Targeting Cancer

June 3, 2015 2 Forward-Looking Statements

These slides and the accompanying oral presentation contain "forward-looking" statements within the meaning of the Private Securities Litigation Reform Act of 1995. These statements may be identified by the use of forward looking terminology such as "anticipate", "believe", "continue", "could", "estimate", "expect", "intend", "may", "might", "plan", "potential", "predict", "should" or "will" and include statements regarding Threshold's product candidates and clinical trial progress and results. These forward-looking statements are based on our current expectations, speak only of the date of this presentation and involve risks and uncertainties, many of which are outside of our control, that can cause actual results to differ materially from those in the forward-looking statements. Potential risks and uncertainties include, but are not limited to, our ability to complete our anticipated clinical trials, the time and expense required to conduct such clinical trials, the ability to manufacture clinical or commercial product, issues arising in the regulatory process and the results of such clinical trials (including product safety issues and efficacy results). Further information is included in Threshold’s periodic reports filed with the SEC at www.sec.gov. We disclaim any duty to update any forward-looking statements.

3 2015 Is a Key Year for Threshold (Nasdaq: THLD)

• We are nearing data readouts for two pivotal Phase 3 trials of evofosfamide (TH-302) – Soft tissue sarcoma and – Events expected to be reached in 2H:15 with data shortly thereafter • We have a global partnership with Merck KGaA for evofosfamide – Brings critical mass to program – Covers 70% of all evofosfamide related expenses – THLD retains co-commercialization rights in the U.S. • We own worldwide rights to TH-4000 – Molecularly-targeted, hypoxia-activated, irreversible EGFR-TKI – Phase 2 clinical trials planned to begin this year • We are the world’s leader in developing therapies that target tumor hypoxia

Tumor Hypoxia The Result of Abnormal Vasculature

Normal Tissue Tumor Tissue Vasculature Vasculature

4 5 The Tumor Microenvironment Subregional hypoxia as a defining feature of solid tumors

Blood Vessels Compilation of studies 2000+ cancer patients Hypoxic region Tumor Tissue Normal Tissue Necrosis Tumor Type Median pO2 Median pO2 mm Hg (# pts) mm Hg Human colon cancer xenograft Pancreas 2 (8) 57 Brain 13 (104) 26 Head & Neck 10 (592) n/a Lung 16 (26) n/a Breast 10 (212) 52 Cervix 9 (730) 42 Liver 6 (4) 30 Prostate 2,5,10,11,21 n/a (57,55,55,10,13) 150µm Pimonidazole staining Sarcoma 14 (283) 51 of hypoxic regions Melanoma 12 (18) 41 Blood vessels in blue Lymphoma 18 (8) n/a

Reviewed in Vaupel, P. et al., Antioxidants & Redox Signaling, 9: 1221, 2007 Minchinton, A. and Tannock, I. Nat. Rev. Cancer. 6: 583, 2006 6 Significance of Tumor Hypoxia Solid tumors

60 years ago tumor hypoxia was identified as underlying resistance to radiation…

Br J Radiol 26: 638, 1953

…tumor hypoxia also underlies resistance to The More Hypoxic, the Worse the Prognosis Four examples from many published studies

Soft Tissue Sarcoma Head(n = 63and pts) Neck Cancer 100 (n = 28 pts) 100 (n = 63 pts)

80 80

pOpO2 2 > _? 1010 mmHgmmHg pOpO2 2 >_? 19 mmHg (n = 14) (n = 19) 60 60

40 40 Survival(%) Survival(%) Survival(%) pOpO2 <2 <10 10 mmHg mmHg pO2 < 19 mmHg (n = 44) 20 (n = 14) 20

0 0 0 12 24 36 48 60 0 12 24 36 48 Time (months) Time (months)

100 Cervical Cancer Multiple Myeloma (n = 89 pts) (n = 42 pts)

80

HIF-1α - 60 (n = 23) pO2 ?>_ 10 10 mmHgmmHg (n(n == 41)41) 40 Survival(%) Survival(%) pO2 < 10 mmHg HIF-1α 20 pO2 < 10 mmHg + (n = 48)(n = 48) (n = 19)

0 0 10 20 30 40 50 60 70 80 Time (months) Time (months)

7 STS, Nordsmark, ’01; H&N, Brizel, ’99; Cerv, Hockel, ‘96; Myeloma, Ria ‘14 8 Reactive Chemotherapy Does Not Penetrate Tumors Targets Cells Adjacent to Blood Vessels

