…and a ray of light: genasense

www.bearstearns.com MAY 9, 2003

European Equity Research Pharmaceuticals

Stock Rating Outperform Sector Rating Market Overweight Target Price €60

Valuation Method Price / Pre-amortisation EPS 52-Week Range €80-€38 EPS 2002: €2.64 2003E: €2.82 2004E: €3.56

Post Options EPS 2002: €2.43 2003E: €2.60 2004E: €3.36

P/E 2003E: 16.7x 2004E: 13.2x Dividend (2003E) €0.7 Yield (2003E) 1.5% Common Shares (m) 788m Equity Market Capitalisation (m) €36,800 Book Value per Share (2003E) €16 02–05E EPS CAGR 16.1%

Alexandra Hauber

Eric Hoffman, PhD

Regulated by the United Kingdom Financial Service Authority

U.S. investors contact your Bear, Stearns & Co. Inc. registered representative at 1-800-999-2000

…And a Ray of Light: Genasense ��NEAR-TERM OPPORTUNITY ALMOST IGNORED. Despite its possible

launch in first-quarter 2004, investors in Aventis have not yet become excited about Genasense, a new anti-cancer antisense compound licensed from Genta. They are sceptical because: (1) antisense is still unproven technology; (2) Bcl-2 has never been targeted before; (3) there have been recent setbacks for other targeted therapy approaches; and (4) there are higher regulatory hurdles for combination therapy than monotherapy.

��BCL-2 SET TO BE VALIDATED THIRD-QUARTER 2003. We will learn this summer if targeting Bcl-2 enhances the efficacy of chemotherapy, when headline data from at least one of the three concurrent registration trials is released. Positive data would allow a first-quarter 2004 launch and, importantly, would validate Bcl-2 as a target for much larger, solid tumour indications.

��CLOSER TO GLIVEC/HERCEPTIN THAN EGFR/VEGF THERAPIES. We have developed a framework that could help differentiate various targeted therapies. We conclude that Genasense is closer to Glivec or Herceptin than to Erbitux or Avastin, and thus should have a higher chance of clinical benefit. Genta may avoid the clinical regulatory problems often associated with combination therapies, as all three registration trials are randomised and controlled.

��WE REITERATE OUR OUTPERFORM RATING. We estimate a Genasense market potential in the three initial indications of more than US$500 million. Assuming it works in solid tumours, Genasense could be a multi-billion dollar compound. Investors’ concerns about Allegra, Lovenox and the potential cash drain from non-core businesses seem to be overly discounted in Aventis’ current share price. In our view, investors are not paying for the high-risk, but potentially high-reward, Genasense.

Please read the important disclosure information on the last page of this report.

(AVEP.PA-€44.63;AVE-$51.00)

AVENTIS (AVEP.PA-€44.63) Company Description: Aventis Group generated sales in 2002 of €20,622m. It has broken the back of its disposal plan in 2001, following the disposal of Aventis Crop Sciences and Aventis Animal Nutrition. The largest business is pharma, which generated €16,026m in 2002. The largest products were Allegra (2002 sales of €2,032m), Lovenox (2002 sales of €1,563m) and Taxotere (2002 sales of €1,261m). It also has a significant Vaccines business (2002 sales of €1,580m) and a therapeutic proteins business (2002 sales of €1,068m), which is slated for disposal.

Key Upcoming Events/Developments: �� Week of 28 July 2003: Second-quarter 2003

results.

�� 3Q03: Release of headline data from first phase 3 Genasense (cancer) data.

�� 3Q03: Filing of insulin Glulusine (short-acting insulin analogue).

�� 3Q03: Filing of Alvesco (asthma) in the United States.

�� 2H03: Potential filing of Genasense (cancer) in the United States.

CAPITALISATION (31/12/02) 2002A %

Interest-Bearing Debt 4,752 32 Pension Provisions 3,328 22 Other Provisions 4,702 37 Cash & MS (1,299) 9 Net Debt and Provisions 12,350 100

KEY FINANCIAL RATIOS 2002A 2003E EBITDA Margin 24.7% 30.9% EBITA Margin 9.0% 19.7% Return on Book Equity 18.4% 17.6%

Source: FactSet Research Systems Inc

VALUATION (DEC 2003E) INTERIM EARNINGS PER SHARE 2003E 2004E 2002A 2003E

Price / Earnings 16.7x 13.2x Q1 €0.53 €0.25 Price / Pre-Amortisation Earnings 13.6x 11.2x Q2 €1.34 €0.85 EV/EBITDA 8.7x 7.6x Q3 €0.72 €0.88 EV/ Free Cashflows 20.5x 13.6x Q4 €0.04 €0.84

FY €2.64 €2.82

FINANCIAL SUMMARY (€ in millions, except per share data)

FreeYear- Gross SG&A Operating Net Diluted Cash FlowEnd Revenue Income Margin Expenses % Sales Income Margin Income EPS EBITDA per Share 2008E 23,517 17,300 73.6% (6,975) -29.7% 6,519 27.7% 4,115 5.52 7,909 5.81 2007E 22,365 16,393 73.3% (6,693) -29.9% 6,084 27.2% 3,744 4.95 7,445 5.18 2006E 21,225 15,496 73.0% (6,377) -30.0% 5,690 26.8% 3,404 4.45 7,030 4.65 2005E 20,277 14,789 72.9% (6,024) -29.7% 5,475 27.0% 3,192 4.12 6,803 4.27 2004E 19,506 14,329 73.5% (6,121) -31.4% 4,950 25.4% 2,782 3.56 6,271 3.71 2003E 18,449 13,484 73.1% (5,871) -31.8% 4,494 24.4% 2,225 2.82 5,706 (0.04)2002A 20,622 14,044 68.1% (6,773) -32.8% 3,851 18.7% 2,092 2.64 5,096 6.43

Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar AprMay

40

50

60

70

80

15.6

31.2

EUR

Thousands

5/7/02 to 5/8/03541683EUR3.82AVENTIS SA EUR3.82

46.40Last: 38.00Low: 78.20High:

European Equity Research

Table of Contents

…And a Ray of Light: Genasense ............................................................................................................................. 5

Investment Positives and Risks.................................................................................................................................. 8

Introduction to Genasense ....................................................................................................................................... 11

The Development Programme ................................................................................................................................. 14

Headline Data Available Between 21 June and 21 September ...................................................................... 14

Three Potential Registration Trials ................................................................................................................ 15

Aventis’ Solid Tumour Trials ........................................................................................................................ 17

Potential in Indications Outside Cancer......................................................................................................... 17

Safety Database.............................................................................................................................................. 17

Route of Administration................................................................................................................................. 17

The Development Challenge.................................................................................................................................... 19

Targeted Therapy – The Promise................................................................................................................... 19

Not All Targeted Therapies Are the Same..................................................................................................... 20

Concerns in Combination Trials .................................................................................................................... 23

Powering Up the Melanoma Trial.................................................................................................................. 24

Genasense Mechanism and Antisense Technology Review.................................................................................... 26

The Target: Bcl-2 ..................................................................................................................................................... 28

Bcl-2, A Novel Specific Protein Target ......................................................................................................... 28

Apoptosis Activation, A Novel Class Of Target............................................................................................ 30

Competition – Bcl-2 Inhibitors ...................................................................................................................... 31

Competition – Apoptosis Activators.............................................................................................................. 32

Aventis/Genta Licensing Agreement ....................................................................................................................... 33

Appendix 1: Genasense Sales Projections ............................................................................................................... 37

Appendix 2: Genasense Patents and Intellectual Property ...................................................................................... 40

Appendix 3: Antisense and the Case of Affinitak.................................................................................................... 41

Summary Financials................................................................................................................................................. 44

Pipeline Summary.................................................................................................................................................... 48


BEAR, STEARNS INTERNATIONAL LTD.

… And a Ray of Light: Genasense

Figure 1. Aventis’ Share Price Evolution Since June 1998 (€)

Figure 2. Aventis – Share Price Evolution Since June 2001 (€)

30

40

50

60

70

80

90

100

Jun-

98Oc

t-98

Feb-

99Ju

n-99

Oct-9

9Fe

b-00

Jun-

00Oc

t-00

Feb-

01Ju

n-01

Oct-0

1Fe

b-02

Jun-

02Oc

t-02

Feb-

03

peak = € 94.329/06/01

Merger Concluded 15/12/99

30

40

50

60

70

80

90

100

Jun-

01

Sep-

01

Dec-0

1

Mar-0

2

Jun-

02

Sep-

02

Dec-0

2

Mar-0

3

DJ Stox x

Av entis-17%

Source: Datastream.

Executive Summary

In our view, Aventis is a classic case for value investing. The company is cheap, in our opinion, not only on P/E multiples, but also on all cashflow-based valuation metrics which take potential liabilities from Aventis’ conglomerate past into account (see our report from 8 May 2003, ‘On the Dark Side of the Moon’). The problem with value investing is the fact that it might take time until this value is unlocked, particularly in the absence of a catalyst.

In this report we have identified a near-term catalyst: this summer, Genta and Aventis are expected to provide headline data for at least one of the three concurrently conducted registration trials for Genasense. Genasense is a new anti-cancer compound based on antisense technology, licensed from U.S. biotech company Genta in May 2002.

Enrolment in the three registration trials for the indications of malignant melanoma, multiple myeloma and CLL (chronic lymphatic leukaemia) was completed in first-


AVENTIS: AND A RAY OF LIGHT: GENASENSE

quarter 2003 (January for melanoma and myeloma, March for CLL). Genta will see the data from all three trials at roughly the same time and, after initial review, will focus analysis and preparation for registration on the most promising trial, in order to ensure fast submission. Positive data, which would be released by the companies in ‘summer’ (i.e. 21 June – 21 September 2003) in at least one of those trials would allow submission to the FDA soon after, in third-quarter 2003. In the case that the data were positive in either malignant melanoma or multiple myeloma – for which Aventis/Genta have been awarded fast-track status – expedited review and a launch in first-quarter 2004 might be possible.

