This Management’s Discussion and Analysis of Financial Condition and Results of Operations for the three months ended March 31, 2019 and 2018 should be read in conjunction with our annual report on Form 10-K for the year ended December 31, 2018 filed with the Securities and Exchange Commission and “Risk Factors” section included elsewhere in this Form 10-Q. This Form 10-Q contains forward-looking statements within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, and Section 27A of the Securities Act of 1933, as amended. When used in this report, the words “believe,” “anticipate,” “intend,” “plan,” “estimate,” “expect,” “may,” “will,” “could,” “potentially” and similar expressions are forward-looking statements. Such forward-looking statements are based on current expectations and beliefs. Any such forward-looking statements are not guarantees of future performance and involve risks and uncertainties. Actual events or results may differ materially from those discussed in the forward-looking statements as a result of various factors.

Forward-looking statements made in this report include, for example, statements about:

18

Forward-looking statements are not guarantees of future performance and involve risks and uncertainties. Actual events or results may differ materially from those discussed in the forward-looking statements as a result of various factors. For a more detailed discussion of such forward looking statements and the potential risks and uncertainties that may impact upon their accuracy, see the “Risk Factors” section and “Overview” section of this Management’s Discussion and Analysis of Financial Condition and Results of Operations. These forward-looking statements reflect our view only as of the date of this report. We undertake no obligations to update any forward-looking statements. You should also carefully consider the factors set forth in other reports or documents that we file from time to time with the Securities and Exchange Commission.

Overview

We are a biopharmaceutical company with research and development programs broadly falling into two categories: (i) new chemical entities derived from our Epigenetic Regulator Program, in which we attempt to discover and develop molecules which have not previously been approved and marketed as therapeutics, and (ii) proprietary pharmaceutical programs, in which we apply our formulation expertise and technologies largely to active pharmaceutical ingredients whose safety and efficacy have previously been established but which we aim to improve in some manner through a new formulation. We also manufacture and sell osmotic pumps used in laboratory research and design, develop and manufacture a wide range of standard and custom biodegradable polymers and excipients for pharmaceutical and medical device clients for use as raw materials in their products. In addition, we conduct research and development of pharmaceutical products in collaboration with third party pharmaceutical and biotechnology companies.

Our product pipeline currently consists of multiple investigational drug candidates in development.  DUR‑928, a new chemical entity in Phase 1 and 2 development, is the lead candidate in DURECT’s Epigenetic Regulator Program.  An endogenous, orally bioavailable small molecule, DUR-928 has been shown in preclinical studies to play an important regulatory role in lipid homeostasis, inflammation, and cell survival.  Human applications may include acute organ injury such as alcoholic hepatitis (AH) and acute kidney injury (AKI), chronic metabolic diseases such as nonalcoholic fatty liver disease (NAFLD), nonalcoholic steatohepatitis (NASH) and other liver diseases, and inflammatory skin conditions such as psoriasis and atopic dermatitis.  DURECT’s proprietary oral and injectable delivery technologies are designed to enable new indications and enhanced attributes for small-molecule and biologic drugs.  One late-stage development program in this category is POSIMIR® (bupivacaine extended-release solution), an investigational analgesic product intended to deliver bupivacaine to provide up to three days of pain relief after surgery.

A central aspect of our business strategy involves advancing multiple product candidates at one time, which is enabled by leveraging our resources with those of corporate collaborators. Thus, certain of our programs are currently licensed to corporate collaborators on terms which typically call for our collaborator to fund all or a substantial portion of future development costs and then pay us milestone payments based on specific development or commercial achievements plus a royalty on product sales. At the same time, we have retained the rights to other programs, which are the basis of future potential collaborations and which over time may provide a pathway for us to develop our own biopharmaceutical commercial, sales and marketing organization.

Additional details of these programs and related strategic agreements are contained in our annual report on Form 10-K for the year ended December 31, 2018 and in Note 2 of the financial statements included in Item 1 above.

Epigenetic Regulator Program and New Chemical Entities

Epigenetic regulation involves biochemical modification of either DNA itself or of proteins that intimately associate with DNA.  These modifications lead to changes to gene expression that facilitate downstream biological effects.  

19

DURECT’s Epigenetic Regulator Program involves a multi-year collaborative effort with the Department of Internal Medicine at Virginia Commonwealth University (VCU), the VCU Medical Center and the McGuire VA Medical Center. The knowledge base supporting this program is a result of more than 3 0 years of lipid research by Shunlin Ren, M.D., Ph.D., Professor of Internal Medicine at the VCU Medical Center and a recipient of multiple grants from the National Institutes of Health (NIH) and other agencies for metabolic disease research. The lead compound from this program , DUR-928 , is an endogenous, orally available small molecule that modulates the activity of various nuclear receptors that play important regulatory roles in lipid homeostasis, inflammation and cell survival.   Under a license with VCU, w e hold the exclusive royalty- bearing worldwide right to develop and commercialize DUR-928 and related molecules discovered in the program.

The biological activity of DUR-928 has been demonstrated in over a dozen different animal disease models involving three animal species. Several of these disease models represent chronic metabolic disorders of hepatic lipid accumulation and dysfunction (e.g., nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH)) and several represent acute organ injuries (e.g., endotoxin shock, acute oxidative damage, ischemic-reperfusion kidney injury, and stroke models).

Our major product research and development efforts planned for DUR-928 in 2019 are set forth in the following table:

In pharmacokinetic and/or toxicity studies conducted in mice, hamsters, rats, rabbits, dogs, minipigs and monkeys, DUR-928 has been found to be orally available, locally tolerable and safe by all routes tested to date. These non-clinical results support the use of DUR-928 in the completed, ongoing and planned human safety, pharmacokinetics (PK) and proof-of-concept trials.

Acute Organ Injury Program with Injectable DUR-928

Market Opportunity .    Acute organ injury is another area of major unmet medical need for which effective pharmaceutical treatment is often lacking. Alcoholic liver disease (ALD) is a syndrome characterized by progressive inflammatory liver injury associated with long-term heavy intake of alcohol, and involves a spectrum that ranges from mild injury to severe, life threatening injury.  Alcoholic hepatitis (AH), is an acute, inflammatory form of ALD for which there are no effective therapeutics available. The prevalence of AH is believed to be 10-35% of heavy drinkers.  There were over 320,000 hospitalizations related to alcoholic hepatitis in 2010, and the hospitalization costs amounted to nearly $50,000 per patient. The cost of a liver transplant exceeds $800,000. Acute kidney injury (AKI), a sudden loss of kidney function due to renal failure or injury, affects approximately 2.8 million patients per year in the United States and is associated with increased mortality, prolonged hospital stays, kidney dialysis and progression to chronic kidney disease. There are various forms of acute organ injury affecting the liver, the kidney or other organs for which we are or may seek to develop DUR-928.

