Background: Metastatic pancreatic adenocarcinoma (mPDAC) carries a dismal prognosis, with often a poor response to treatment (Tx). Although gemcitabine with nab-paclitaxel (GnP) is now a standard 1st-line (1L) Tx in mPDAC, provincial policies limit its use. To understand the implications of the restriction, this real-life study was conducted. Methods: A historical prospective data was compiled using the local electronic medical record (Endovault Oncology v3.0, NY) at the Jewish General Hospital-McGill. All patients (pts) diagnosed with mPDAC that received GnP from 2013-2016 were identified. Demographics and Tx data was obtained. Primary objective was to evaluate tolerability in all lines of Tx [grade >2 CTCAE v4.03, mean Tx delay (mTxd)]. Secondary objectives were Tx attrition and outcomes [RECIST 1.1, complete response (CR), stable disease (SD) and progression disease (PD)]. Results: Of the 40 pts (female 60%, median age 64.3), 25 (62.5%) had primary at the head of the pancreas and 30 (75%) liver metastasis. 1L Tx: GnP 24 (60%), FOLFIRINOX 10 (FFX, 25%), Gemcitabine 5 (Gc, 12.5%) and FOLFOX 1 (2.5%). 70% of both FFX and Gc pts had grade 2 peripheral neuropathy (PN), neutropenia (NEUT) and fatigue. Half GnP pts had grade 2 NEUT and fatigue. mTxd in GnP, FFX and Gc were 7, 14 and 7 days. All FFX and Gc pts had PD. GnP cohort had 1 CR and 7 SD. 2L Tx was in 22 (55%): GnP 13 (59.5%), FFX 3 (13.5%), Capecitabine 2 (9%), FOLFOX 2 (9%) and Gc (9%). FFX pts 66.6% had grade 3 fatigue. GnP pts had grade 2 NEUT, PN and fatigue in 53.8%, 15.3% and 46.2%. mTxd were GnP 14 days and FFX 14 days. Except 1 SD, all FFX pts had PD. GnP pts had 1 CR, 6 SD, and 6 PD. All pts on Capecitabine, FOLFOX and Gc had PD. 3L Tx was in 6 (15%): 3 GnP (50%) and 3 on clinical trials (50%). All pts had PD and < 3 cycles. In 1L FFX pts, 7/10 and 3/6 went on to receive GnP in 2L and 3L. All 1L Gc pts had 2L GnP. Only 3 pts and 1 pt from 1L GnP pts went on to receive 2L and 3L Tx. Conclusions: Due to access limitations to GnP, its use was likely reserved for poor performance status pts. Instead, 1L FFX was used when possible. 1L FFX pts received more Tx lines due to initial pt selection. Further studies are needed in broader populations to evaluate optimal 1L Tx selection accounting for expected attrition and overall survival.