Investors in Aventis have not yet become excited about Genasense, despite the fact that there will be newsflow within the next four months. There are at least four reasons why investors are sceptical at this stage:

1) As an antisense compound, Genasense is based on little proven technology;

2) The target Bcl-2 has never been targeted before;

3) After recent setbacks with other compounds, targeted therapy seems to be harder to realise than originally thought; and

4) Genasense is developed in combination with chemotherapy, for which the regulatory hurdle for approval seems higher than for monotherapy approval.

In this report we show the market potential of Genasense, and address the concerns of the investment community regarding the technological and regulatory hurdles that Aventis/ Genta will have to overcome.

Market potential. We estimate that the market potential of Genasense in the three initial ‘proof of concept’ indications is over US$500 million. However, assuming it works as well in solid tumours, Genasense could be a multi-billion dollar compound. Therefore, the publication of the headline data would allow an early filing and launch of Genasense. More importantly, we view the registration trials in the three ‘niche’ indications as ‘proof of concept trials’, which could validate Bcl-2 as a target to enhance the efficacy of chemotherapy in much larger solid tumour indications.

Antisense technology risk. The risk of antisense as a technology with one product (Vitravene from Isis for treatment of cytomegalovirus [CMV] retinitis in AIDS patients intolerant to other therapies) on the market is decreasing. Unfortunately, the recent failure of another antisense compound, ISIS/LLY’s Affinitak in phase III, has not helped to boost investors’ confidence in antisense technology. However, in our view, the failure seems primarily related to the target (PKC), not the antisense technology. It is also possible that the Affinitak study was simply underpowered, as indicated by the fact that a secondary endpoint was met, and also supported by subset analysis.

Risk of an unproven target. Bcl-2 is well understood from a biological viewpoint, and its central role in resisting apoptosis has been confirmed. However, the ultimate



proof is in the data to be released this summer. We have to point out that it is possible that the trials might fail to demonstrate a clear clinical benefit (because the trials might turn out not large enough, for example), which could cause a delay.

Risk from concept of targeted therapy. After recent setbacks with other compounds (Erbitux/Iressa/Avastin), targeted therapy once hailed as the ‘Midas touch’ seems to be harder to realise than originally thought. Initial clinical success with compounds such as Glivec from Novartis, or Herceptin from Genentech/Roche could not be replicated with compounds such as Erbitux or Iressa (targeting the EGF receptor) or Avastin (to block VEGF).

We have developed a framework that could help differentiae between various targeted therapies. Based on this framework, we conclude that Genasense is closer to Glivec or Herceptin than to Erbitux or Avastin, and consequently should have a higher chance of proving clinical benefit.

Development challenge of combination therapy. Genta might avoid clinical regulatory problems often associated with development of combination therapies, as it has set up all three of the registration trials as randomised trials with a control arm.

Continuous infusion regime a minor burden. Genasense is given as a continuous intravenous infusion using a portable pump (about the size of a VCR cassette which can be attached to a belt), typically over the period of five to seven days at the same time as chemotherapy, followed by at least two weeks off. This cycle may be repeated up to 12 times, depending on the indication. This continuous infusion via an ambulatory pump represents some additional burden for patients, as most chemotherapy regimens include single-day outpatient infusions, but one that we believe would be tolerated if it provides additional efficacy.

Investment Summary

In our view, Genasense is a high-risk, but potentially high-reward opportunity, for which investors are currently not paying. If data on at least one of the three registration trials, expected to be released by the companies in ‘summer’ (i.e. 21 June – 21 September 2003) turns out to be positive, we would expect to see significant upside in the share price.

We reiterate our Outperform rating, and our €60 price target (upside 34%).



Investment Positives and Risks

Investment Positives – Reasons to Buy

Strong earnings growth driven by margin expansion. A key part of Aventis’ investment thesis is the capacity for margin expansion, despite OTC and generic competition to its largest product, Allegra, in the U.S. market. We forecast Aventis’ EBITA margin will expand from 26% in 2002 to 29% in 2006E. This is driven by the core blockbusters comprising an increasing proportion of sales, as well as a cost containment programme, which management formally announced at the FY2002 analysts meeting.

Strong product management capabilities. Aventis has undoubtedly been extremely successful from a product management perspective. As an example, it has made Lovenox the anti-coagulant of choice, by a combination of generating critical mass of clinical data and pricing strategies. It has also beaten the market’s expectations on the launch of Lantus, the long-acting insulin analogue, since its U.S. launch in second-quarter 2001.

New product opportunities not priced in. Most investors take a cautious view on Aventis’ pipeline, but we are likely to see the filing of three NCEs in 2003: HMR 1964 (rapid-acting insulin analogue), Alvesco, for which Aventis has the important U.S. rights, and Genasense (cancer), as well as the launch of an NCE in the U.S. – Ketek (antibiotic).

Investment Risks

OTC/generic concerns surrounding its largest product, Allegra. OTC versions of Claritin, the key competitor to Allegra in the U.S. market, were launched in December 2002. This caused sales of Rx versions of Claritin to fall by over 70%. However, Allegra continues to grow on a year-on-year basis, in terms of TRx and NRx. Aventis is still confident that it can continue to grow Allegra in the U.S. market, despite market concerns.

Two sets of concerns surrounding Lovenox. Lovenox is one of Aventis’ key growth drivers, with sales of €1,563 million in 2002. There are worries about possible generic competition and also risks to the product’s growth rate:

��Investors became concerned about Lovenox generics when a paragraph IV challenge was erroneously listed on the FDA website in early April 2003. It was subsequently removed, but speculation remains that the FDA may be reviewing a potential generic filing and it was only the timing surrounding the filing that was wrong. Aventis’ recent filing of a citizen’s petition against the ‘possibility’ of a Lovenox ANDA and its application for the reissuance of the ’618 patent has done nothing to reduce the speculation. However, we are unsure whether there can be true generics in the near term, due to difficulties in establishing bioequivalence tests.



��Investors are concerned about Lovenox’s growth rates following the March 2002 ‘Dear Doctor’ letter, warning against using it in patients undergoing heart valve replacement surgery. Lovenox’s growth has slowed. Aventis has increased the level of sales support behind it and is seeking to emphasise the fact that the side effect was only seen in pregnant women. However, risk does remain about reaching the stated 2003 sales growth target of 15%.

Concerns surrounding strength of late-stage pipeline. There are concerns in the market about Aventis’ top-line growth rate. We acknowledge that the group does have some products that we expect to be launched over the next two years, as outlined previously. However, there is a lack of potential blockbusters, in mid- to late-stage clinical trials, to drive strong top-line growth, given the likely loss of Allegra sales.

Delay in launch of important new product. Ketek (antibiotic) has twice been delayed by the FDA, most recently on 24 January 2003. While the FDA has not asked for any additional trials, we believe it is looking at the overall safety profile with particular emphasis on visual side effects. We still expect the product to be launched in the U.S. in particular, as it is already on the market in key European markets; however, in our view these delays are affecting sentiment to the stock.

Distinction between core and non-core business likely to remain in 2004. Last year Aventis said it would end the distinction between core and non-core business by FY2003. However, it has not yet managed to dispose of Aventis Behring (despite discussions with CSL), so has indicated that the distinction is likely to remain this year. In the current era of concern about financial reporting standards, we believe this will affect investors’ confidence in its financial results.

Investment Thesis

Our analysis of Aventis’ balance sheet reveals significant liabilities, which are related to the company’s conglomerate past. However, we believe these liabilities are more than reflected in the current share price.

In our view, the simplest way to capture the valuation risk from Aventis’ cash exposure from its legacy business is to declare all provisions and (quantified) contingencies as debt and deduct it from Aventis’ enterprise value. (Pessimists might want to throw in some extra debt as a ‘safety margin’ in the case that Aventis hasn’t adequately provisioned for its liabilities – we have contemplated this move but decided against it, as even the liabilities reflected on the balance sheet in the form of provisions may never materialise.)

Such a kitchen sink approach still reveals that Aventis is good value on all multiples – including CF-based ones, as shown in the tables below. At our price target of €60, Aventis trades at a 10% discount to the sector based on 2003/04 price to pre-amortisation EPS. This discount is appropriate, in our view, given concerns about Aventis’ top-line growth rate.



Figure 3. Price / Pre-Amortisation EPS 2002A 2003E 2004E 2005EAstraZeneca 22.6 22.2 19.1 16.0 Aventis 14.1 13.6 11.2 9.9 GlaxoSmithKline 17.0 16.6 16.1 15.3Novartis 20.3 19.4 16.0 13.0 Roche 17.7 18.6 15.7 13.8 Sanofi-Synthélabo 20.5 18.0 15.3 13.3 Average 18.9 18.4 16.0 13.9 Source: Company data; Bear, Stearns International Limited estimates.

Figure 4. EV / EBITDA Multiples – Pharma Sector 2002A 2003E 2004E 2005EAstraZeneca 13.8 14.4 12.2 10.0 Aventis 10.7 8.7 7.6 6.7 GlaxoSmithKline 10.8 10.5 10.1 9.5Novartis 16.5 15.0 13.3 11.0 Roche 15.0 13.4 11.5 10.2 Sanofi-Synthélabo 14.3 11.7 9.6 7.9 Average 13.8 12.7 11.2 9.7 Source: Company data; Bear, Stearns International Limited estimates.

Figure 5. EV / EBITDA Multiples – Chemicals Sector 2002A 2003E 2004EAkzo Nobel 7.3 6.8 6.4BASF 7.1 6.3 5.6Bayer 10.1 8.7 7.9Ciba SC 8.2 7.7 7.4Clariant 8.4 7.6 6.9Solvay 5.8 5.3 4.9Syngenta* 7.5 6.8 6.2 Average 7.8 7.0 6.5 Aventis 10.6 8.6 7.6Source: Company data; Bear, Stearns International Limited estimates.

Figure 6. Enterprise Value to Free Cashflow Multiples 2002A 2003E 2004E 2005E 2006EAventis 93.2 20.5 13.6 12.0 11.0 GlaxoSmithKline 19.3 17.9 16.3 15.3 14.1 Sanofi-Synthelabo 39.3 21.3 16.8 13.5 11.1 Roche AG -70.7 135.9 26.4 22.9 19.6Novartis AG 59.4 23.2 18.5 15.4 12.6AstraZeneca 17.7 18.9 16.3 12.9 10.4 Average 26.4 39.6 18.0 15.3 13.1 Source: Company data; Bear, Stearns International Limited estimates.