20

Clinical Program .     In 2018, we began a Phase 2a clinical trial to evaluate intravenously infused DUR-928 in patients with moderate and severe alcoholic hepatitis (AH).  This is an open label, dose escalation (30, 90 and 150 mg), multi-center U.S. study, originally designed to be conducted in two sequential parts.  Part A includes patients with moderate AH and Part B includes patients with severe AH (as determined by the Model of End-Stage Liver Dis ease (MELD) scores, a common scoring system to assess the severity and prognosis of AH patients).  Dose escalation may occur following review of safety and pharmacokinetic (PK) results of the prior dose level by a Dose Escalation Committee (DEC). The targe t number of patients for the study is 4 per dose group. The objectives of this study include assessment of safety, PK and pharmacodynamic (PD) signals, including liver biochemistry and biomarkers. After completing dosing for the low-dose 30 mg cohort (n=4) of Part A (moderate AH patients), the DEC approved commencement of the 90 mg cohort in Part A while simultaneously commencing recruitment for Part B (severe AH patients) with the 30 mg dose.  Upon completion of dosing for the 30 mg cohort (n=4) of Part B, the enrollment of which was more rapid than Part A, the DEC approved commencement of the 90 mg cohort in Part B. In parallel with our ongoing trial, we are supporting Dr. Craig McClain’s efforts to initiate an NIH-funded study of DUR-928 in AH patients at the University of Louisville.

Ten patients have completed dosing with DUR-928 to date in the ongoing open label, dose-escalation, multi-center U.S. trial. Eight patients (4 moderate and 4 severe) have been treated with DUR-98 at the 30 mg dose, and two patients (1 moderate and 1 severe) at the 90 mg dose.

Lille scores are used in clinical practice to help determine the prognosis for AH patients after 7 days of treatment. Patients with a Lille score below 0.45 have an 85% 6-month survival rate vs. those with Lille scores of above 0.45, have only a 25% 6-month survival rate ( Louvet A et al. Hepatology 2007; 45: 1348-54 ). The lower the Lille score, the better the prognosis is for the AH patient. In our study, the median Lille score for the 9 AH patients treated with DUR-928 who returned for their Day 7 visit is 0.04, with a range of 0.01 to 0.19. The median Lille score among a cohort of 15 patients treated with either supportive care or supportive care with corticosteroids at the University of Louisville (UL) is 0.41 (shown as historical control). ¹

The chart below shows individual patient Lille scores plotted as a function of their initial MELD scores.

21

Mathurin et al. proposed three prognostic classifications of AH patients in response to treatment based on Lille scores and their correlation with 28-day survival rates from a meta-analysis of four randomized controlled trials evaluating the effectiveness of corticosteroids in an aggregate of 324 patients with severe AH. ( Methurin P. et al. Gut 2011; 60:255-260 ) The following table shows the percentage of patients from three AH data sets, including DUR-928, in each Mathurin classification based on patients’ Lille scores. Seventy-eight percent (7/9) of the DUR-928 treated AH patients with Lille scores are classified as complete responders, 22% (2/9) are partial responders and none (0/9) were null responders.  

Bilirubin is formed by the breakdown of red blood cells in the body. The level of total bilirubin in the blood is an indication of how the liver is functioning.  Compared to baseline (n=10), the median reduction in total bilirubin in the DUR-928 treated patients was 16% at Day 7 (n=9) and 41% at Day 28 (n=8) compared to 3% at Day 7 and 35% at Day 28 in the UL patients.  The chart below shows the percent change in total bilirubin at Day 7 and 28 compared to baseline (Day 0) for both the separate UL (as historical control) and DUR-928 studies.  

P-Values calculated with Wilcoxon's Signed Rank Test

22

Model of End-Stage Liver Disease (MELD) score is another common scoring system used to assess the severity and prognosis of AH patients. Patients with MELD scores of 11-19 are classified as having moderate AH and patients with MELD scores of 20-30 are clas sified as having severe AH.  As with Lille scores, the lower the MELD score, the better the prognosis for the AH patient. In our study (shown in the chart below), the median reduction from baseline (Day 0, prior to treatment) (n=10) in MELD in the DUR-928 treated patients was 4% at Day 7 (n=9) and 21% at Day 28 (n=8) compared to a 4% increase at Day 7 and 6% reduction at Day 28 in the UL patients (as historical control).

MELD is calculated based on (a) bilirubin, (b) serum creatinine (sCr), and (c) International Normalized Ratio (INR), which is a measure of prothrombin time.    P-Values calculated with Wilcoxon's Signed Rank Test

To date there is no statistical difference in the pharmacokinetic profiles between moderate and severe AH patients treated with DUR-928. There have been no drug-related adverse events in the DUR-928 treated patients to date. The data presented are preliminary and will be finalized upon completion of the trial. There can be no assurance that additional patients treated with DUR-928 will have similar results as those reported here.

Phase 1 trials of DUR-928 administered through injection have supported the development of DUR-928 in AH. The initial Phase 1 trial in healthy subjects was a single-site, randomized, double-blinded, placebo-controlled, single-ascending-dose study that evaluated the safety, tolerability and PK of intramuscular (IM) injected DUR-928. The 24-subject study (16 healthy volunteers on the drug and 8 on placebo) of four escalating dose levels resulted in dose proportional systemic exposure of DUR-928 with peak plasma concentrations greater than 1000-fold higher than endogenous levels. DUR-928 was well-tolerated at all dose levels, with no serious treatment-related adverse events reported. We also conducted a multiple-dose study involving 10 healthy subjects, in which participants received IM-injected DUR-928 for 5 consecutive days (8 subjects on the drug, 2 on placebo) using the next to highest dose from the single dose study. No serious treatment related adverse events were reported, no subjects withdrew from the study, no accumulation in plasma concentrations were observed with repeat dosing, and the pain scores and injection site reactions were minimal. We also conducted a single-ascending dose intravenous infusion (IV) study with 16 healthy subjects and observed no treatment-related serious adverse events. The systemic exposure following IV infusion was dose proportional.

A Phase 1 drug-drug interaction study conducted in healthy subjects demonstrated that neither orally administered nor intravenously injected DUR-928 at doses tested affected the safety and PK of midazolam, a drug metabolized by CYP3A4, which is one of the important enzymes associated with clinically relevant drug-drug interactions.