Vessels: Red Doxorubicin: Blue Hypoxia: Green

Mouse Mammary Tumor

Minchinton, AI, Tannock, IF. Nat Rev Cancer. 2006 Aug 6(8):583-92. Tumor Hypoxia Impact on the ‘Hallmarks of Cancer’ and promotion of more malignant phenotype

Hypoxia leads to induction of Hypoxic regions of tumors the glycolytic phenotype play an important role in (Warburg Effect) the formation, function, Semenza, Curr Opin Genet Dev 2010; Kroemer, Cancer Cell 2008; Zeng, Cancer Lett 2014; Masson, Cancer Metab 2014; Semenza, J Clin Invest 2013 and recruitment of myeloid-derived Hypoxia inhibits apoptosis suppressor cells (MDSC) and regulatory T-cells (programmed cell death), Corzo, J Exp Med 2010; Sceneay, Oncoimmunol 2013; Tartour, Cancer Metastasis Rev 2011; Yotnda, Methods Mol Biol 2010 leading to resistance to chemotherapy and radiation Graeber, Nature 1996; Erler, MCB 2004; Weinmann, FASEB J 2004; Koshikawa ,Oncogene 2006 Hypoxia regulates pro- Hypoxia induces both genetic inflammatory process in the (mutations) and epigenetic tumor microenvironment Holger, N Engl J Med 2011; Shay, Semin Cell Dev Biol 2012; Mamlouk, Int J Cancer 2013; (chromatin) modifications Finger, Cancer Metastasis Rev 2010; Becker, Cancer Immunol Immunother 2013 Brisow, Nat Rev Cancer 2008; Bindra, Cancer Met Rev 2007; Huang, J Mol Med 2007; Brigati Mediators Inflamm 2010

Hypoxia stimulates Hypoxia induces the angiogenesis through HIF-1α epithelial-to-mesenchymal and VEGF signaling Pugh, Nature Med 2003; Semenza, Crit Rev Biochem 2000; transition (EMT) that allows Yang, Cell Biochem 3013 Hypoxia promotes the cell motility and metastasis Hill, Semin Radiat Oncol 2009; Tsai, J Biomed Sci 2012; Marie-Egyptienne ,Cancer Lett 2013; expression of enzymes that Jiang, Cancer Biol Ther 2011 degrade the extracellular matrix (ECM) and promote invasion Pennacchietti Cancer Cell 2003; Boccaccio Nat Rev Cancer 2006; Finger Cancer Metastasis Rev 2010 9 Hanahan and Weinberg, Cell 144: 646, 2011 10 Evofosfamide Overview

• Evofosfamide is a small molecule prodrug that releases the potent

DNA-alkylating agent, Br-IPM, via reductases under hypoxic conditions.

Evofosfamide (inert prodrug) Br-IPM (active cytotoxin) H C O O N Br Oxygen Gradient O BrCl P - N P N O H O N N 2 HN H Br N H H BrCl

Hypoxia 5% O2 0.5% O2 Trigger Normoxia Hypoxia

• >1500 patients have been treated with evofosfamide in clinical trials to date demonstrating antitumor activity: – across multiple tumor types – as a single agent and in combination – with a consistent safety profile and manageable toxicity

Duan J-X, et al. J Med Chem 2008;51:2412-24120 11 Partial Response in SCLC (the ‘401’ trial)

Baseline Post Cycle 2

Liver 39y ♂ Mets Lung cancer (SCLC)

Third-line Cisplatin/Etoposide Paclitaxel/Avastin

Lung TH-302 480 mg/m2

RECIST status: RECIST status: eight target lesions 44% decrease SLD Monotherapy ~80% volume reduction 12 Partial Response in Metastatic Melanoma (the ‘401’ trial)

Baseline Post Cycle 2

Liver 74y ♂ Mets Melanoma

2nd-Line DTIC

TH-302 2 Lung 670 mg/m Mets

RECIST status: RECIST status: six target lesions 53% decrease SLD ~85% volume reduction Monotherapy 13 Phase 1/2a Monotherapy Trial (the ‘401’ trial) Summary: efficacy (RECIST) and safety

• Overall efficacy signals – >50% of 129 patients with all tumor types achieved SD or better

Historical TH-302 Agent Indication Tumor Tumor Response Response st nd Dacarbazine 1 or 2 line metastatic melanoma 9.8% 22%

Temozolomide 1st or 2nd line metastatic melanoma 14.4% (8/36)