Introduction to Genasense

Genasense is an antisense therapeutic designed to make cancer cells more susceptible to chemotherapy or other cancer therapy, currently in three separate phase III trials. In more detail, it is an antisense 18 base DNA oligonucleotide that blocks production of Bcl-2, a protein that contributes to the survival of cells generally and the resistance of cancer cells to chemotherapy specifically.

U.S. biotech Genta developed Genasense and licensed it to Aventis in April 2002, when the companies announced they would jointly develop and co-promote Genasense in the United States. Currently in phase III trials for chronic lymphocytic leukaemia (CLL), malignant melanoma and multiple myeloma, we expect headline results from at least one of these trials and an NDA filing on at least one of the indications by summer 2003, defined as 21 June to 21 September. Fast-track status for melanoma and myeloma would allow launch as early as first-quarter 2004. The drug is in 11 non-blinded NCI trials for 10 distinct indications, including several solid tumours representing larger market opportunities, and expansion of the label to these indications could well follow relatively quickly.

The risks surrounding the drug arise because Bcl-2 is a novel target and, more broadly, increasing a cell’s susceptibility to apoptosis is a novel approach. Additionally, antisense has not yet truly been established as a viable technology in the clinic, despite one antisense drug being already approved and marketed1. Further, Isis Pharmaceuticals’ antisense compound Affinitak recently failed in a phase III trial treating non-small cell lung cancer, casting further doubt on the field.

On the other hand, there are several reasons for optimism. Bcl-2 is a good target, in our opinion, as the protein is overexpressed in 70% of cancers. It has become a very well understood and validated target, emerging as a key player in the balance between apoptosis and survival in cells after a tremendous amount of industry and academic work on the molecule2. Additionally, increasing apoptosis is an attractive alternative approach to treating cancer and would synergize with most standards of care. Pre-clinical and clinical data support both single-agent and combination therapy regimens, although only combination use will be pursued for registration. A good safety database exists with over 1,000 treated patients. Three concurrent registration trials maximise the potential for success.

As the molecular mechanisms behind cancer started to become clearer, targeted therapies were for a time thought of as a kind of magic bullet, much like monoclonal antibodies in the past. Knowing of a protein that drove cancer meant having a target

1 Isis Pharmaceuticals’ marketed antisense drug is Vitravene, an antisense inhibitor of ICAM-1 to treat cytomegalovirus (CMV) retinitis in AIDS patients intolerant of other therapies. Some debate exists that it may function through non-specific immune stimulation, rather than antisense reduction of the target protein.

2 Bcl-2 has over 11,000 entries in Medline in the last decade. The 2002 Nobel Prize in Physiology or Medicine was awarded for work on Bcl-2, the worm version of the protein called ced-9 and apoptosis as a key process in organ development.

GENASENSE TARGETS BCL-2, MAKING CANCER CELLS EASIER TO KILL



to attack. Targeted therapies promised effective treatments for cancer with high specificity and low toxicity. However, despite some resounding success stories, such as Glivec and Herceptin, the story is often more complicated. Despite this, we believe Bcl-2 is more similar to the Glivec and Herceptin targets, and thus has a higher chance of success than VEGF- or EGFR-targeted therapies.

Figure 7. Genasense Timeline

Genasense headline results from one of the three phase III

trials – melanoma, myeloma, CLL (Summer 2003

defined as June 21st to Sept 21st)

Potential Genasense launch

in melanomaor myeloma

2Q03 3Q03 4Q03 2005+2Q04 3Q04 4Q041Q04

Potential Genasense launch

in CLL

Genasense regulatory filing with the FDA in U.S. in

one of the three lead indications

Genasense regulatory filing with the EMEA in E.U.

Source: Bear, Stearns International Limited estimates.

Market Potential

We estimate the market potential of Genasense in the three initial ‘proof of concept’ indications at over US$500 million. Assuming it also works in solid tumours, Genasense could be a multi-billion dollar compound. Therefore, the publication of the headline data would allow an early filing and launch of Genasense. More importantly, we view the registration trials in the three ‘niche’ indications, as ‘proof of concept trials’, which could validate Bcl-2 as a target to enhance the efficacy of chemotherapy in much larger solid tumour indications.

Figure 8. Summary of Genasense Market Potential Indication Potential Launch 2008E+ Sales Initial Indications Malignant melanoma 1Q04 US$350 Multiple myeloma 1Q04 US$140 Chronic lymphocytic leukaemia 3Q04 US$70 Initial Indications Total US$560 Solid Tumour Indications Refractory NSCLC 2007 US$200 First-Line NSCLC 2007 US$450 Other Solid Tumours 2008+ Multi billion dollar peak sales Source: Bear, Stearns International Limited estimates. Notes: Assumes US$2,000 per Genasense course (a course is one week on, then two weeks off) and an average of six courses per chemotherapy cycle.



Among the initial three indications, malignant melanoma should have the greatest sales potential of around US$350 million, as the incidence of malignant melanoma is higher than in the other two indications. In addition, the medical need seems to be greatest. The official FDA-approved regime dacarbazine, which serves as base therapy in the registration trial, is no longer widely used on its own in the market. However, newer regimes (such as the ‘Dartmore regime’ or CVD, which both add cisplatin) haven’t shown that they make a difference to survival. Hence Genasense could fill an important medical need, and consequently we would assume fast penetration of the target population.

In multiple myeloma, we estimate sales potential of about US$140 million as the target population, with a prevalence of only 50,000 patients in the U.S., is quite small.

Finally, in CLL we estimate a sales potential of about US$70 million. The target population prevalence is of a similar size as for multiple myeloma; however, based on the trial design, we estimate that there will be fewer cycles per treatment and thus a lower sales potential than in multiple melanoma.

Aventis/Genta are conducting a 280-patient trial for Genasense + Taxotere in second-line NSCL cancer. The companies are also in the process of designing a larger phase II trial which looks at Genasense + Taxotere as first-line treatment in NSCL cancer. We estimate that the opportunity in NSCL cancer alone could exceed US$600 million, and this is just the first solid tumour indication.

‘NICHE INDICATIONS’

SOLID TUMOUR INDICATIONS



The Development Programme

Genta initially decided to market Genasense on its own, focusing development efforts on smaller high-need indications. An earlier effort in the first targeted indication, NHL, was de-emphasised following the approval in November 1997 and success of Rituxan in the indication.

In 1999, Genta initiated three phase III trials in melanoma, myeloma and CLL, which are currently ongoing. The choice of three concurrent trials was designed to maximise the potential for one of the trials to work and serve as a registration trial for initial approval.

After the licensing deal was announced (in May 2002), Aventis/Genta initiated the first solid tumour trial, which looks at Genasense in addition to Taxotere in second-line NSCL cancer. Further trials in solid tumour indications are currently in the design phase; however, we assume that trial enrollment will not start before results from the registration trials become available.

Headline Data Available Between 21 June and 21 September Genta indicates that we will see headline results data from at least one trial and a U.S. NDA filing in summer 2003 (21 June to 21 September). It will look at initial data from all three trials quickly and then assign staff to analyse the most favourable one fully, with the goal of filing an NDA in summer.

There is orphan drug designation for these three indications and fast track for myeloma and melanoma. Genta is even aiming for approval before year-end, which appears rather tight. We would assume a launch in melanoma or myeloma in first-quarter 2004 at the earliest.

There are interim analyses built into each of the trials, but none has been carried out at this point. Genta agreed with Aventis and the FDA that it will not undertake the interim analysis in the melanoma trial, although no decision on the other trials has been made.

All of the ongoing trials are unblinded, because Genasense is delivered via continuous infusion over a week. Thus there is no ethical way to do a placebo-controlled study, as it would over-burden already sick patients. While concerns can occur over physician bias or placebo effects because of this, in our view they are minor and, in any event, seem unavoidable.

The companies also plan to present data from a non-randomised Phase II clinical trial evaluating Genasense in prostate cancer at the ASCO meeting in May 2003, which is not part of the initial registration programme.



Three Potential Registration Trials

There are three concurrently conducted phase III trials ongoing that could serve as registration trials and lead to approval, in malignant melanoma, multiple myeloma and CLL.

Figure 9. Phase III Registration Genasense Clinical Trials Trial Indication Phase No. of

Patients Combination Therapy

Combination Details* Line of Treatment

Primary Endpoints

Timing Sponsor Design Details

Malignant melanoma

Phase III 750 Dacarbazine (DTIC)

FDA standard of care First line Survival Results summer 2003

Aventis/Genta Random, unblinded

Multiple myeloma Phase III 200 High dose dexamethasone

FDA standard of care, Up to 12 cycles.

Second line Time to progression

Results summer 2003


Chronic lymphocytic leukaemia

Phase III 200 Fludarabine and cyclophosphamide

FDA standard of care Second line Complete and partial responses

Results summer 2003


Source: Company reports; Bear, Stearns International Limited research. * Genasense is always administered as a continuous infusion for a week followed by two weeks off.

The trial in invasive malignant melanoma seems to be the one with the most focus from the companies and investors. The trial is now by far the largest after two increases in size, and the indication has the largest market potential. Additionally, it is the only trial with survival as the primary endpoint.

Description. The malignant melanoma trial includes 750 patients and enrolment completed in late January 2003. It tests Genasense as first-line therapy in combination with dacarbazine. The primary endpoint is survival. There is orphan drug and fast-track designation for this indication.

Current standard of care. The standard of care for malignant melanoma differs between the approved agents and the one more commonly used by clinicians. As per the FDA, the standard is dacarbazine (DTIC); by contrast clinicians more frequently add cisplatin or tamoxifen to dacarbazine. The companies are required to test Genasense against the approved therapy, so not much can be done to avoid this. To compensate to some extent, the companies are including some of the more accepted clinical standards in the NCI trials.

Powering. The malignant melanoma trial included 270 patients originally. Genta then expanded it to 450 patients (which was over-enrolling), and then both companies expanded it again, to 750 patients.