We have also conducted a Phase 1b study with injected DUR-928 in patients with impaired kidney function (stage 3 and 4 chronic kidney disease (CKD)) and matched control subjects (MCS), matched by age, body mass and gender with normal kidney function. This study was a single-site, open-label, single-ascending-dose study. in two successive cohorts (first a low dose of 30 mg and then a high dose of 120 mg) evaluating safety and PK of intramuscular injected DUR-928. The low dose cohort consisted of 6 patients with CKD and 3 MCS; the high dose cohort consisted of 5 CKD patients and 3 MCS.  In this trial, DUR-928 was well tolerated among all subjects and the PK parameters between the kidney function impaired patients and the MCS were comparable. While the number of subjects involved was small and not designed to assess efficacy, we did observe decreases in bilirubin and CK-18 when those levels were elevated at baseline, although the results were not statistically significant. The results of this Phase 1b trial were presented at Kidney Week 2018 in San Diego, CA.

Skin Inflammatory Disorder Program with Topical DUR-928

Market opportunity.   Psoriasis is an inflammatory skin disease and an immune-mediated condition that causes the body to make new skin cells in days rather than weeks.  In the United States, there are about 150,000 new cases of psoriasis every year and it affects more than 8 million Americans.  According to the International Federation of Psoriasis Associations (IFPA), nearly 3% of the world’s

23

population , or about 125 million peopl e , has some form of psoriasis. Psoriasis causes itchiness and irritation and may be painful. There i s no cure for psoriasis, but treatment can ease symptoms.   

Approximately 80% of patients with psoriasis have localized disease. Topical agents remain the mainstay of localized psoriasis treatment. Other inflammatory skin disorders, such as atopic dermatitis, which affects approximately 32 million Americans, represent significant unmet medical needs. Most currently available topical treatments, such as steroids, are typically employed as first-line therapy and either slow down excessive skin cell proliferation or reduce inflammation.

Clinical program.   We have conducted an exploratory Phase 1b trial in psoriasis patients (9 evaluable patients) in Australia.  The double-blinded and placebo-controlled trial was conducted using a micro-plaque assay with intralesional injections of DUR-928.  We believe that the initial results were encouraging and warrant further investigation.  As a result, we are conducting a Phase 2a, randomized, double-blind, vehicle-controlled proof-of-concept clinical trial, in which DUR-928 is applied topically once-daily for 28 days in patients with mild to moderate plaque psoriasis. The trial is being be conducted at multiple clinical sites in the U.S.  Twenty patients are planned to be enrolled to obtain approximately 15 evaluable patients. Patients serve as their own controls, applying DUR-928 to the plaque on one arm and the vehicle (placebo) to a similar plaque on the other arm. After the treatment period, patients will be followed for an additional four weeks. The primary efficacy endpoint will be change in local psoriasis scores from baseline in the DUR-928-treated plaques compared to that in the vehicle-treated plaques. We began dosing patients in March 2019 and expect to announce top-line data from this study in the second half of 2019.

Chronic Liver Disease Program with Orally Administered DUR-928

Market Opportunity.     Non-alcoholic fatty liver disease (NAFLD) is the most common form of chronic liver disease in both children and adults.   It is estimated that NAFLD affects approximately 20% to 30% of adults and 10% of children in the United States. Non-alcoholic steatohepatitis (NASH), a more severe and progressive form of NAFLD, is one of the most common chronic liver diseases worldwide, with an estimated prevalence of more than 10% of adults in the United States, Europe, Japan and other developed countries. No drug is currently approved for treatment of NAFLD or NASH.  Moreover, alcoholic fatty liver disease (AFLD), including its more advanced stage, alcoholic steatohepatitis (ASH), develops in approximately 90% of individuals who drink more than 60 grams/day of alcohol, but may occur in individuals who drink less, and is a major contributor to the global burden of liver cirrhosis.  In addition to these liver diseases, there are a number of orphan liver diseases for which we may seek to develop DUR-928.

Clinical Program.     In March 2019 we began enrolling patients in a Phase 1b randomized and open-label clinical study being conducted in the U.S. to evaluate safety, pharmacokinetics and signals of biological activity of DUR-928 in NASH patients with stage 1-3 fibrosis.  Three doses of DUR-928 (50 mg QD, 150 mg QD and 300 mg BID) will be administered orally for 28 consecutive days with approximately 20 patients per dose group for a total of approximately 60 patients in the trial as shown below. Key endpoints include safety and pharmacokinetics (PK), clinical chemistry and biomarkers (e.g., bilirubin, lipids, liver enzymes, CK-18s, and inflammatory cytokines) as well as liver imaging. We expect to announce initial data from this study in the second half of 2019.  

DUR-928 Phase 1b 28-Day Daily Oral Dosing Clinical Trial Schema

Screening

Run-in  2 weeks

( baseline data)

28-day dosing

28-day follow-up

DURECT has completed multiple single-site Phase I trials in healthy subjects with orally administered DUR-928. One of the studies was a randomized, double-blinded, placebo-controlled, single-ascending-dose study that evaluated the safety, tolerability and PK of DUR-928. The 30-subject study evaluated DUR-928 in five cohorts of healthy volunteers receiving DUR-928 (n=20 on drug, 10 on placebo) at escalating doses that resulted in peak plasma concentrations greater than 100-fold higher than endogenous levels. DUR-928 was well-tolerated at all dose levels, with no serious treatment-related adverse events reported. Dose related increases in plasma concentrations were observed with peak plasma concentration at approximately 2-6 hours after dosing. We subsequently conducted a randomized, double-blinded, placebo-controlled, multiple-ascending-dose study in 20 healthy subjects (n=16 on drug, 4 on placebo).  Following multiple dosing, DUR-928 was well-tolerated at all doses, with no serious drug-related adverse events reported and no accumulation in plasma concentrations observed with repeat dosing.  We also conducted a food effect study with 8 healthy subjects and observed no food effect on DUR-928 absorption.

We also conducted a Phase 1b trial in cirrhotic and non-cirrhotic NASH patients and matched control subjects (MCS) (matched by age, body mass index and gender with normal liver function) utilizing orally administered DUR-928. This was an open-label,

24

single-ascending-dose safety and PK study conducted in Australia in two successive dose cohorts (first a low dose of 50 mg and then a high dose of 200 mg ) . Both cohorts consisted of 10 NASH patients and 6 MCS. Data from this study was presented at the International Liver Congress™ 2017 organized by the European Association for the Study of the Liver (EASL) in Amsterdam on April 22, 2017. All patients and MCS in this study tolerated DUR-928 well.  One patient (with a prior history of arrhythmia and an ongoing viral infection) in the high dose cohort experienced a serious adverse event (shortness of breath), whi ch occurred without unusually abnormal biochemical changes and resolved without intervention but was considered possibly treatment related by the physician due to its temporal association with dosing.  In both the low and high dose cohorts, the PK paramete rs were comparable between the NASH patients and the MCS.  In addition, the systemic exposure following the low and high doses of DUR-928 was dose dependent.