20% 2nd line SCLC refractory (no response Topotecan or PD within 90 days) 2% - 7% (2/10)

25% Erbitux 2nd line Head & Neck 13% (2/8)

• Safety profile with weekly regimen – MTD at 575 mg/m2; skin and mucosal toxicities are dose limiting (dose proportional, predictable, reversible); No dose limiting myelosuppression 14 Evofosfamide (TH-302) Clinical Development

Trial Phase Phase Phase Phase TH-CR-_ _ _ 1 1/2 2b 3 401 / 402 / 403 Initial clinical studies (mono / combo) 404 Pancreatic Cancer (+ ) 406 Soft Tissue Sarcoma (+ doxorubicin) MAESTRO Pancreatic Cancer (+ gemcitabine) 415 n-s NSCLC (+pemetrexed) 413 Melanoma: monotherapy (PETimaging) 414 QT Study: monotherapy EMR-002 Japanese Pancreatic (mono /+ gem) 407 Advanced leukemias: monotherapy 408 Multiple Myeloma (+/- bortezomib) EMR-006 Pancreatic Cancer (+ gem / Abraxane) 410 RCC & GIST (+ sunitinib) IST Glioblastoma (+ bevacizumab) IST Various solid tumors (+ pazopanib) IST HCC (+ sorafenib)

Investigator = ongoing = completed sponsored sponsored sponsored 15 Soft Tissue Sarcoma & Pancreatic Cancer Status of Phase 3 Pivotal Trials Soft Tissue Sarcoma Pancreatic Cancer TH-CR 406: Enrollment of 640 MAESTRO: Enrollment of 660 patients completed December 2013 patients completed November 2014 • 85% power to detect 33% • 90% power to detect 33% improvement (434 events) improvement (508 events) • Partnership with SARC • Metastatic or locally advanced unresectable front-line patients • Special Protocol Assessment (SPA) & Fast Track Designation • Top-Line Data for Both Trials Expected ~EOY 2015

Chawla, et al. J Clin Oncol. 2014 Oct 10;32(29):3299-306. Phase II study of the safety and antitumor activity of the hypoxia-activated prodrug TH-302 in combination with doxorubicin in patients with advanced soft tissue sarcoma.

Borad, et al. J Clin Oncol. 2014 Dec 15 [Epub ahead of print]. Randomized Phase II Trial of Gemcitabine Plus TH-302 Versus Gemcitabine in Patients With Advanced Pancreatic Cancer

http://www.thresholdpharm.com/scientific_publications#evofosfamide 16 Phase 2 Study in Non-Squamous NSCLC (TH-CR-415) Randomize Randomize EVO (400 mg/m2) 2 n-s NSCLC + pemetrexed (500 mg/m ) Second-line (prior TKI monotherapy OK) 1:1 (N=440) pemetrexed (500 mg/m2) + placebo

Primary Overall survival (OS): 321 events (deaths) to assure 85% power to Endpoint detect a 40% improvement in OS (HR=0.715) with a two-sided α of 0.05

• First patient treated June 2014 • Expect enrollment to be completed in 2016 • Phase 1/2 data: Borad et al. ESMO Annual Meeting, October 2010

17 GBM: Current Treatment and Outcomes Phase 1/2 IST of Evofosfamide plus Bevacizumab

Treatment Setting Therapy Median PFS Median OS Newly Diagnosed temozolomide (TMZ) ~7 months ~15 months First-line setting, after + radiation (RT) surgical resection followed by TMZ + RT 1 provides ∆2.5 maintenance TMZ months benefit over RT alone First Recurrence bevacizumab ~4 months2 ~9 months2 Second-line setting Bevacizumab- Another ~38 days3 82 days3

refractory bevacizumab- Salvage setting containing regimen

EVO dose- escalation/expansion in combination with bevacizumab 1. Stupp R. et al N Engl J Med. 2005;352(10):987-996. 2. Friedman HS et al J Clin Oncol. 2009;27:4733-4740. 3. Quant EC et al Neuro-Oncology 2009;11:550-555. 18 Change in Tumor Size and Response Assessment According to Response-Assessment Neuro-oncology (RANO) Criteria

18% ORR (N=22 evaluable)

Brenner AJ, et al. Neuro Oncol (2014) 16 (suppl 5): v10-v11 (Abstract AT-12). 19 Salvage Therapy Outcomes in Recurrent GBM Following Bevacizumab (BEV) Failure

Median PFS, months Treatment Regimen Number Patients ORR (%) (95% CI) 1 BEV + evofosfamide (TH-302) 22 18 2.8 (1.9, 3.9)