The first expansion was decided in summer 2002, was preplanned and was approved by the FDA. The second expansion was a joint decision, based on what they would have done to power the study with current funding levels. The FDA approved the move in advance and there was no statistical penalty, as the data was not examined in either increase. Adding patients to an ongoing clinical trial, not just once, but twice, could raise concerns that Genasense is not working as well as was expected or that the study was underpowered, but we believe Aventis simply decided that Genta had not been conservative enough and the companies simply wanted more comfortable powering of the trial. The trial completed enrolment in January 2003, which means an 80% increase in trial size in seven months. Genta still believes filing this summer is possible because mortality in malignant melanoma is unfortunately fast.

MALIGNANT MELANOMA



Single registration trials generally require more benefit to justify approval. As an example, if a 5% benefit is required in two separate trials, a single trial might need to show a 20% benefit. Median survival rates seen for invasive malignant melanoma first-line patients are typically about six months.

With 48,000 annual new cases of invasive malignant melanoma in the U.S. expected to continue growing, the market potential is substantial. Assuming a cost of Genasense at US$2,000 per course and six courses of chemotherapy/Genasense in a typical treatment cycle, we see peak worldwide sales of US$350 million in 2009.

Description. The multiple myeloma trial includes 200 patients and enrolment completed in early January 2003. It tests Genasense as second-line therapy in combination with high-dose dexamethasone. The primary endpoint is time to progression. There is orphan drug and fast-track designation for this indication.

Current standard of care. As with multiple myeloma, the standard of care for multiple myeloma differs between the approved agents and the one more commonly used by clinicians. As per the FDA, the standard is dexamethasone, but clinicians more frequently use melphan/prednisone or dexamethasone/thalidomide. Again, the companies are required to test Genasense against the approved therapy, but are including some more commonly accepted clinical standards with the NCI trials. The median survival time is roughly 2.5 to 3 years in these patients, but time to progressive disease (TTP) is significantly shorter, on the order of six months, to allow the trial to complete more quickly.

There are 15,000 annual new cases of multiple myeloma, 11,500 deaths each year and over 50,000 people currently living with the disease in the United States. The market potential is less substantial than for melanoma, but still reasonable. Again assuming a cost of Genasense at US$2,000 per course, and eight courses of chemotherapy/Genasense in a typical treatment cycle, we see peak worldwide sales of US$140 million in 2009.

Description. The CLL trial includes 200 patients, and enrollment completed in early March 2003. It tests Genasense as second-line therapy in combination with fludarabine and cyclophosphamide, the current FDA-approved standard of care. The primary endpoint is complete and partial responses. There is orphan drug designation, but not fast-track designation for this indication.

Current standard of care. The current FDA-approved standard of care is fludarabine and cyclophosphamide. The clinical standard of care is evolving standard with addition of monoclonal antibody Campath to treatment options. Combined complete and partial response rates in refractory patients on fludarabine (with or without cyclophosphamide) are typically 40%-50%.

There are 7,500 annual new cases of CLL, around 5,000 deaths each year and around 50,000 people with the disease in the United States. The market potential in CLL is the smallest of the three indications that are in regulatory trials now. Again assuming a cost of Genasense at US$2,000 per course and four courses of

MULTIPLE MYELOMA

CHRONIC LYMPHOCYTIC LEUKAEMIA



chemotherapy/Genasense in a typical treatment cycle, we see peak worldwide sales of US$70 million in 2009.

Aventis’ Solid Tumour Trials

After the licensing deal was announced (in May 2002), Aventis/Genta initiated the first solid tumour trial, which looks at Genasense in addition to Taxotere in second-line NSCL cancer. Further trials in solid tumour indications are currently in the design phase; however, we assume that trial enrollment will not start before results from the registration trials will be available.

Potential in Indications Outside Cancer

Potential exists in multiple sclerosis, rheumatoid arthritis and a variety of infectious diseases, but that is not the focus for now. Aventis may explore these areas post approval. Aventis already owns all rights and Genta receives royalties on all sales, regardless of indication.

Safety Database

There are minor theoretical concerns over side effects of antisense, but preclinical, phase I and phase II studies have shown Genasense to be well tolerated. Genasense has been tested in patients for a comparatively long time. Since 1995, more than 700 patients have been treated with Genasense. The most common side effects observed in clinical trials have been low-grade fever and fatigue. Thrombocytopenia (low platelet counts) has been observed in some patients who are also receiving myelosuppressive (bone marrow-suppressing) drugs. Importantly, most of these reactions do not overlap with the common side effects of standard anticancer therapy. The Bcl-2 deficient mouse model had melanin and immune system problems, but otherwise developed normally and had few problems.

Toxicity in the CLL trial was a concern in the past, but is minimal now. Initial suggestions from the 1997 paper showed CLL and NHL patients could be more sensitive to Genasense. Phase I data from December 2001 showed CLL patients have reactions qualitatively similar to, but less severe than, Rituxan: fever, low blood pressure and occasionally other complications related to the rapid killing of tumour cells. The dose used in the phase III trial was safe, potentially effective and can even induce tumour death without chemotherapy.

Route of Administration

Administration is via continuous intravenous infusion using a portable pump, typically over the period of five to seven days, followed by two weeks off. In clinical trials, this three-week cycle may be repeated for up to 12 months, depending on the trial. Chemotherapy is already routinely done using infusion pump and an ambulatory pump allows treatment at home, not just hospital. Other delivery methods are being explored.

AVEN

TIS: AN

D A R

AY OF L

IGH

T: GEN

ASENSE


Figure 10. Genasense Clinical Trials Trial Indication Phase No. of

Patients Combination Therapy Combination

Details* Line of Treatment Primary

Endpoints Rationale or Purpose Timing Sponsor, Other Details

Ongoing Trials . . . . . . . . . Malignant melanoma Phase III 750 Dacarbazine (DTIC) FDA standard of care First line Survival Proof of concept, registration trial Results summer

2003 Aventis/Genta, Random, unblinded

Multiple myeloma Phase III 200 High dose dexamethasone FDA standard of care. Up to 12 cycles.

Second line Time to progression

Proof of concept, registration trial Results summer 2003

Aventis/Genta, Random, unblinded


Phase III 200 Fludarabine and cyclophosphamide

FDA standard of care Second line Complete and partial responses

Proof of concept, registration trial Results summer 2003

Aventis/Genta, Random, unblinded

Non-small cell lung Phase II/III 280 Taxotere (docetaxel) Standard of care Second line Survival Expansion to solid tumour indications Initiated1Q03 Aventis/Genta, Random, unblinded Multiple myeloma Phase II NA Dexamethasone and

thalidomide Newer therapy Second line NA Combination with newer therapy Started 3Q02 NCI, Non-randomised, unblinded

Prostate

Phase II NA Taxotere (docetaxel) NA Second line NA Expansion to solid tumour indications Nearly complete NCI, Non-randomised, unblinded

Acute myeloid leukaemia

Phase II NA Mylotarg (toxic Mab) Newly approved standard of care.

Second line NA Combination with newer therapy NA NCI, Non-randomised, unblinded. Patients over 60.


Phase II NA None Monotherapy trial Second line NA Monotherapy trial NA NCI, Non-randomised, unblinded

Non-Hodgkin's lymphoma

Phase II NA Rituximab (MAb) NA Second line NA Combination with newer therapy NA NCI, Non-randomised, unblinded

Chronic myelogenous leukaemia

Phase II NA Imatinib mesylate NA Second line NA Expansion to other indications NA NCI, Non-randomised, unblinded

Small Cell Lung Cancer

Phase II NA Carboplatin and etoposide NA NA NA Expansion to solid tumour indications NA NCI, Non-randomised, unblended. Extensive disease.

Mantle cell lymphoma Phase II NA Monotherapy or R-CHOP (MAb/chemo combination)

NA NA NA Combination vs. monotherapy Recently started NCI, Non-randomised, unblinded

Colorectal Cancer Phase I/II NA Oxaliplatin, fluorouracil (5-FU), and leucovorin calcium

NA NA NA Expansion to solid tumour indications NA NCI, Non-randomised, unblinded. Advanced tumours.

Hepatocellular Carcinoma

Phase I/II NA Doxorubicin NA NA NA Dosing, safety, efficacy NA NCI, Non-randomised, unblinded. Advanced other incurable solid tumour.


Phase I NA Cytarabine and daunorubicin NA First line NA Expansion to first line NA NCI, Non-randomised, unblinded. Patients over 60.

Advanced Solid Tumours

Phase I NA Carboplatin and paclitaxel NA NA NA Expansion to solid tumour indications NA NCI, Non-randomised, unblinded

Pediatric Solid Tumours

Phase I NA Cyclophosphamide, doxorubicin, and dexrazoxane

NA Second line NA Expansion to pediatric solid tumour indications

NA NCI, Non-randomised, unblinded

Trials In Planning . . . . . . . . . Non-small cell lung Phase III in

design NA NA NA First Line NA First line therapy in large solid tumour

indication NA NA

Prostate Phase III in design

NA NA NA First Line NA First line therapy in large solid tumour indication

NA NA


Phase III in design

NA NA NA First Line NA First line for AML NA NA

Source: Company reports; Bear, Stearns International Limited research. Notes: NA – not available; *Genasense is always administered as a continuous infusion for a week followed by two weeks off.



The Development Challenge

Targeted Therapy – The Promise

As the molecular mechanisms behind cancer started to become clearer, targeted therapies have become possible. Knowing of a protein that drove cancer meant having a target to attack. Targeted therapies promise effective treatments for cancer with high specificity and low toxicity. However, despite some resounding success stories, the story is often more complicated. Targeted therapies may yet make substantial changes in how we treat cancer, but the path there is trickier than first imagined.

After a few clear successes in Herceptin and Glivec, targeted therapeutics as a class have had a few stumbles. The EGFR inhibitors had Erbitux’s regulatory filing turned away by the FDA. Iressa’s INTACT trials failed to meet the primary endpoints. Avastin, one of the VEGF inhibitors, and most recently Affinitak, targeting PKC-alpha, both also failed in phase III trials.

We believe these difficulties have caused investors to sour on the initial promise, and caused two misconceptions to arise:

1) Successful targeted therapies are going to be the rare exception rather than the rule; and

2) Combination trails bring numerous complications.

In contrast, based on our analysis of similarities and differences among a number of approaches to targeted anti-cancer therapy (see Figure 11 below), we have come to believe that not all targets are the same and that combination trials simply need to be well designed and controlled.