While this study was not designed to assess efficacy, we observed significant reductions from baseline of several biomarkers after both doses of DUR-928.   A single oral dose of DUR-928 statistically significantly reduced the levels of both full-length and cleaved cytokeratin-18 (CK-18), bilirubin, hsCRP, and IL-18 in these subjects.  The mean decrease of full-length CK-18 (a generalized cell death marker) at the measured time point of greatest effect (12 hours after dosing) was 33% in the low dose cohort and 41% in the high dose cohort.  The mean decrease of cleaved CK-18 (a cell apoptosis marker) at the measured time point of greatest effect (12 hours after dosing) was 37% in the low dose cohort and 47% in the high dose cohort.  The mean reduction of total bilirubin (a liver function marker) at the measured time point of greatest effect (12 hours after dosing) was 27% in the low dose cohort and 31% in the high dose cohort.  The mean decrease of high sensitivity C-Reactive Protein (hsCRP) (a marker of inflammation) at the measured time point of greatest effect (24 hours after dosing) was 8% in the low dose cohort and 13% in the high dose cohort.   The mean decrease of IL-18 (an inflammatory mediator) at the measured time point of greatest effect (8 hours after dosing) was 4% in the low dose cohort and 8% in the high dose cohort.

Collectively, the reduction of these biomarkers plus results from our animal and cell culture studies suggest potential therapeutic activity of DUR-928 for patients with liver diseases.  However, additional studies are required to evaluate the safety and efficacy of DUR-928, and there is no assurance that these biomarker effects will be associated with clinically relevant benefits, or that DUR-928 will demonstrate safety or efficacy in treating liver diseases in future controlled trials.  

Additional Proprietary Pharmaceutical Programs

POSIMIR ^® (Extended Release Bupivacaine)

Our post-operative pain relief depot, POSIMIR, is a sustained release injectable using our proprietary SABER ^® delivery system to deliver bupivacaine, an off-patent pharmaceutical agent. SABER is a controlled drug delivery technology that is administered via the parenteral (i.e., injectable) route to deliver drugs that act systemically or locally. POSIMIR is designed to be administered to a

25

surgical site at the end of surgery for post-operative pain relief and is intended to provide local analgesia for up to 3 days, which we believe coincides with the time period of the greatest need for post-surgical pain control in most patients.

Status.   In April 2013, we submitted an NDA as a 505(b)(2) application, which relied in part on the FDA’s findings of safety and effectiveness of a reference drug. In February 2014, we received a Complete Response Letter from the FDA. Based on the Complete Response Letter and subsequent communications with the FDA, we conducted a new POSIMIR Phase 3 clinical trial (the PERSIST trial) consisting of patients undergoing laparoscopic cholecystectomy (gallbladder removal) surgery to further evaluate the benefits and risks of POSIMIR. We began recruiting patients for this trial in November 2015 comparing POSIMIR to placebo. Based on advice from the FDA received subsequent to the start of the trial, in April 2016 we decided to amend the PERSIST trial. Starting in August 2016, we began implementing Part 2 of the PERSIST trial to evaluate POSIMIR against standard bupivacaine HCl rather than placebo as we had been doing initially in the study. In October 2017, we reported that the PERSIST trial did not meet its primary efficacy endpoint of reduction in pain on movement over the first 48 hours after surgery as compared to standard bupivacaine HCl.   While results trended in favor of POSIMIR versus the comparator, they did not achieve statistical significance.

In total, we have completed 16 clinical studies with POSIMIR, in seven different surgical procedures, including inguinal hernia repair, shoulder surgery (primarily subacromial decompression), appendectomy, abdominal hysterectomy, open laparotomy, laparoscopic cholecystectomy, and laparoscopic colectomy. The incision lengths treated ranged from a few centimeters for laparoscopic portals, to open laparotomy incisions of up to 35 cm. The seriousness of the surgery ranged from day surgery hernia repair in relatively healthy patients to major abdominal surgery for colon cancer in elderly patients with substantial co-morbidity who were often hospitalized for a week or more. The safety experience from this variety of procedures and patients was designed to allow for extrapolation of the safety and efficacy data to a broad surgical population. Our POSIMIR clinical development program has been devised to establish the safety and efficacy of POSIMIR for the treatment of post-surgical pain for 3 days.

26

Efficacy

We believe that two of the completed clinical trials support the efficacy of POSIMIR in post-operative pain and meet the requirements to be considered as adequate and well-controlled pivotal clinical trials:

Hernia pivotal efficacy trial

The hernia pivotal efficacy clinical trial was designed to evaluate the tolerability, activity, dose response and pharmacokinetics of POSIMIR in patients undergoing open inguinal hernia repair. The trial was conducted in Australia and New Zealand as a multi-center, randomized, double blind, placebo-controlled study in 122 patients. Study patients were randomized into three treatment groups: patients that were treated with POSIMIR 2.5 mL (n=43), POSIMIR 5 mL (n=47) and SABER placebo (n=32). The co-primary efficacy endpoints for the study were Mean Pain Intensity on Movement area under the curve (AUC), a measure of pain over a period of 1-72 hours post-surgery, and the proportion of patients requiring supplemental opioid analgesic medication during the study (defined as 0-15 days).

In relation to the co-primary endpoint of pain reduction as measured by Mean Pain Intensity on Movement AUC 1-72 hours post-surgery, the patient group treated with POSIMIR 5 mL reported thirty-one percent (31%) less pain versus placebo, and the result was statistically significant (p=0.0031). Fifty-three percent (53%) of the study patients in the POSIMIR 5 mL group took supplemental opioid analgesic medications versus seventy-two percent (72%) of the placebo patients (p=0.0909). Although this positive trend for this co-primary endpoint in favor of the POSIMIR 5 mL group was not statistically significant, both secondary endpoints measuring opioid analgesic medication consumption were met at a statistically significant level. During the periods of 1-24 hours, 24-48 hours and 48-72 hours after surgery, placebo patients consumed approximately 3.5 (p=0.0009), 2.9 (p=0.0190) and 3.6 (p=0.0172) times more supplemental opioid analgesic medications (mean total daily consumption of opioid analgesic medication in morphine equivalents), respectively, than the POSIMIR 5 mL treatment group. The median decrease in supplemental opioid analgesics taken over the first three days after surgery was 80% (p=0.0085) for the POSIMIR 5 mL group as compared to the placebo group.

Shoulder pivotal efficacy trial

The shoulder pivotal efficacy trial was a multicenter, randomized, double-blind, active- and placebo-controlled, parallel-group, dose-response trial conducted at 9 investigational centers in Europe. In this study, 107 patients were randomly assigned to one of three treatment groups prior to undergoing elective arthroscopic shoulder surgery: POSIMIR 5 mL (n=53), SABER-Placebo (n=25) or bupivacaine HCl solution (n=29). All patients were given a background pain treatment consisting of a daily dose of two or four grams (depending on the patient’s weight) of paracetamol (acetaminophen). In addition, each patient was provided supplemental opioid rescue medication, if needed. With respect to efficacy, the primary endpoints of the study were to demonstrate: (1) an improvement in terms of pain intensity on movement area under the curve (AUC) during the period 1–72 hours post-surgery, and (2) a decrease in the total use of opioid rescue analgesia 0–72 hours post-surgery.

