Immune related adverse events (irAE) with platinum chemotherapy (CT) with durvalumab (D) ± tremelimumab (T): CCTG IND226

3058

Background: CT is immunomodulatory and requires corticosteroids (CS) premedication. We hypothesized that the incidence of irAE may be lower when D ± T is given with CT or CS. Methods: Patients (pts) receive CT (pemetrexed, nabpaclitaxel, etoposide or gemcitabine + cisplatin or carboplatin; usual 4-6 cycles) with D ± T, followed by D ± T alone (1 year total); pts with ≥ g2 (selected) or ≥ g3 irAE discontinued D ± T. Cycles were coded as CT + D ± T or D ± T; pts could contribute to both. CS: high (dexamethasone > 10mg/day for 5 days) or low CS. irAE were D ± T related gastrointestinal (GI), skin, endocrine, neurologic, hypersensitivity, pneumonitis (PN) or other immune (nephritis (GN), pancreatitis, hepatitis). Biochemistry (BIO; all causality): creatinine, transaminases/bilirubin (LFTS) and amylase/lipase was summarised. Results: In this ongoing study, 118 pts received 723 cycles. Pts had good performance status (PS 0-1), 78 had thoracic malignancies and 84 no prior CT. 44 pts continue on D ± T alone; 32 pts continue on CT + D ± T while 76 pts discontinued D ± T primarily due to disease progression; 15 discontinued for ≥ g2 irAE [PN (3), hepatitis (1), GN (2), adrenal (1), myocarditis (1), GI (3), thrombocytopenia (1), hyperthyroidism (1), encephalitis (1), pt decision (1)]. 67 pts had high CS cycles while 78 pts had low. 50% pts had irAE and 10% had ≥ g3 irAE, most commonly skin and GI. GI (15 vs 11%), skin (26 vs 20%) and PN (3 vs 0%) were reported in more pts during CT + D ± T cycles (non significant (NS)) ; hypothyroidism was more common with D ± T alone (18 vs 10%; p = NS). IrAE rates and severity were similar between high (67 pts) or low CS (78 pts) except for GI (19 vs 10%; p = NS). BIO were more common during CT + D ± T (74% of pts vs 48% p = 0.003); rates in high CS were similar to low CS. LFTs (ALT/AST - 41% vs 16%; 38% vs 9%; p = 0.005) and amylase/lipase (18 vs 9%; 19 vs 14%; p = NS) were more common in pts with CT + D ± T cycles vs pts with D ± T alone cycles. Conclusions: There is no evidence that CT or CS abrogates irAE in this exploratory analysis. GI, skin, pneumonitis, LFTS and amylase/lipase were more common during CT + D ± T suggesting additive/multifactorial causes; hypothyroidism is more common in D ± T cycles, which may reflect time on treatment. Clinical trial information: NCT02537418.

Does blood sugar impact on metastatic non-small cell lung cancer (NSCLC) outcome?

e20579

Background: Diabetes is a frequent diagnostic (9.3%) in general Western population, and because of its complications, becomes a comorbidity in NSCLC patients that could impact on prognostic. Impact of diabetes on NSCLC prognosis is unknown but the majority of patients fear eating desserts because feeding mostly the NSCLC tumor cells. Methods: Using our SARDO database and hospital files, we retrospectively reviewed 548 patients with metastatic NSCLC diagnosed and treated in our University hospital between 2006 and 2015. Uncontrolled diabetes was established when Hb1Ac > 7% or with more than 2 hospital glycemia over 11 mmol/L. Results: Amongst 418 metastatic NSCLC patients, 294pts (70%) had synchronous metastasis at diagnostic and 117pts (30%) have metachronous metastasis. As expected metachronous metastasis patients had a better survival (mOS 25mo) than synchronous metastasis (mOS 6mo). Then, we divided our patients in 4 subgroups: 1) Patients with diagnosed and uncontrolled diabetes (n = 78); 2) Patients with high glycemia but not previously diabetes diagnostic (n = 19); 3) Patients with controlled diagnosed diabetes (n = 59); 4) Control patients with normal glycemias or HBA1c (n = 262). Some tumor stage and demographic disparities were seen between the 4 groups but we controlled for them. In upfront metastatic NSCLC patients, mOS was 4.0 months for group 1, 4.0 months for group 2, 4.0 months for group 3 and 7.0 months for our control group, not statistically different, which is lower than literature but expected for this unselected population treated before 2015. In metachronous patients, group 1 mOS was 21mo, group 2 was 58mo, group 3 was 30mo and mOS was 23mo for control patients, with no statistically difference between groups. Women had a better mOS 8 mo compared to 4 mo in men (p = 0,003) independently of diabetes or sugar status. Conclusions: This retrospective data is reassuring regarding patients blood glycemia during the metastatic lung cancer treatment. Even in a long term metachronous disease perspective, blood sugar or diabetes diagnostic did not seem to influence the outcome. However, these data should be validated in a prospective study that could acknowledge if blood sugar control could improve lung cancer outcome.