DIFFICULTIES AFTER INITIAL SUCCESSES



Figure 11. Selected Targeted Cancer Therapeutics Drug Target Class of Drug Lead Indication(s) Stage Company/

Partner(s) Target a Good Biomarker? Target a Prognostic or

Response Indicator? Herceptin

Her2 (EGFR 2) MAb Breast cancer Launched Genentech Yes, multiple gene amplifications in 25% of breast cancers

Yes

Glivec

Bcr-abl Small molecule B cell lymphoma Launched Novartis Yes, Philadelphia chromosome is clear

Yes

Iressa

EGFR TK Small molecule NSCLC Approved AstraZeneca Yes, but overexpression only Not clearly

Tarceva

EGFR TK Small molecule NSCLC Phase III OSI Pharma Yes, but overexpression only Not clearly

Erbitux

EGFR MAb NSCLC Phase III ImClone/Merck KgaA/BMY

Yes, but ovexpression only Not clearly

Avastin

VEGF MAb Breast, colorectal and NSCLC

Failed Phase III Genentech Not perfect, secreted marker Some correlation

Affinitak

PCK alpha Antisense NSCLC Failed Phase III ISIS/Lily Difficult, many close isotypes Some correlation, can have mixed effects

Tariquidar

P-gp Small molecule NSCLC Phase III Xenova/QLT No, induced by chemo, expression can be transient

Likely, but too difficult to test in many cases

Genasense

Bcl-2 Antisense Melanoma, myeloma, CLL, NSCLC

Phase III Genta/Aventis Yes, some mutations, but also some simple overexpression and also induced by chemo

Yes

Source: PharmaProjects; Bear, Stearns International Limited research. Notes: * e.g. NEJM 1989 320:1047-1054, Blood 1997 90: 244-251 Eur J Cancer 1999 Aug;35(8):1214-9, Blood 1999 93: 3081-3087.

Not All Targeted Therapies Are the Same

Not all targeted therapies are the same, and much of the difference is related to the characteristics of the target itself. What sets the targets of Glivec and Herceptin apart from the targets of the drug candidates that stumbled? The two clear differences for the success stories are:

1) The target’s upregulation, which includes the permanence and level of that upregulation prior to treatment.

This can also be thought of as how good a biomarker the target can be. Can patients who should respond be selected easily based on the expression of the target? Glivec’s target Bcr-abl and Herceptin’s target Her2 are both permanent mutations that cause manifold amplifications of the protein. As such, they are good biomarkers for selecting patients likely to respond. By contrast, a poor biomarker won’t allow patient selection, risking underpowered or unwieldy trials.

2) The ability of the target to predict a tumour’s response to therapy, time to disease progression or outcome.

Does the target actually drive the cancer? Again, Glivec’s target Bcr-abl and Herceptin’s target Her2 both signal the cancer cells to proliferate and have a strong correlation to prognosis. Poor correlation with prognosis risks failure because correlation does not always equal causation.

Based on these differences, which we believe to be crucial, we have developed a framework that could help differentiate between various targeted therapies. By sorting the targeted therapies based on target upregulation and prognostic power, Genasense does not group with either the Glivec/Herceptin or EGFR/VEGF inhibitors, but in our view falls closer to the former than to the latter.

HOW DOES GENASENSE COMPARE?



Bcl-2 overexpression is not a perfect biomarker compared to the Bcr-abl mutation or the Philadelphia chromosome, but it is upregulated in 70% of cancers and does correlate well with response to therapy and disease prognosis, as several studies have suggested3. We conclude that Genasense is closer to Glivec or Herceptin than to Erbitux or Avastin, and thus should have a higher chance of clinical benefit.

An additional third axis that we could add to our framework could have been the redundancy of the target. In other words, one can ask if there are other molecules or pathways in the cancer cell that can accomplish the same job as the target. In the case of signalling pathway targets that drive proliferation, such as Her2, Bcr-abl, EGFR and PKC-alpha, there are typically other pathways also driving a cancer’s division. By contrast, there is far less evidence that VEGF or Bcl-2 have redundant systems that can accomplish the same function. Thus, they could be more critical to a tumour’s progression and be better targets.

3 NEJM 1989 320:1047-1054 shows this in large cell lymphoma, Blood 1997 90: 244-251 shows this in NHL, Eur J Cancer 1999 Aug;35(8):1214-9 shows this in ovarian cancer, Blood 1999 93: 3081-3087 shows this in indolent follicular lymphoma



Figure 12. Selected Targeted Cancer Therapeutics Comparison

The Target as a Biomarker

Iressa, Tarceva,Erbitux –

EGFR pathway

Herceptin –Her2

The Target as a Prognostic Indicator

Genasense –Bcl-2

Gleevec –Bcr-abl

Affinitak –PKC alpha

VEGFInhibitors

Tarquidar –P-gp

• Is the target a permanent mutation and over expressed at high levels?

• Is the over expression only mild, reversible or only induced by chemotherapy?

• Does the target correlate well with response to chemotherapy, disease progression or outcome?

Source: Bear, Stearns International Limited research.

Combination Trials

Trials with a single-agent treatment represent a significantly less risky regulatory approach than developing a compound only in combination with another treatment. Obviously, data are easier to interpret when using only a single active drug.

Developing a compound (A) in combination with another therapy (B) brings complications, as we will describe further below. Indeed a number of notable setbacks at regulatory filing have recently occurred for drugs used in combination trials, including Erbitux and Iressa.

ImClone’s Erbitux application was rebuffed by the FDA. Apart from issues with regards to patient selection (it was questionable whether all patients were truly refractory to irinotecan, the combination partner (B) in the trial), it was also criticised that it was a small, uncontrolled study. In addition, the lack of monotherapy data was also criticised. In short, ImClone had only conducted a trial (A)+(B), but had neither provided data on (B) vs. (A)+(B) in the form of a randomised controlled trial, nor (A) vs. (A)+(B) to establish whether the compound has any activity as monotherapy.

RECENT SETBACKS



Iressa failed in combination with chemotherapy. (Iressa has shown activity as monotherapy, i.e. (A) alone has shown an effect on response rates, a surrogate marker, however in a controlled setting added to chemotherapy (B) vs. (A)+(B) failed to show a difference in survival for the combination). These results are still controversial. Some argue that the chemotherapy (B) might somehow have cancelled out the benefits of Iressa (A), there is no precedent in oncology to say that a drug would work alone but not with standard chemotherapy. It might equally be possible that the study was just underpowered due to insufficient patient selection.

Concerns in Combination Trials

In our view there are three key issues which need to be addressed in combination trials.

1) A regulatory filing based on data from combination trial (A)+(B) may require additional supporting information about the new compound (A) on safety and pharmacokinetic parameters. Hence additional monotherapy trials, apart from the pivotal registration trials, should be conducted. (In some cases, monotherapy trials may even also yield some efficacy data about monotherapy.) Genta has established that Genasense is safe and has some direct killing effects on myeloma cells. Single-agent trials have also included an NHL trial and various phamacokinetic studies.

2) Testing in combination (A)+(B) compared to an appropriate base (B) is also a concern. In many cases, the FDA-approved standard of care (B) is different from the clinical standard of care (C?). The fields advance faster than regulatory agencies, but registration trials require the FDA-approved background therapy (B). This might later cause problems in clinical practice as the data required to achieve approval might be of little use to help oncologists in their daily practice. We believe that some of the trials conducted by the NCI should help to bridge this gap. As per the FDA standard for myeloma, the phase III trial is with dexamethasone. Clinicians use dexamethasone/thalidomide more frequently, and this combination is the subject of an NCI trial.

3) Some argue that if (A) by itself doesn’t cause the tumour to shrink, different endpoints other than response rates are required, such as time to progression (TTP) or survival data. This in turn would lead to longer trails, as survival data in particular require a long time horizon to be established. We believe that if appropriate control arms (B) exist, response rates (or an improved response rate of (A)+(B) over the control arm (B)) should still allow endpoints such as response rates to be used. All of Genta’s three registration trials have a control arm (i.e. they look at (A)+(B) vs. (B)). The conservative approach of doing controlled trials is more costly than the often-used one-arm only salvage therapy registration trials; hence funding bottlenecks might incur the risk of underpowering.



Earlier Genasense data point to combination with chemotherapy being the best course. The biology of Bcl-2 suggests that it should function best when sensitising cancer cells to chemotherapy, as opposed to killing them on its own. Further, preclinical studies have shown that Genasense synergizes with most types of anticancer treatment, including chemotherapy, radiation and immunotherapy. For example, based on preclinical data, Genta has conducted preliminary clinical trials using Genasense in combination with paclitaxel (Taxol), irinotecan (Camptosar), fludarabine (Fludara), cyclophosphamide (Cytoxan), docetaxel (Taxotere), Mylotarg, cytosine arabinoside and dexamethasone

The initial Genasense registration filing expected in 2003 will only be for use in combination with existing chemotherapy regimes rather than the ones used currently in the clinic. However, Genasense may avoid clinical regulatory problems often associated with combination therapies, as all three registration trials are randomised and controlled.

In 1999, Genta CEO Raymond Warrell made a somewhat controversial decision to forego testing Genasense in the type of midsize trials that backfired for Erbitux. It became clear to Genta that the NCI trails may well yield good information, but wouldn’t lead to registration. Smaller trials with no control group also often yield ambiguous data. This approach could end up being both lower risk and lower cost than ImClone’s strategy.

The NCI trials will now serve as supporting and marketing trials for putting the drug in many clinicians’ hands and encouraging rapid uptake in other indications and academic trials.

Genasense appears to have some direct killing effects on myeloma cells and can inhibit angiogenisis. Single-agent trials have included 1997 NHL trial and various phamacokinetic studies. These are not required in the NDA, but should support the application.

Powering Up the Melanoma Trial

The Affinitak failure may have been because of an insufficient pharmacologic target, but the results suggest the control group did better than expected, indicating that the trial may have been underpowered (see Appendix 3 for more detail). This same risk exists for the Genasense phase III melanoma trial. The increased sample sizes in this trial may have been put into place because all patients were living longer than anticipated.