Results from this study demonstrate that the POSIMIR group experienced a statistically significant reduction in pain intensity of approximately 20% (p=0.012) versus SABER-Placebo. Applying the appropriate statistical test given the data distribution, the POSIMIR group showed a statistically significant reduction of approximately 67% (p=0.013) in median opioid use in favor of POSIMIR. No statistical differences were found when POSIMIR was compared to bupivacaine HCl.

Safety

Bupivacaine is a well-known drug with an extensive understanding of its risks and benefits. DURECT has completed 16 clinical trials involving POSIMIR. A total of 1463 patients participated in these trials, 1144 of whom were exposed to POSIMIR or SABER-Placebo in volumes ranging from 2.5 to 10 mL. A total of 876 patients were exposed to POSIMIR with the dose of bupivacaine ranging from 330 to 990 mg. Our planned commercial dose is 5 mL (660 mg). In addition, a total of 272 patients were treated with bupivacaine HCl in control groups and 268 patients received SABER-Placebo in control groups.

Overall, the POSIMIR patient groups showed a similar systemic safety profile as the patient groups treated with SABER-Placebo and bupivacaine HCl. Local site reactions were observed more frequently in the POSIMIR and SABER-Placebo groups than in the active comparator groups, most frequently in abdominal surgeries; most of these observations were discolorations (e.g., surgical bruising), the majority of which resolved without treatment during the observation period. There was little difference in the incidence of severe or serious adverse events between the POSIMIR, SABER-Placebo and bupivacaine HCl treatment groups. Most of the serious adverse events seen in these trials appear to be due to complications of surgery, anesthesia, analgesics, or co-morbidity and not POSIMIR-related. The clinical history for serious adverse events has been reviewed, and we do not believe POSIMIR was associated with any bupivacaine toxicity. The adverse event data was analyzed in a variety of ways to detect any evidence of bupivacaine central nervous system or cardiac toxicity or other unexpected effects. No patients treated with POSIMIR had an instance of a severe central nervous system or cardiac adverse event traditionally associated with bupivacaine toxicity.

27

Plans for FDA Submission

After carefully reviewing the existing POSIMIR data and evaluating the feedback we have received from the FDA, including the CRL and other correspondence, we plan to submit a response to the CRL to the FDA in the second quarter of 2019. The submission will request FDA approval of POSIMIR based on what the Company and its advisors believe is adequate evidence of both safety and efficacy. As the submission is intended to be a response to a CRL, we expect a 6-month FDA review period.  

Market Opportunity .    According to data published by the Center for Disease Control and Prevention, there are approximately 72 million ambulatory and inpatient surgical procedures performed annually in the U.S. Insufficient postoperative pain control remains a significant problem, with studies indicating that roughly 65% of patients experience moderate-to-extreme pain after surgery. The current standard of care for post-surgical pain includes a variety of opiate and non-opiate analgesics and muscle relaxants. While systemic opioids can effectively reduce post-surgical pain, they commonly cause side effects including drowsiness, constipation, nausea and vomiting, and cognitive impairment. Effective pain management can be compromised if patients fail to adhere to recommended dosing regimens because they are suffering from these side effects. Post-surgical pain also can be treated effectively with local anesthetics; however, their usefulness often is limited by their short duration of action .

PERSERIS ^™ (risperidone)

In September 2017, we entered into an agreement with Indivior, under which we assigned to Indivior certain patents that may provide further intellectual property protection for PERSERIS, Indivior’s extended-release injectable suspension for the treatment of schizophrenia in adults.  In consideration for such assignment, Indivior made an upfront non-refundable payment to DURECT of $12.5 million.  Indivior also paid a $5 million milestone payment to DURECT in August 2018 following the FDA approval of PERSERIS. Under the terms of the agreement with Indivior, DURECT receives quarterly earn-out payments that are based on a single digit percentage of U.S. net sales of PERSERIS.   

Indivior has announced the commercial launch of PERSERIS in February 2019 with a 50-person field force.  While Indivior has projected a potential peak net revenue goal of $200 to $300 million per year for PERSERIS, there can be no assurance such revenues will be achieved.       

ORADUR-ADHD Program

Market Opportunity .    Attention Deficit Hyperactivity Disorder (ADHD) is a neurobehavioral condition that is estimated to affect over 6 million (approximately 9%) of U.S. children ages 2-17, according to the U.S. Department of Health and Human Services. The principal characteristics of ADHD are inattention, hyperactivity, and impulsivity. The condition presents itself in childhood and can be life long as a significant number of children with ADHD continue to present symptoms as adults. 62% of children with ADHD are estimated to be under treatment with medication, utilizing stimulants such as amphetamine or methylphenidate as first-line treatments. U.S. sales of ADHD treatments were approximately $10.4 billion in 2016. The 2016 National Survey on Drug Use & Health estimates that 612,000 Americans over the age of 12 abuse stimulants for euphoric highs and increased performance or wakefulness.

Development Strategy .    In collaboration with Orient Pharma, we have developed a drug candidate (ORADUR- Methylphenidate ER) based on our ORADUR Technology for the treatment of ADHD. This drug candidate is intended to provide once-a-day dosing with added tamper resistant characteristics to address common methods of abuse and misuse of these types of drugs. In August 2009, we entered into a development and license agreement with Orient Pharma, a diversified multinational pharmaceutical, healthcare and consumer products company with headquarters in Taiwan, under which we granted to Orient Pharma development and commercialization rights in certain defined Asian and South Pacific countries to ORADUR-Methylphenidate ER. We retain rights to North America, Europe, Japan and all other countries not specifically licensed to Orient Pharma.

Orient Pharma conducted a Phase 3, multi-center, randomized, double-blind, placebo-controlled, two-way cross-over study designed to demonstrate the efficacy and safety of ORADUR-Methylphenidate ER (referred to as Methydur Sustained Release Capsules or Methydur) in children and adolescents with ADHD between the ages of 6 and 18 years.  According to Orient Pharma, the study was conducted in Taiwan and enrolled 110 subjects, of which 100 evaluable subjects completed the study.  For the primary efficacy endpoint, Orient Pharma observed a statistically significant difference between Methydur and Placebo treatments in the mean change of total score for the Swanson, Nolan, and Pelham-IV (SNAP-IV) teacher form (p=0.0044 for the intent to treat population and p=0.0104 for the per protocol population).  Orient Pharma’s safety analysis indicates that the incidence of adverse events with Methydur was similar to other approved methylphenidate products.

