or sequencing immune checkpoint inhibitors that target distinct pathways has the potential to improve survival outcomes compared to either agent alone. This randomized phase II trial (CheckMate 064) previously demonstrated improved response rates with NIVO followed by IPI vs. IPI followed by NIVO in patients (pts) with advanced melanoma. Here, we report the first overall survival (OS) data from this study. Methods: Pts (N = 140) were randomized 1:1 to receive sequential induction treatment with NIVO 3 mg/kg Q2W x 6 doses followed by IPI 3 mg/kg Q3W x 4 doses (cohort A; N = 68 treated) or IPI 3 mg/kg Q3W x 4 doses followed by NIVO 3 mg/kg Q2W x 6 doses (cohort B; N = 70 treated). Following induction, both cohorts received NIVO 3 mg/kg Q2W until progression or unacceptable toxicity. The primary endpoint was to evaluate treatment-related grade 3–5 adverse events (AEs) during the induction periods until week 25. Secondary and exploratory endpoints included objective response rate (ORR) by modified RECIST v1.1, progression-free survival rate, and OS. Results: As previously reported, baseline patient characteristics were mostly balanced between cohorts, although a higher percentage of pts in cohort A had an ECOG performance status of 0, PD-L1 tumor expression ≥ 5%, and brain metastases. Similar rates of treatment-related grade 3/4 AEs occurred during both induction periods to week 25 in cohort A (50%) and cohort B (43%). In the updated analysis (minimum follow-up of 14 months), grade 3/4 AEs were 63% in cohort A and 50% in cohort B across all study periods, which led to discontinuation in 25% and 27% of pts, respectively. ORR was higher in cohort A than in cohort B (54% vs. 31%), with more complete responses (11% vs. 6%). A significant difference in OS was observed between cohorts A and B, with the median not reached vs. 16.9 months, and 1-year OS rates of 76% and 54%, respectively. Conclusions: NIVO followed by IPI continues to show improved efficacy outcomes vs. IPI followed by NIVO, with longer OS. A higher frequency of AEs was observed in cohort A, with the types and rates of AEs similar to that reported for NIVO given concurrently with IPI. 

Wishing you all my best.  - c