The estimated sample size and the effect to be observed in control groups are based on published research. In cancer, the most relevant publications with survival data often take many months to complete analyses and publish results even after the last patient is evaluated. This means that the survival data can be out of date by the time they are published and patients, on average, are living longer. This seems to have been the case with Affinitak’s recently failed trial.

BIOLOGY SUPPORTS COMBINATION USE FOR GENASENSE

SLOW AND STEADY WINS THE RACE



It has been disclosed that the companies do not know survival rates by the identity of treatment groups. However, longer survival in both groups would mean that either longer follow-up might be needed to show a treatment effect or a larger sample size might be required to show a smaller difference in treatment effect between treatment arms.

Genta’s critics believe that the only reason for adding more patients to the study at this point is that Genasense is not working well enough. Patients who are taking the drug are not living longer than those who don’t. Increasing patient enrolment essentially retools the trial and gives the drug more chances to produce a statistically significant positive result.

Genta has not performed any type of preliminary analysis of the trial’s results that would have given it an early clue that Genasense wasn’t working. Instead, more patients were added only because the company wanted to increase the trial’s sample size which, from a statistical level, is more likely to produce positive results.



Genasense Mechanism and Antisense Technology Review

Genasense is an antisense therapeutic, based on a technology that isn’t new. Originating in the late 1970s near the time of the first gene-splicing experiments, antisense generally aims to reduce the expression of a specific protein by interrupting the basic biological process of translation.

The central dogma of molecular biology is the transfer of information from DNA to RNA to protein, with DNA being the permanent information storage system, RNA serving as an information transfer system and proteins the primary effectors of biology. Genetic information contained in DNA is copied or transcribed into a messenger RNA (mRNA) and mRNA is subsequently translated into protein.

Antisense drugs are typically small, chemically modified strands of DNA. They block the DNA-RNA-protein process for a particular protein by binding to a specific, complementary mRNA molecule via nucleic acid base pairing. Once the DNA is bound to the mRNA, the mRNA is unable to be translated and subsequently degraded. Thus the specific protein production is stopped.

Figure 13. Genasense Mechanism

ProteinBcl-2 protein

NucleusNucleus

DNABcl-2 gene

Messenger RNABcl-2 mRNA

Transcription

Antisense andmRNA Complex

Bcl-2 mRNA bound toantisense DNA

Transcription

CellCellMembraneMembrane

Antisense DNAComplementaryto Bcl-2 mRNA

Translation

No Protein Reduced Bcl-2

expression

Translation block and

degradation

Source: Bear, Stearns International Limited.

When antisense technology first advanced to the clinic, it came with high expectations, but the path to approval has been long and bumpy. This is not unlike the original promise of monoclonal antibodies (MAbs), hailed as ‘magic bullets’

ANTISENSE TECHNOLOGY

MABS EVENTUALLY OVERCAME SIMILAR CLINICAL HURDLES



when first entering the clinic. For MAbs, there were years of clinical setbacks until various technical challenges were overcome and more appropriate indications chosen. Today, MAbs are beginning to come onto the market in greater numbers and provide significant clinical benefit to patients.

It may well be that we are nearing a similar turnaround for antisense-based drugs, and we will see fewer failures as the technology begins to bear more fruit. Genasense heads the list of antisense therapeutics in clinical development. After a long and bumpy road, is the second antisense drug nearing the clinical finish line?

Figure 14. Antisense Therapeutics in Clinical Development Antisense Drug Target Lead Indication Stage Company Partner(s) Vitravene ICAM-1 Cytomegalovirus Infection Launched ISIS Pharmaceuticals Novartis Ophthalmics Genasense Bcl-2 Melanoma, myeloma, CLL Phase III Genta Aventis, Baxter ISIS-3521 PKC-alpha Non-small cell lung cancer Failed Phase III ISIS Pharmaceuticals Eli Lilly Alicaforsen ICAM-1 Crohn's disease Phase III ISIS Pharmaceuticals Oncomyc-NG c-myc Colorectal cancer Phase III AVI BioPharma ISIS-2503 Ha-ras Non-small cell lung cancer Phase II ISIS Pharmaceuticals Elan ISIS-14803 Polyprotein initiation codon Hepatitis-C virus infection Phase II ISIS Pharmaceuticals Elan ISIS-104838 TNF-alpha Psoriasis Phase II ISIS Pharmaceuticals Elan MG-98 DNA methyltransferase Head and neck cancer Phase II MethylGene British Biotech, MGI Pharma GEM-231 Protein kinase A Cancer Phase II Hybridon GEM-92 HIV gag HIV/AIDS Phase II Hybridon EPI-2010 Adenosine A1 receptor Asthma Phase II EpiGenesis Chiesi, Taisho, Hybridon GTI-2040 R2 ribonucleotide reductase Renal cancer Phase II Lorus Therapeutics TGF-beta antisense TGF-beta Small cell lung cancer Phase II Immune Response NovaRx INX-3280 c-myc Cancer Phase II Inex LE-AON c-raf1 Head and neck cancer Phase II NeoPharm AP-12009 TGF-beta 2 Brain cancer Phase II Antisense Pharma Resten-NG (AVI-4126) c-myc Restenosis Phase II AVI BioPharma Medtronic OGX-011 Clusterin Prostate cancer Phase I OncoGenex Technologies ISIS Pharmaceuticals GTI-2501 R1 ribonucleotide reductase Cancer Phase I Lorus Therapeutics HGTV-43 HIV TAR & two sites of the TAT region HIV/AIDS Phase I Enzo Biochem EN-101 Acetlycholinesterase Myasthenia gravis Phase I Ester Neurosciences AVI-4557 CYP3A4 Unspecified Phase I AVI BioPharma Source: PharmaProjects; Bear, Stearns International Limited research.



The Target: Bcl-2

Bcl-2, A Novel Specific Protein Target

The Bcl-2 protein is part of a complex system that controls apoptosis, or programmed cell death (PCD). Bcl-2 is an attractive target for cancer therapeutics because the basic biology surrounding the molecule is now well known and the protein is very central to resisting apoptosis. Bcl-2 as a target is unique from other new generation cancer therapeutics, such as EGFR pathway inhibitors, in that it sensitises the cancer cells to death, rather than limiting proliferation.

Bcl-2 was first discovered as a common mutation found in B cell lymphomas, in which a chromosome translocation resulted in excess production of the protein. It was subsequently found that the protein, named B cell lymphoma gene-2 (Bcl-2), confers resistance to apoptosis and allows the lymphoma cells to live for an excessively long time.

Bcl-2 is a member of a family of proteins that exert diverse effects on apoptosis. The individual proteins pair up in various combinations and these protein dimers act on the apoptosis machinery. The relative protein expression or balance between apoptosis inhibitors (e.g. Bcl-2, Bcl-Xl) and those that drive apoptosis (e.g. Bad, Bax) determine the outcome for a cell.

Figure 15. Bcl-2 Family Members Anti-apoptotic (protect from apoptosis) Notes Bcl-2 Best understood anti-apoptotic family member Bcl-Xl (long form) Long isoform of Bcl-X on balance is anti-apoptotic Bag Less well understood role Pro-apoptotic (drive apoptosis) Notes Bax Best understood pro-apoptotic family member, Bcl-2 cleavage product Bcl-Xs (short form) Short isoform of Bcl-X is pro-apoptotic Bid May also have some anti-apoptotic activity Bad Less well understood role Bak Less well understood role Source: Bear, Stearns International Limited.

Bcl-2 localises predominantly to the outer mitochonrial membrane, the surface of the energy-producing organelle in a cell. During apoptosis, cytochrome c and apoptosis-inducing factor (AIF) are released from the mitochondria, which then induce caspase activation and apoptosis. Increased levels of Bcl-2 can prevent the release of these molecules, whereas caspase inhibitors cannot.

There are several different ways in which Bcl-2 may function. Bcl-2 may form homodimers to actively block apoptosis or it may form heterodimers with a proapoptotic family member to block by sequestering those apoptosis-inducing proteins. What is important is the ratio or balance of antiapoptotic Bcl-2 family members to proapoptotic family members in determining if apoptosis will occur. In either event, excess Bcl-2 protein expression promotes survival.

BCL-2 HELPS A CELL RESIST APOPTOSIS

MECHANISM OF ACTION



Figure 16. Bcl-2 Mechanism of Action

MitochondriaMitochondria

CellCellMembraneMembrane

Activationof effectorcaspases

Internal Signal:e.g. DNA damage or

oxidative stress

External Signal:TNF alpha family members

e.g. FAS ligand/FAS

Bcl-2

Apoptosis orProgrammed

cell death

Cyt C and AIF Cyt C and AIF

Bcl-2 protectscells from apoptosis

by multiple mechanisms

Enzymatictriggering

of apoptoticcascade

Serine/threoninephosphorylation and

initiator caspases

Frequentlymediated via

p53

Initiatorcaspases

Cytochrome C and AIF (apoptosis inducing factor)

released frommitochondria

Block Block

Block


Apoptosis may be induced by a variety of signals including irreparable DNA damage, oxidative stress or signalling through ‘death domain’ surface receptors. This form of cellular suicide prevents the expansion of damaged or out of control cells. Many chemotherapy agents induce DNA damage that in turn causes apoptosis. Bcl-2 works to prevent apoptosis and, therefore, its overexpression can prevent apoptosis in cells that are damaged. This can lead to the continued division of the mutated cells lines and eventually cancer. Additionally, overexpression of Bcl-2 can contribute to metastasis in certain cancers and indicates poor prognosis in many cancers. Bcl-2 is upregulated in 70% of cancers (see Figure 17 below) and blocking it should help sensitize cancer cells to chemotherapy-induced death.

LINK TO CANCER



Figure 17. % of Tumours with Bcl-2 Upregulation

0102030405060708090

100

Breast Colorectal Melanoma Prostate(HR)

SCLC AML CLL Ly mphoma My eloma

Solid Tumours Hematologic Cancers

Source: Genta presentation.

Apoptosis Activation, A Novel Class Of Target

Apoptosis is a necessary and normal process, by which a cell commits an efficient and organised form of suicide, for the benefit of the organism as a whole. It is required to accommodate the billions of new cells produced daily and to eliminate old, damaged or out of control cells. Apoptosis can be thought of as a default self-destruct mechanism in a cell, used whenever anything goes wrong. The process can be started either internally or externally, in response to a variety of signals.