In September 2018, Orient Pharma informed us that it had obtained marketing authorization for Methydur from the Ministry of Health and Welfare in Taiwan.  Methydur is indicated for the treatment of ADHD and will be available in three strengths (22 mg, 33 mg and 44 mg) in Taiwan. Orient Pharma also has stated that it expects to make Methydur commercially available in Taiwan in 2019, while seeking a partner in China and pursuing regulatory approvals in selected other countries in Southeast Asia where it has

28

commercialization rights and a commercialization presence. We are seeking potential development and commercialization partners for ORADUR- Methylphenidate ER Capsules for major markets not licensed to Or ient Pharma.

REMOXY ^® ER

In December 2002, we entered into an agreement with Pain Therapeutics, amended in December 2005, under which we granted Pain Therapeutics the exclusive, worldwide right to develop and commercialize selected long-acting oral opioid products using our ORADUR technology incorporating four specified opioid drugs.

On March 20, 2019, we received written notice from Pain Therapeutics that, effective June 18, 2019, Pain Therapeutics is terminating their agreement with us. As a result of this termination, Pain Therapeutics will have no license under DURECT’s technology and will be returning its exclusive, worldwide commercialization rights to develop and market REMOXY ER to the Company.

Drug Delivery Technologies and Programs

Our drug delivery technologies are designed to deliver the right drug to the right place, in the right amount and at the right time to treat a variety of chronic, acute and episodic diseases and conditions. We aim to improve therapy for a given disease or patient population by controlling the rate and duration of drug administration. In addition, if advantageous for the therapy, our technologies can target the delivery of the drug to its intended site of action.

Our technologies are suitable for providing long-term drug therapy because they can often store highly concentrated, stabilized drugs in a small volume and protect the drug from degradation by the body. This, in combination with the ability to continuously deliver desired doses of a drug, can extend the therapeutic value of a wide variety of drugs, including, in some cases, those which would otherwise be ineffective, too unstable, too potent or cause adverse side effects. In some cases, delivering the drug directly to the intended site of action can improve efficacy while minimizing unwanted side effects elsewhere in the body, which often limit the long-term use of many drugs. Our pharmaceutical systems may thus provide better therapy for chronic diseases or conditions, or for certain acute conditions where longer drug dosing is required or advantageous, by replacing multiple injection therapy or oral dosing, improving drug efficacy, reducing side effects and ensuring dosing compliance. Our technology may thereby improve patients’ quality of life by eliminating more repetitive treatments, reducing dependence on caregivers and allowing patients to lead more independent lives.

We currently have several major active drug delivery technology platforms:

The SABER and CLOUD Bioerodible Injectable Depot Systems

Our bioerodible injectable depot systems include our SABER and CLOUD platform technologies. SABER uses a high viscosity base component, such as sucrose acetate isobutyrate (SAIB), to provide controlled release of a drug. When the high viscosity SAIB is formulated with drug, biocompatible excipients and other additives, the resulting formulation is easily injectable with standard syringes and needles. After injection of a SABER formulation, the excipients diffuse away, leaving a viscous depot which provides controlled sustained release of drug. CLOUD is a class of bioerodible injectable depot technology which generally does not contain SAIB but includes various other release rate modifying excipients and/or bioerodible polymers to achieve the delivery of drugs for periods of days to months from a single injection. We are researching and developing a variety of controlled-release products based on the SABER and CLOUD technologies. Based on research and development work to date, our bioerodible injectable depot technologies have shown the following advantages:

29

The SABER technology is the basis of POSIMIR (described above). The SABER technology is also utilized in our ophthalmic program with Santen Pharmaceutical Co., Ltd. (Santen), as well as multiple feasibility programs. In our clinical studies thus far, our bioerodible injectable depot formulations have been observed to be generally safe and well-tolerated.

The SABER technology is also the basis for SucroMate™ Equine, an injectable animal health drug utilizing our SABER technology to deliver the peptide deslorelin. This was the first FDA approved SABER injectable product when it was launched in 2011 by CreoSalus, Inc.

The ORADUR Sustained Release Gel Cap Technology

We believe that our ORADUR sustained release technology can transform short-acting oral capsule dosage forms into sustained release oral products. Products based on our ORADUR technology can take the form of an easy to swallow capsule that uses a high-viscosity base component such as sucrose acetate isobutyrate (SAIB) to provide controlled release of active ingredients for an extended period of time. Oral dosage forms based on the ORADUR gel-cap may also have the added benefit of being less prone to abuse (e.g., by crushing and then snorting, smoking, injecting or extracting by mixing with alcohol or water). These properties have the potential to make ORADUR-based products an attractive option for pharmaceutical companies that seek to develop abuse-deterrent oral products.

The ORADUR technology is the basis of our ORADUR-Methylphenidate ER and REMOXY ER programs (described above).

Depot Injectable Programs

In addition to biologic drugs, many traditional small molecule drugs have to be given by frequent injections, which is costly, inconvenient and may result in either unwanted side effects or suboptimal efficacy. We have active programs underway to improve our depot injectable systems and to apply those systems to various drugs and drug candidates, and have entered into a number of feasibility studies with biotechnology and pharmaceutical companies to test their products in our systems. The ophthalmic program with Santen is a project which started as a depot injectable feasibility project and then matured into a development and license agreement.

Product Revenues

We also currently generate product revenue from the sale of the following three products:

Because we consider our core business to be developing and commercializing pharmaceuticals, we do not intend to significantly increase our investments in or efforts to sell or market any of our existing product lines. However, we expect that we will continue to make efforts to increase our revenue related to collaborative research and development by entering into additional research and development agreements with third-party collaborators to develop product candidates based on our drug delivery technologies.

Operating Results

Since our inception in 1998, we have had a history of operating losses. At March 31, 2019, we had an accumulated deficit of $475.8 million. Our net loss was $7.1 million for the three months ended March 31, 2019. Our net losses were $25.3 million and $3.7 million for the years ended December 31, 2018 and 2017, respectively. These losses have resulted primarily from costs incurred to research and develop our product candidates and to a lesser extent, from selling, general and administrative costs associated with our operations and product sales. We expect our research and development expenses in the near future to increase compared to the first quarter of 2019 as we experience higher research and development expenses related to DUR-928. We expect selling, general and administrative expenses in the near future to be lower compared to the first quarter of 2019 due to lower patent related expenses. We do not anticipate meaningful revenues from our products in development, should they be approved, for at least the next twelve months. Therefore, we expect to incur continuing losses and negative cash flows from operations for the foreseeable future.