Adjuvant chemotherapy outcome in unselected single-center non-small cell lung cancer (NSCLC) population

e20089

Background: Adjuvant chemotherapy with Cisplatinum/Vinorelbine is an established regimen since a decade in selected resected NSCLC population in order to improve DFS and overall survival. However, the outcome of patients treated outside the research protocols is not as good as in clinical trials. Methods: We retrospectively reviewed 209pts resected stage I to III NSCLC patients using our SARDO database, treated in our University hospital between 2006 and 2015 with a median follow-up of 13mos. Any adjuvant chemotherapy was recorded. Results: Amongst 209 resected NSCLC patients, 185pts (89%) had negative margins and 24pts (11%) have positive margins resection. The observation group and the chemotherapy group in both settings were well-balanced in terms of demographic, co-morbidities and tumor characteristics. As expected, the median overall survival (mOS) in negative margins was significantly better than in positive margins patients (43mos vs 24mos p = 0.029). In negative margins patients, no survival benefit was seen in stage I (A and B) (57 mos for chemo group vs 77 mos for observation group p = 0,217) and stage III (A and B) (39mos mOS for chemo group vs 40mos for observation group, p = 0.37), but a trend for improved mOS was seen in stage II (52 mos for chemo group vs 38 mos for observation group, p = 0.114), also consistent with mPFS results for each group. However, for the small number of patients with positive margins, a significant improvement in survival was seen in chemotherapy arm (mOS 37mos) compared to observational arm (mOS 11mos p = 0,039). No significant difference in survival was seen between Carbo/Taxol regimen and Cisplatinum/Vinorelbine regimen (43 mos vs 38 mos p = 0,658). Conclusions: In our unselected population, adjuvant chemotherapy did not show the improvement in survival expected with margin negative resected NSCLC. In positive resected NSCLC, adjuvant chemotherapy benefit was statistically significant, emphasizing need to treatment to these patients.

Phase III study of nab-paclitaxel (nab-P) plus gemcitabine (Gem) versus Gem alone in patients (pts) with metastatic pancreatic adenocarcinoma (mPC): Subgroup analysis of Canadian pts from the MPACT trial

439

Background: Weekly nab-P + Gem is a new option for first-line treatment (Tx) of mPC. In the MPACT trial, nab-P/Gem demonstrated superior overall survival (OS; primary endpoint) vs Gem alone as first-line Tx of mPC (Table). Here we report a subgroup analyses evaluating the efficacy and safety outcomes with nab-P + Gem vs Gem alone from the Canadian cohort of the MPACT trial. Methods: Previously untreated pts (N = 861) with mPC were randomized 1:1 (stratified by Karnofsky Performance Status [KPS], region, and the presence of liver metastases) to receive nab-P 125 mg/m2 + Gem 1000 mg/m2 on days 1, 8, and 15 of each 28-day cycle or Gem 1000 mg/m2 weekly for 7 weeks followed by 1 week of rest (cycle 1) and then days 1, 8, and 15 of each 28-day cycle (cycle ≥ 2). Results: 63 pts from Canada enrolled in the MPACT trial. Baseline pt characteristics were well balanced. Median age was 61 years and KPS was similar for both groups and comparable to the intent-to-treat (ITT) populations. Primary lesion in the pancreatic head was more common among pts in the nab-P + Gem vs Gem arm (55% vs 30%); use of biliary stent was similar (33% nab-P + Gem; 27% Gem). Median OS and progression-free survival (PFS) were longer with nab-P + Gem vs Gem (Table). Median Tx duration was 4.2 mo with nab-P + Gem vs 3.2 mo with Gem. Use of subsequent therapy was 30% in the nab-P + Gem arm vs 43% in the Gem arm. The median relative dose intensity for Gem was similar in each arm (81% nab-P + Gem vs 85% Gem). The most common grade ≥ 3 AEs for nab-P + Gem vs Gem were neutropenia (22% vs 10%), fatigue (34% vs 33%), and neuropathy (25% vs 0%). Conclusions: Canadian pts participating in MPACT were similar to the ITT population and nab-P + Gem was well tolerated and showed improved median OS, PFS, and ORR vs Gem alone, although not statistically significant (likely due to the small number of pts). Clinical trial information: NCT00844649. [Table: see text]