2 BEV + carboplatin + irinotecan 25 0 2.3 (1.8, 3.6)

3 Various chemo 19 NR 2.0 (1.2, 3.3)

4 BEV + various chemo 23 0 1.8 (NR)

5 BEV + temsirolimus 13 0 1.8 (NR) BEV + TMZ or 6 23 0 1.8 (1.0, 2.1) BEV + etoposide 7 BEV + chemo 54 (35 GBM) 0 1.3 (1.2, 1.5) 8 BEV + irinotecan 19 0 1.1 (NR) 9 BEV + dasatinib 14 0 0.9 (0.9 - 1.2) 10 MPC-6827 24 4 0.7

NR = not reported 1. Brenner 2014 SNO 6. Reardon 2011 J. Neurooncol 2. Reardon 2011 Cancer 7. Quant 2009 Neuro Oncol 3. Iwamoto 2009 Neurology 8. Kreisl 2009 JCO 4. Norden 2008 Neurology 9. Lu-Emerson 2011 Neurooncol 5. Lassen 2013 Anticancer Res 10. Grimm 2010 SNO

20 Summary Patients with Bevacizumab-Refractory Recurrent Glioblastoma

TH -302 has a manageable safety profile when used in combination with BEV

Early signals of activity were observed (N=22 evaluable patients) • 18% objective response rate (1 CR, 3 PR) • >70% SD or better • Median PFS 2.8 months compares favorably with benchmark of ~1.2 months • Study expanded to include DFCI • Compelling rationale to move to front-line in MGMTmethyl- patients

21 TH-CR-408: Phase 1/2 in Relapsed/Refractory MM Part 1: EVO + Low-dose Dexamethasone

• Dose-escalation/dose-expansion; conducted by Dana-Farber Cancer Institute and the Blood Cancer Research Partnership • Heavily pretreated patients enrolled (median 6.5 prior therapies): bortezomib (100 %); lenalidomide/thalidomide (100%); an alkylator (92%) • EVO 340 mg/m2 established as maximum-tolerated dose (MTD); 31% Clinical Benefit Rate (partial + minimal responses); 88% SD or better

(PR) (MR) (SD)

(PD) (CBR)

Laubach J et al., 2014 ASCO 22 TH-CR-408: Phase 1/2 in Relapsed/Refractory MM Part 2: Evo + Bortezomib + Low-dose Dexamethasone (EBorD)

• A total of 25 patients with relapsed/refractory MM have been enrolled in the EBorD component of the study as of May 1, 2015. • At ASCO 2015, preliminary safety and efficacy analyses from 18 patients were presented. • Patients had received a median of 8 prior systemic therapy regimens including at least one regimen with bortezomib and one regimen with thalidomide/lenalidomide.

Laubach JP et al. ASCO 2015 23 Nadir Percent Change in Paraprotein N = 18 patients (4 at 240 mg/m2 and14 at 340 mg/2)

83% SD or better

340 340 340 340 240 340 340 340 240 340 340 340 240 340 340 340 240 340

Laubach JP et al. ASCO 2015 24 IMWG Best Overall Response

29% Clinical Benefit Rate at RP2D in pts with 8 Prior Therapies

Daratumumab Ph 2 SIRIUS Trial showed 29% CBR, including 1 CR, for pts with 5 prior therapies…filed for registration

IMWG: International Myeloma Working Group Durie B, et al. International uniform response criteria for multiple myeloma. Leukemia 1-7, 2006. Rajkumar SV, et al. Consensus recommendations for the uniform reporting of clinical trials: report of the International Myeloma Workshop Consensus Panel 1. Blood 117: 4691-4695, 2010 RP2D: Recommended Phase 2 Dose

Laubach JP et al. ASCO 2015 25 Certain Cancers are Dependent on EGFR Signaling Current EGFR TKI Therapy Is Not Optimal

Non-squamous NSCLC • NSCLC patients with initial sensitivity to EGFR TKIs eventually develop 1L EGFR mut resistance US/EU ~15% Asia 40-50% EGFR TKI • EGFR signaling contributes to tumor

Tarceva () US/EU growth in head and neck cancers Gilotrif (afatinib) US/EU (squamous cell) Iressa (gefitinib) Asia • Presence of wild type EGFR is a factor in resistance to EGFR TKIs

• Approved EGFR inhibitors are unable 2L T790M+ T790M- to optimally silence wild type EGFR at ~50% of pts ~50% of pts tolerable doses