30

Depen ding on whether we enter into additional collaborative agreements in the near term and the extent to which we earn milestone revenues, we may be required to raise additional capital through a variety of sources .

Critical Accounting Policies and Estimates

The preparation of financial statements in conformity with generally accepted accounting principles requires management to make estimates and assumptions that affect the reported amounts of assets and liabilities and the disclosure of contingent assets and liabilities at the dates of the financial statements and the reported amounts of revenues and expenses during the reporting periods. The most significant estimates and assumptions relate to revenue recognition, the recoverability of our long-lived assets, including goodwill and other intangible assets, accrued liabilities, contract research liabilities, inventories and stock-based compensation. Actual amounts could differ significantly from these estimates. There have been no material changes to our other critical accounting policies and estimates as compared to the disclosures in our annual report on Form 10-K for the year ended December 31, 2018.

Results of Operations

Three months ended March 31, 2019 and 2018

Collaborative research and development and other revenue

We recognize revenues from collaborative research and development activities and service contracts. Collaborative research and development revenue primarily represents reimbursement of qualified expenses related to collaborative agreements with various third parties to research, develop and commercialize potential products using our drug delivery technologies, and revenue recognized from ratable recognition of upfront fees and milestone payments in connection with our collaborative agreements.

We expect our collaborative research and development and other revenue in the near future to be comparable to the first quarter of 2019. We expect our collaborative research and development and other revenue to fluctuate in future periods pending our efforts to enter into potential new collaborations, our existing third party collaborators’ commitment to and progress in the research and development programs, and any royalty or earn-out revenue recognized from collaborators or counterparties. The collaborative research and development and other revenues associated with our major collaborators or counterparties are as follows (in thousands):

Product revenue

A portion of our revenues is derived from product sales, which include our ALZET mini pump product line, our LACTEL biodegradable polymer product line and certain excipients that are included in Methydur and in a currently marketed animal health product. Net product revenues were $2.6 million and $2.4 million in the three months ended March 31, 2019 and 2018, respectively. The increase in the three months ended March 31, 2019 was primarily attributable to higher revenue from our LACTEL product line as a result of higher units sold, partially offset by lower revenue from our ALZET mini pump product line as a result of lower units sold compared to the corresponding period in 2018.

Cost of product revenues

31

Cost of p roduct revenues w as $ 1.1 million and $1.2 million for the three months ended March 31, 2019 and 2018 , respectively. The decrease in the cost of product revenue in the three months ended March 31, 2019 were primarily the result of lower cost of goods sold related to our LACTEL product line arising from lower manufacturing costs for units sold as well as lower cost of goods sold related to our ALZET product line arising from lower units sold compared to the corresponding period in 2018 . Cost of product revenues and gross profit margin will fluctuate from period to period depending upon the product mix in a particular period and unit volumes sold. Stock-based compensation expense recognized related to cost of product revenues was $2 1 ,000 and $25,000 for the three months ended March 31, 2019 and 2018, respectively .

We had 22 manufacturing employees as of March 31, 2019 compared with 21 as of March 31, 2018. We expect the number of employees involved in manufacturing will remain comparable in the near future.

Research and development

Research and development expenses are primarily comprised of salaries, benefits, stock-based compensation and other compensation cost associated with research and development personnel, overhead and facility costs, preclinical and non-clinical development costs, clinical trial and related clinical manufacturing costs, contract services, and other outside costs.

Research and development expenses were $6.3 million and $7.0 million for the three months ended March 31, 2019 and 2018, respectively. The decrease in the three months ended March 31, 2019 was primarily attributable to lower research and development costs associated with POSIMIR, ORADUR-ADHD, REMOXY ER and other research programs, partially offset by higher research and development costs associated with depot injectable programs, DUR-928 and the Santen ophthalmic program compared to the corresponding period in 2018, as more fully discussed below. Stock-based compensation expense recognized related to research and development personnel was $174,000 and $353,000 for the three months ended March 31, 2019 and 2018, respectively. As of March 31, 2019, we had 44 research and development employees compared with 48 as of March 31, 2018. We expect research and development expenses in the near future to increase compared to the first quarter of 2019 as we expect to incur higher research and development expenses for DUR-928.

Research and development expenses associated with our major development programs approximate the following (in thousands):

DUR-928

Our research and development expenses for DUR-928 were $4.2 million in the three months ended March 31, 2019 compared to $4.1 million for the corresponding period in 2018. The increase in the three months ended March 31, 2019 was primarily due to higher clinical trial expenses and higher employee-related expenses incurred for this drug candidate compared with the corresponding period in 2018.

Depot injectable programs

Our research and development expenses for depot injectable programs were $1.3 million in the three months ended March 31, 2019 compared to $973,000 for the corresponding period in 2018. The increase in the three months ended March 31, 2019 was primarily due to higher employee-related costs for these programs compared with the corresponding period in 2018.

POSIMIR

Our research and development expenses for POSIMIR were $722,000 in the three months ended March 31, 2019 compared to $1.7 million for the corresponding period in 2018. The decrease in the three months ended March 31, 2019 was primarily due to lower outside expenses and lower employee-related expenses for POSIMIR compared with the corresponding period in 2018.

32

ORADUR-ADHD

Our research and development expenses for ORADUR-ADHD were $17,000 in the three months ended March 31, 2019 compared to $32,000 for the corresponding period in 2018. The decrease in the three months ended March 31, 2019 was primarily due to lower employee-related costs for this drug candidate compared with the corresponding period in 2018.

Santen ophthalmic program

Our research and development expenses for the Santen ophthalmic program were $12,000 in the three months ended March 31, 2019 compared to $8,000 for the corresponding period in 2018. The increase in the three months ended March 31, 2019 was primarily due to higher employee-related costs associated with this drug candidate compared with the corresponding period in 2018.

REMOXY ER

Our research and development expenses for REMOXY ER were $2,000 in the three months ended March 31, 2019 compared to $4,000 for the corresponding period in 2018. The decreases in the three months ended March 31, 2019 was primarily due to lower employee-related costs for REMOXY ER compared with the corresponding period in 2018.

Other DURECT research programs

Our research and development expenses for all other programs were $37,000 in the three months ended March 31, 2019 compared to $109,000 for the corresponding period in 2018. The decrease in the three months ended March 31, 2019 was primarily due to lower employee-related costs incurred as well as lower outside expenses associated with these programs compared with the corresponding period in 2018.

We expect our research and development expenses in the near future to increase compared to the first quarter of 2019 as we expect to incur higher research and development expenses for DUR-928. The duration of development of our research and development programs may span as many as ten years or more, and estimation of completion dates or costs to complete are speculative and subjective due to the numerous risks and uncertainties associated with developing pharmaceutical products, including significant and changing government regulation, the uncertainties of future preclinical and clinical study results, the uncertainties with our collaborators’ commitment to and progress in the programs and the uncertainties associated with process development and manufacturing as well as sales and marketing. In addition, with respect to our development programs subject to third-party collaborations, the timing and expenditures to complete the programs are subject to the control of our collaborators. Therefore, we cannot reasonably estimate the timing and estimated costs of the efforts necessary to complete the research and development programs. For additional information regarding these risks and uncertainties, see “Risk Factors” below.