NSCLC driver mutations in the Quebec population: Epidemiologic and clinical evaluation

e22159

Background: Recent advances in lung cancer treatment embrace the recognition of molecular pathways implicated in its pathogenicity, paving the path to personalized therapies. We conducted a retrospective analysis to characterize the molecular features of the population treated for non-squamous non-small cell lung cancer (NS-NSCLC) in the province of Quebec. Methods: 622 patients with NS-NSCLC and adequate tumor blocks, treated at the CHUM between 2006 and 2008, were included. All samples were tested for ALK translocations (by IHC and FISH), EGFR classical exon 19 and 21 mutations by PCR (fragment analysis and qPCR) and for KRAS codon 12 and 13 mutations by mismatch PCR-RFLP. Molecular features were matched to demographic characteristics and clinical outcomes. Results: So far, complete results are available for 153 patients. Considering the amount of tumor tissue available, this population is largely represented by patients with local or loco-regional disease (n= 140, 91.5%). A minority of patients (10.3%) was never or light smokers (< 10 pack-yrs). Only 2 patients (1.3%) were of Asian descent. The following table depicts the outcomes of this cohort of patients segregated according to mutation status and extent of disease. Conclusions: ALK rearrangements were not identified in this unselected NS-NSCLC population characterized by localized disease and strong smoking history. ALK translocation prevalence in different populations is likely to be largely influenced by its tumor stage distribution, tobacco exposure and the use of selection criteria for molecular testing. An expanded cohort of patients will be presented at the meeting. [Table: see text]

Recurrent anal squamous carcinoma: Is HPV a risk factor?

e15094

Background: Anal squamous carcinoma (ASC) is a rare cancer, representing only 1.5 to 2% of all gastrointestinal (GI) malignancies. Combined chemo-radiotherapy therapy (CRT) is the treatment of choice with a response rate of 80% and recurrence rate of 10 to 15%. HPV is a risk factor for anal cancer but its role in cancer recurrence is less clear. We conducted a retrospective study to analyze clinical outcome of ASC patients in function of their HPV status. Methods: Patients with non-metastatic ASC treated between 2005-2011 at the Centre Hospitalier de l’Université de Montréal were eligible. Patient characteristics, clinical outcomes, and tumoral HPV status were retrospectively identified. Results: Seventy patients (M:31; F:39) were included in the analysis. The median age of patients was 51 years. Patient and tumor characteristics are shown in the Table. CRT was done in 57 patients. Median follow-up was 37.6 months. Eleven patients had recurrence of whom 10 were HPV-positive. Median TTR was 13.2 months. The 3 deaths related to cancer occurred during the first year of diagnosis. The 3 year overall survival was 90.08%. In multivariate analysis neither the advanced stage of disease nor the HPV-positive status had a statistically significant impact on relapse. Conclusions: In this small cohort, we found a high prevalence of HPV positivity in ASC but HPV infection couldn’t be identified as a risk factor for ASC relapse. Inclusion of men in the HPV vaccination program may help in reducing HPV related ASC. [Table: see text]

Assessment of safety of bevacizumab (bev) use for metastatic colorectal cancer (mCRC) in elderly patients treated at the Centre Hospitalier de l’Université de Montréal (CHUM): A retrospective study

677

Background: A 2004 randomized study showed a meaningful improvement in survival with bev added to chemotherapy (CT) for patients with mCRC. This combination is now widely used and generally well tolerated although adverse events (AEs) are described associated with the use of bev, some of which may be increased in incidence in an elderly population. In this study, we assessed these AEs in a population of elderly patients treated with bev in a major center in Montréal.

Methods: Patients with mCRC who received bev at our institution outside a clinical trial were retrieved from the pharmacy registry. Medical records were sought from patients who qualified for the inclusion criteria: ≥ 65 years, bev therapy at 5mg/kg every two weeks, combination with CT, period 2007-2011. A retrospective analysis of specific AEs was done and grade according to NCI-CTC (version 4.0) was assessed: febrile neutropenia, hypertension, proteinuria, diarrhea, venous thromboembolic event (VTE), arterial thromboembolic event (ATE), bleeding and gastrointestinal perforation.

Results: 47 patients received a total of 486 cycles of bev. Out of these cycles, 210 were given with FOLFOX, 199 with FOLFIRI and 77 with other combinations. Median age of the patients was 69 years. 3 episodes of febrile neutropenia were noted. Grade 3 or 4 arterial hypertension occurred in 4 patients of which 2 had to definitively stop bev. Grade 3 proteinuria was reported with definitive cessation of bev in 2 patients. Grade 3 or 4 diarrhea was observed in 2 cases. We reported 1 cerebral ATE leading to discontinuation of bev and grade 2 VTE in two patients. One grade 5 gastrointestinal perforation was reported. No major bleeding was noted.

Conclusions: The number of adverse events observed in this population of elderly patients is comparable to other studies. Thus, age should not be an exclusion criteria for treatment with bev and CT.