Most cells seem to have this self-regulation mechanism firmly in place, and can normally activate apoptosis internally. A cell seems to continually assess its own situation, asking itself:

��Do I have DNA damage or unacceptable levels of oxidative stress?

��Am I receiving too many growth signals or is proliferation going too fast?

��Has cell division started, but not yet completed it in a timely manner?

��Have I been infected by a virus?

If a cell answers yes to any of the above questions, it normally initiates its own programmed cell death, otherwise known as apoptosis.

Alternatively, external signals can also push the self-destruct button on a cell, starting apoptosis. Immune system cells that detect a lurking virus in a cell provide a signal to that cell through an appropriately named death receptor. This causes the infected cell to begin apoptosis, limiting infection of neighboring cells.

Normal cells can activate apoptosis in response to a variety of stimuli. However, in the case of cancer cells and some viral infections, this process can be subverted or

REACTIVATING THE SELF-DESTRUCT SYSTEM FOR A CELL

APOPTOSIS AND ITS ROLE IN CANCER



disabled, allowing rapidly proliferating or infected cells to go on living to the detriment of the organism as a whole.

The protein players and their interaction in this complicated and intricate process have become better known in the last decade. While Bcl-2 is a key player in these events, numerous other proteins have roles as well, and can be targets for therapeutics. However, many of these potential targets are upstream of Bcl-2 and will eventually downregulate Bcl-2. As an example, some standard chemotherapy drugs, such as paclitaxel, vincristine, and docetaxel, indirectly reduce the amount of Bcl-2 protein but this is not their primary mechanism of action.

Traditional chemotherapy agents or radiation treatment simply kill quickly dividing cells preferentially. Most chemotherapy drugs act by damaging DNA or causing cell cycle arrest, which then induces apoptosis. New generation drugs target the drivers of cancer more specifically. Many of the new generation cancer therapeutics aim to limit proliferation. Others activate apoptosis and sensitise the cancer cells to die more easily.

Figure 18. Two Mechanistic Contributors to Cancer: Excess Proliferation and Inadequate Death

Excess Proliferation Inadequate CellDeath (apoptosis)

Growth receptor or signalling moleculeover expression/ constitutive activation (EGFR, ras) or cell cycle control protein

defect (p21, p27)

Survival factor over expression (Bcl-2) orapoptosis mediator defect (p53)

Cancer:Inappropriate control

of the cell cycle

Most cancers will have multiple mutationscausing both proliferation and inadequate apoptosis


Competition – Bcl-2 Inhibitors

There are few direct competitors to Genasense in development. Of those that directly target Bcl-2, only Genasense has advanced beyond preclinical development. While Aventis and Genta have a substantial lead, there does seem to be active work in the area by major pharma competitors.

GENASENSE IS THE ONLY ANTI-BCL-2 AGENT IN CLINIC



Figure 19. Therapeutics Targeting Bcl-2 in Development Antisense Drug Lead Indication(s) Stage Company Partner(s) Genasense Melanoma, myeloma, CLL Phase III Genta Aventis, Baxter INGN-261 Colorectal cancer Preclinical Introgen Therapeutics Bcl-2 inhibitors Breast cancer Preclinical Shire Bcl-2 inhibitors Cancer Preclinical Wyeth Bcl-2/Bcl-xL inhibitor Cancer Preclinical Idun Abbott GX-015 Cancer Preclinical Gemin X Biotechnologies HA-14-1 Cancer Preclinical Academic Source: PharmaProjects; Bear, Stearns International Limited research.

Competition – Apoptosis Activators

A number of compounds are in development that also activate apoptosis, or make cells more sensitive to apoptosis signals. However, most of these directly or indirectly influence Bcl-2. Examples of targets to reactivate apoptosis are:

��Retinoic acids, which may transcriptionally control Bcl-2 levels.

��NF-kB, which activates Bcl-2 expression in lymphoma cells.

��Inhibitors of phosphodiesterase (PDE) type IV, the PDE predominantly expressed in B-CLL cells, have been shown to induce cAMP-dependent apoptosis. PDE IV inhibitors suppress the expression of Bcl-2 and induce the pro-apoptotic protein Bax.

��Bcl-2 binding peptides that inactivate the protein and antimycin A that binds to Bcl-2 related proteins.

The drugs in development can induce or increase sensitivity to apoptosis, either as a primary mechanism or secondary to some other disruption in the cell. Most are upstream of Bcl-2 and increase sensitivity to apoptosis via the Bcl-2 family of proteins, which are central to the apoptosis machinery/decision. However, none of these are yet in phase III trials.

The real competitors will be other drugs that make chemotherapy better, such as Erbitux, Iressa, Tarceva, Affinitak and Avastin. Because Genasense and these drugs act on differing pathways leading to cancer, they could be used in combination and potentially synergize. However, the first ‘add on’ to conventional chemotherapy will make the path for those following more difficult.

OTHER TARGETS ALSO ACTIVATE APOPTOSIS

THE REAL COMPETITORS



Aventis/Genta Licensing Agreement

In April 2002, Genta and Aventis Pharmaceuticals announced that they would jointly develop and co-promote Genasense in the United States. This is the second-largest transaction for a single biotech product, following the ImClone/Bristol-Myers Squibb agreement on Erbitux. A comparison is provided in the table below.

Figure 20. Erbitux and Genasense Agreements Comparison Deal Component Imclone/Bristol-Myers Squibb Revised Imclone/Bristol-Myers Squibb Aventis/Genta Date Announced September 19, 2001 March 5, 2002 April 29, 2002 Compound Erbitux - Genasense Class of Molecule Monoclonal antibody - Antisense Target EGFR (proliferation inhibition) - Bcl-2 (apoptosis activation) Lead Indication(s) Colorectal cancer - Melanoma, myeloma, chronic lymphocytic leukaemia Total Upfront US$1.2B US$200M on signing of new agreement US$132M Upfront Cash US$200M US$140M in lieu of the US$300M milestone

payment US$50M

Upfront Equity Investment US$1B (20% of ImClone) - US$72M (9% of Genta) Upfront Other - US$60M on 1st anniversary of revised deal US$10 convertible debt Milestones US$300M on regulatory filing US$250M on 1st indication approval in lieu

of US$500M milestone US$75M and US$20M convertible debt on 1st FDA approval

US$500M on FDA approval US$250M on 2nd indication approval US$40M on 2nd FDA approval US$20M each on 1st and 2nd European approval US$40M each on 1st solid tumour and 1st front line “major”

tumour approval in U.S. US$45M on 2nd “major” tumour approval in U.S. Development Costs Shared for U.S. and Japan

development No change Aventis pays 75% of U.S development; 100% other

development costs; 100% marketing costs and funds 50 Genta sales reps

Royalty Rate to Biotech Partner

Distribution fee of 39% of N. American sales; Cost plus 10% for supply in N. America; 50% of profits in Japan, plus co-promotion rights.

No change 18-22% estimated in U.S., estimated 15% in Europe

Rights Acquired by Pharma Partner

Exclusive N. American and Japan rights

No change Co-promotion in U.S; Exclusive rights ex-U.S.

Total Deal Cost US$2B Reduced total cost by US$100M, spread payments out

US$477M

Expiration 2018 2018 NA Source: Company reports; Bear, Stearns International Limited research.

The deal includes up to US$477 million in cash, equity, milestones and convertible debt to Genta. Aventis has exclusive rights ex-U.S. and the companies will co-promote the drug in the United States. Genta will receive royalties on all worldwide sales of Genasense, which we estimate to be around 18%-22%, which scales up with sales. The royalty will be slightly less in Europe, we estimate at 15%.

Specific milestones to Genta are detailed in Figure 20 above. In sum, the bulk (US$300 million of the US$477 million) of the payments from Aventis to Genta are tied to regulatory success of the drug. This is a more conservative approach to deal making, but one could argue the best for all, assuming the drug works.

In March 2003, the companies amended the agreement to pull forward a US$40 million revolving line of credit due at the first U.S. approval. Genta intends to begin drawing it down in first-quarter 2003.



Genta retains responsibility for worldwide manufacturing and U.S. regulatory filings. Joint management teams, including representatives from both partners, will oversee the alliance.

Aventis funds 75% of all U.S. NDA-directed development costs and 100% of all other development costs, within the U.S. and worldwide. Aventis pays 100% of marketing costs and funds 50 full-time Genta sales reps.

Genta estimates the first five years of the agreement cost savings will be US$500 million, placed onto Aventis. In exchange, Aventis owns rights to all indications, cancer or otherwise, and Genta receives royalties on everything. Potential exists in MS and RA for Bcl-2, but that is not the focus for now.

Comparison to the BMY/IMCL Erbitux Deal

The Genta/Aventis agreement was the second-largest biotech deal for a single product, behind the ImClone/Bristol Myers agreement. Aventis clearly learned from Bristol’s pain and the key difference to this deal following the difficulties with Erbitux include a more conservative structuring, with greater back loading, and a smaller equity investment at less of a premium.

On 19 September 2001 Bristol Myers-Squibb agreed to pay about US$1 billion for a 20% stake in ImClone at US$70 a share, a 40% premium over its price at the time. Bristol would also pay US$200 million cash upfront and another US$800 million as three separate milestone payments tied to the development and approval of Erbitux. Bristol would market the drug, taking an estimated 40% of profits. This US$2 billion was and remains the biggest single-product deal ever made in biotech.

The rationale seemed reasonable at the time. If Erbitux turned out to be a US$1 billion a year drug, Bristol could reap as much as US$400 million annually. The US$1 billion equity investment would provide an additional benefit if the drug was successful and ImClone’s stock climbed. However, in hindsight the deal was risky with so much money being paid upfront, as Bristol has learned to its detriment.

In contrast to the BMY/IMCL deal that likely influenced the subsequent Genasense agreement, the Aventis/Genta deal is far more back-end loaded. While Aventis is paying to support the development of the drug in larger solid tumour indications, less will be lost if the drug fails.

Aventis’ Contribution to the Party

Aventis has brought substantially more resources to the table, as well as a more conservative, longer-term view. These resources include financial strength to fund bigger trials, and wide clinical and regulatory experience in oncology. Along with the resources, Aventis brings a keen interest in the larger market potential of solid tumour indications.