Selling, general and administrative. Selling, general and administrative expenses are primarily comprised of salaries, benefits, stock-based compensation and other compensation cost associated with finance, legal, business development, sales and marketing and other administrative personnel, overhead and facility costs, and other general and administrative costs.

Selling, general and administrative expenses were $3.5 million for the three months ended March 31, 2019 compared to $3.2 million for the corresponding period in 2018. The increase in selling, general and administrative expenses in the three months ended March 31, 2019 was primarily due to higher patent related expenses compared with the corresponding period in 2018. Stock-based compensation expense recognized related to selling, general and administrative personnel was $242,000 and $283,000 for the three months ended March 31, 2019 and March 31, 2018, respectively.

We had 22 selling, general and administrative employees as of March 31, 2019 compared with 23 as of March 31, 2018. We expect selling, general and administrative expenses in the near future to be lower compared to the first quarter of 2019 due to lower patent related expenses.

Other income (expense). Interest and other income was $209,000 for the three months ended March 31, 2019 compared to $158,000 for the corresponding period in 2018. The increase in interest and other income in the three months ended March 31, 2019 was primarily the result of higher interest income generated from our investments as a result of higher yields in the three months ended March 31, 2019 compared with the corresponding period in 2018.  

Interest and other expense was $629,000 for the three months ended March 31, 2019 compared to $623,000 for the corresponding period in 2018. The increase in interest and other expense in the three months ended March 31, 2019 was primarily due to higher interest rates associated with the term loan compared with the corresponding period in 2018.

Liquidity and Capital Resources

33

We had cash, cash equivalents and investments totaling $ 28. 8  million at March 31, 2019 compared to $ 34.5 million at December 31, 201 8 . These balances include $150,000 of interest- bearing marketable securities classified as restricted investments on our balance sheets a s of March 31, 2019 and December 31, 201 8 . The de crease in cash, cash equivalents and investments during the three months ended March 31, 2019 was primarily the result of ongoing operating expenses and interest payments, partially offset by payments receiv ed from collaboration partners and customers and cash received from the sale of our common stock .

We used $5.7 million of cash in operating activities in the three months ended March 31, 2019 compared to $6.7 million received for the corresponding period in 2018. The cash used for operations was primarily to fund operations as well as our working capital requirements which involved a decrease in net loss of $1.2 million, partially offset by the changes in account receivable, prepaid expenses and other assets, and accrued and other liabilities.

We received $1.6 million of cash from investing activities for the three months ended March 31, 2019 compared to $2.5 million for the corresponding period in 2018. The decrease in cash received from investing activities was primarily due to a decrease in proceeds from maturities of available-for-sale securities for the three months ended March 31, 2019 compared to the corresponding period in 2018. We anticipate incurring capital expenditures of approximately $100,000 in 2019 to purchase research and development and other capital equipment.

We received $57,000 of cash from financing activities for the three months ended March 31, 2019 compared to $14.1 million for the corresponding period in 2018. The decrease in cash received from financing activities was primarily due to lower net proceeds received from issuances of common stock in the three months ended March 31, 2019 compared with the corresponding period in 2018 . During the three months ended March 31, 2019, we raised net proceeds (net of commission) of approximately $184,000 from the sale of 243,000 shares of common stock at a weighted average price of $0.78 per share in the open market pursuant to the October 2018 registration statement . The net proceeds were further reduced by net issuance costs of approximately $123,000.

We anticipate that cash used in operating activities in the near future will increase compared to the first quarter of 2019 due to higher research and development expenses related to DUR-928.

In November 2015, we entered into the 2015 Sales Agreement with Cantor Fitzgerald, acting as agent. As of December 31, 2018, approximately 24.6 million registered shares of common stock were sold through the 2015 Sales Agreement for total net proceeds (net of commissions) of approximately $38.3 million.  In August 2018, we filed a shelf registration statement on Form S-3 with the SEC, which, upon being declared effective in October 2018, allowed us to offer up to $175.0 million of securities from time to time in one or more public offerings, inclusive of up to $75.0 million of additional shares of common stock which we may sell, subject to certain limitations, under the 2015 Sales Agreement through Cantor Fitzgerald, acting as agent. As of May 3, 2019, we raised net proceeds (net of commissions) of approximately $184,000 from the sale of 242,750 shares of the Company’s common stock in the open market at a weighted average price of $0.78 per share pursuant to the October 2018 registration statement. Any material sales in the public market of our common stock, under the Controlled Equity Offering program with Cantor Fitzgerald or otherwise under the October 2018 shelf registration statement, could adversely affect prevailing market prices for our common stock.

During the three months ended March 31, 2019, there were no significant changes in our commercial commitments and contractual obligations as compared with the information presented in our Annual Report on Form 10-K for the year ended December 31, 2018.

We believe that our existing cash, cash equivalents and investments and anticipated revenues will be sufficient to fund our planned operations, existing debt and contractual commitments and planned capital expenditures through at least the next 12 months from the date the financial statements are filed. We may consume available resources more rapidly than currently anticipated, resulting in the need for additional funding. Additionally, we do not expect to generate significant revenues from our pharmaceutical products currently under development for at least the next twelve months, if at all. Depending on whether we enter into additional collaborative agreements in the near term and the extent to which we earn milestone revenues, we may be required to raise additional capital through a variety of sources, including:

There can be no assurance that we will enter into additional collaborative agreements in the near term, will earn milestone revenues or that additional capital will be available on favorable terms, if at all. If adequate funds are not available, we may be required to significantly reduce or refocus our operations or to obtain funds through arrangements that may require us to relinquish rights to certain of our products, technologies or potential markets, either of which could have a material adverse effect on our

34

business, financial condition and results of operations. To the extent that additional capital is raised through the sale of equity or convertible debt securities, the issuance of such securities would result in ownership dilution to our existing stockholders (assu ming convertible debt securities were converted into shares).

Our cash and investments policy emphasizes liquidity and preservation of principal over other portfolio considerations. We select investments that maximize interest income to the extent possible given these two constraints. We satisfy liquidity requirements by investing excess cash in securities with different maturities to match projected cash needs and limit concentration of credit risk by diversifying our investments among a variety of high credit-quality issuers.

Off-Balance Sheet Arrangements

As of March 31, 2019, we did not have any off-balance sheet arrangements, as defined under SEC Regulation S-K Item 303(a)(4)(ii).