A tumor hypoxia-released inhibitor that TH-4000 • 3rd Gen EGFRi binds both mutant & wild type EGFR AZ, Clovis, Astellas, Novartis, Hanmi should permit targeted activity while Taiho minimizing systemic exposure

1) Franovic et al, 2007, PNAS; 104:13092; 2) Wang et al., 2009, Nature Med, 15:319; 3) 2010, Carcinogenesis, 31:1202; 4) Minakata et al, Cancer Sci. 2012, 103(11), 1946-1954; 5) Taniguchi et al, Cancer Sci. 2008, 99: 929-35; 6) Yun et al (2008), Proc Natl Acad Sci;105:2070–2075; 26 EGFR TKIs Perform Worse in EGFR WT-Containing Patients

Taniguchi et al, Cancer Sci. 2008 (n=21) 27 TH-4000 Is a Novel, Hypoxia-Activated, pan-ErbB Inhibitor Combines molecular targeting and hypoxia targeting in one drug

• TH-4000 is a small molecule prodrug that releases the potent, irreversible Epidermal Growth Factor Receptor (EGFR) Tyrosine Kinase Inhibitor (TKI) via reductases under hypoxic conditions

Comparison of Erlotinib & TH-4000 in Xenograft Models28 Effect of Wild Type EGFR on Efficacy in NSCLC

HCC827 Homozygous Exon19∆ EGFR PC9 Heterozygous Exon19∆ /WT EGFR

Patterson et al., AACR Annual Meeting 2015, Abstract Number: 5358 29 TH-4000 Is More Active than Afatinib Resulting in 100% Tumor Regression

A431 (epidermoid carcinoma) Pure Homozygous WT EGFR

Patterson et al., AACR Annual Meeting 2015, Abstract Number: 5358 30 TH-4000: Phase 1 Results

In a Phase 1 trial of TH-4000 (NCT01631279) a total of 27 patients with advanced solid tumors were treated with weekly 1-hour administration of TH-4000 (dose-escalation from 10 – 200 mg/m2). • Dose-limiting toxicities were facial pain (grade 3) and QT prolongation (grade 4) reported in two patients at 200 mg/m2. • The maximum tolerated dose (MTD) of TH-4000 was established at 150 mg/m2/week(1-hour infusion). • TH-4000 is effective in tumor xenograft models at plasma exposures readily achieved in humans • Tumor targeting may enable effective therapeutic dosing while avoiding the dose-limiting skin rash and GI toxicities associated with currently approved EGFR-TKI therapies

Patterson et al., AACR Annual Meeting 2015, Abstract Number: 5358 31 TH-4000: Next Steps

•Collectively, our data demonstrate that TH-4000 has the potential to address several of the known mechanisms of TKI- resistance in NSCLC –Phase 2 study to commence mid-2015 • Strong clinical rationale supporting therapeutic potential of TH- 4000 in Head & Neck cancer, a disease that has experienced no significant improvements in survival for 20 years –Phase 2 study to commence mid-2015

32 Selected Financial Data

As of March 31, 2015 Cash, cash equivalents, and marketable securities ~$83.1 M Shares outstanding 71.3M Exercisable warrants (@ $2.46) 4.0 M Exercisable warrants (@ $10.86) 8.3M

Merck KGaA Contractual Milestones Received as of 6/30/2014 $110 M Additional development and regulatory milestones $100 M Commercial milestones $340 M TOTAL $550 M 33 Upcoming Milestones

• Mid’15: First patients expected to be enrolled in Phase 2 studies of TH-4000 in patients with NSCLC and SCCHN • EOY: Top-line data expected from pivotal trials in: – Soft Tissue Sarcoma – Pancreatic Cancer • 1H‘16: POC data from TH-4000 in NSCLC and SCCHN • 2016: Completion of enrollment in Phase 2b Evofosfamide study in patients with 2L NSCLC 34 Conclusions

• We are nearing data readouts for two pivotal Phase 3 trials of evofosfamide (TH-302) – If positive, tumor hypoxia will be validated as target for therapeutic intervention and THLD dominates this space • We have a deep pipeline of ongoing and planned studies with evofosfamide in other indications – NSCLC, Multiple Myeloma, Glioblastoma, HCC – Other indications to be initiated • Our partnership with Merck is strong • We own worldwide rights to TH-4000 – Molecularly-targeted, hypoxia-activated, irreversible EGFR-TKI – Phase 2 clinical trials planned to begin this year • We are planning for success