1) Financial Resources

According to the April 2002 agreement, Aventis will pay up to US$477 million in cash, equity, milestones and convertible debt to Genta. Aventis further funds 75% of all U.S. NDA-directed development costs and 100% of all other development costs, within the U.S. and worldwide. Aventis pays 100% of marketing costs and funds 50 full-time Genta sales reps.

In exchange, Aventis owns Genasense rights to all indications, cancer or otherwise, and Genta receives royalties on sales worldwide.

2) Clinical and Regulatory Oncology Experience

Aventis has the largest oncology salesforce in the United States. Aventis also brings access to a wider base of clinicians. We believe the second increase in melanoma trial size was partly because of increased investigator access in Europe via Aventis, in addition to more funding.

Genta retains responsibility for U.S. regulatory filings and members of the Genta team are responsible collectively for 10 NDAs. However, they will no doubt also draw on extensive expertise from Aventis. NDA filing will be in the summer of 2003. There will be a small (two to three month) lag for EMEA filing by Aventis, but is likely to be the same data.

3) Aventis brings interest in larger solid tumour indications

A phase II Aventis-funded trial in refractory NSCLC has already started, a randomised trial in combination with taxotere as second-line therapy in 280 patients. Supporting wider use, NCI-funded trials in prostate, colon, lung, NHL are already ongoing. A total of 11 NCI-funded trials in 10 distinct indications are ongoing. See Error! Reference source not found. for more details.

Randomised trials of first-line NSCLC, first-line prostate and AML trials are in the design process now and should start this year.

Manufacturing

Genta retains responsibility for worldwide manufacturing. It has in place a long-term manufacturing and supply agreement with Avecia, a leading global fine and specialty chemicals company. A recent five-year agreement provides for sufficient quantities to accommodate both existing and new clinical trials, as well as quantities of Genasense for worldwide commercial launch. The drug will be produced at plants in Boston and Scotland.

Genta and Avecia initiated their first manufacturing agreement in 2001, under which Avecia has supplied Genasense for the worldwide clinical trials. Avecia’s biotechnology facility in Milford, MA has been and will continue to be the primary source of supply for Genasense.

LONG-TERM MANUFACTURING AND SUPPLY AGREEMENT WITH AVECIA IN PLACE



The processes involved in manufacturing antisense drugs compare favourably to those of traditional small molecule drugs or protein therapeutics. All DNA-based antisense drugs share a common chemistry, so the manufacturing processes can be standardised easily. Additionally, as synthetic rather than biologic molecules, the expensive equipment and processes for fermentation needed with protein therapeutics are not required.

Genta’s dealing with Avecia includes normal, prudent contractual agreements ahead of an NDA filing and a modest US$5 million advance to facilitate Avecia’s expansions. Avecia is putting this toward the creation of the world’s first 1,000kg DNA medicines production facility.



Appendix 1: Genasense Sales Projections

Figure 21. First-Line Malignant Melanoma Revenue Model Genasense Revenues (patients in thousands) 2004E 2005E 2006E 2007E 2008EUnited States - First Line Malignant Melanoma New cases of invasive malignant melanoma 50 52 55 58 61 % patients receiving first line chemotherapy 75% 75% 75% 75% 75%Patients receiving first line chemotherapy 37 39 41 43 46 Penetration 5% 10% 20% 30% 40%Number of patients on Genasense 1.9 3.9 8.3 13.0 18.2 Chemotherapy cycles per typical treatment 6 6 6 6 6 Annual Genasense courses in US (thousands) 11 24 50 78 109 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End User Revenues (US$m) US$22 US$47 US$99 US$156 US$219 Europe - First Line Malignant Melanoma New cases of invasive malignant melanoma 40 42 44 46 49 % patients receiving first line chemotherapy 75% 75% 75% 75% 75%Patients receiving first line chemotherapy 30 31 33 35 36 Penetration 4% 8% 15% 23% 30%Number of patients on Genasense 1.1 2.4 5.0 7.8 10.9 Chemotherapy cycles per typical treatment 6 6 6 6 6 Annual Genasense courses in Europe (thousands) 7 14 30 47 66 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End user revenues (US$m) US$13 US$28 US$59 US$94 US$131 Total End User Sales (US$m) US$36 US$76 US$159 US$250 US$350Source: Bear, Stearns International Limited estimates.

Figure 22. Second-Line Multiple Myeloma Revenue Model Genasense Revenues (patients in thousands) 2004E 2005E 2006E 2007E 2008EUnited States - Second Line Multiple Myeloma Multiple myeloma patients 50 50 50 50 50 % patients relapsing and receiving second line chemotherapy 25% 25% 25% 25% 25%Patients relapsing and receiving second line chemotherapy 13 13 13 13 13 Penetration 5% 10% 20% 30% 40%Number of patients on Genasense 0.6 1.3 2.5 3.8 5.0 Chemotherapy cycles per typical treatment 8 8 8 8 8 Annual Genasense courses in US (thousands) 5 10 20 30 40 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End User Revenues (US$m) US$10 US$20 US$40 US$60 US$80 Europe - Second Line Multiple Myeloma Multiple myeloma patients 50 50 50 50 50 % patients relapsing and receiving second line chemotherapy 25% 25% 25% 25% 25%Patients relapsing and receiving second line chemotherapy 13 13 13 13 13 Penetration 4% 8% 15% 23% 30%Number of patients on Genasense 0.5 0.9 1.9 2.8 3.8 Chemotherapy cycles per typical treatment 8 8 8 8 8 Annual Genasense courses in Europe (thousands) 4 8 15 23 30 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End user revenues (US$m) US$8 US$15 US$30 US$45 US$60 Total End User Sales (US$m) US$18 US$35 US$70 US$105 US$140Source: Bear, Stearns International Limited estimates.



Figure 23. Second-Line Chronic Lymphocytic Leukaemia Revenue Model Genasense Revenues (patients in thousands) 2004E 2005E 2006E 2007E 2008EUnited States - Second Line CLL CLL patients 50 50 50 50 50 % patients relapsing and receiving second line chemotherapy 25% 25% 25% 25% 25%Patients relapsing and receiving second line chemotherapy 13 13 13 13 13 Penetration 5% 10% 20% 30% 40%Number of patients on Genasense 0.6 1.3 2.5 3.8 5.0 Chemotherapy cycles per typical treatment 4 4 4 4 4 Annual Genasense courses in US (thousands) 3 5 10 15 20 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End User Revenues (US$m) US$5 US$10 US$20 US$30 US$40 Europe - Second Line CLL CLL patients 50 50 50 50 50 % patients relapsing and receiving second line chemotherapy 25% 25% 25% 25% 25%Patients relapsing and receiving second line chemotherapy 13 13 13 13 13 Penetration 4% 8% 15% 23% 30%Number of patients on Genasense 0.5 0.9 1.9 2.8 3.8 Chemotherapy cycles per typical treatment 4 4 4 4 4 Annual Genasense courses in Europe (thousands) 2 4 8 11 15 Cost of one Genasense course US$2,000 US$2,000 US$2,000 US$2,000 US$2,000 End user revenues (US$m) US$4 US$8 US$15 US$23 US$30 Total End User Sales (US$m) US$9 US$18 US$35 US$53 US$70Source: Bear, Stearns International Limited estimates.

Figure 24. Refractory Non-Small Cell Lung Cancer Revenue Model Genasense Revenues (patients in thousands) 2005E 2006E 2007E 2008EUnited States - First Line NSCLC Annual incidence of NSCLC 160 160 160 160 % with stage IIIb/IV NSCLC 50% 50% 50% 50%% stage IIIb/IV receiving first line chemotherapy 90% 90% 90% 90%Stage IIIb/IV receiving first line chemotherapy 72 72 72 72 Penetration 20% 30%Number of patients on Genasense 14.3 21.4 Chemotherapy cycles per typical first line treatment 6 6 Annual Genasense courses in US (thousands) 86 129 Cost of one Genasense course US$2,000 US$2,000 End User Revenues (US$m) US$172 US$257 Europe - First Line NSCLC Annual incidence of NSCLC 160 160 160 160 % with Stage IIIb/IV NSCLC 50% 50% 50% 50%% stage IIIb/IV receiving first line chemotherapy 90% 90% 90% 90%Stage IIIb/IV receiving first line chemotherapy 72 72 72 72 Penetration 15% 22%Number of patients on Genasense 10.7 16.1 Chemotherapy cycles per typical first line treatment 6 6 Annual Genasense courses in Europe (thousands) 64 96 Cost of one Genasense course US$2,000 US$2,000 End user revenues (US$m) US$129 US$193 Total End User Sales (US$m) US$300 US$450Source: Bear, Stearns International Limited estimates.



Figure 25. First-Line, Late-Stage Non-Small Cell Lung Cancer Revenue Model Genasense Revenues (patients in thousands) 2005E 2006E 2007E 2008EUnited States - Refractory NSCLC Refractory NSCLC patients 78 78 78 78 % refractory patients receiving chemotherapy 90% 90% 90% 90%Refractory patients receiving chemotherapy 70 70 70 70 Penetration 10% 20%Number of patients on Genasense 7.1 14.3 Chemotherapy cycles per typical treatment 4 4 Annual Genasense courses in US (thousands) 28 57 Cost of one Genasense course US$2,000 US$2,000 End User Revenues (US$m) US$57 US$114 Europe - Refractory NSCLC Refractory NSCLC patients 78 78 78 78 % refractory patients receiving chemotherapy 90% 90% 90% 90%Refractory patients receiving chemotherapy 70 70 70 70 Penetration 8% 15%Number of patients on Genasense 5.3 10.7 Chemotherapy cycles per typical treatment 4 4 Annual Genasense courses in Europe (thousands) 21 43 Cost of one Genasense course US$2,000 US$2,000 End user revenues (US$m) US$43 US$86 Total End User Sales (US$m) US$100 US$200Source: Bear, Stearns International Limited estimates.



Appendix 2: Genasense Patents and Intellectual Property

Intellectual Property

Aventis/Genta seem to have access to all the necessary intellectual property (IP) surrounding Genasense. The total royalty stack owed on these licences is 5%. The IP is licensed from two sources:

��University of Pennsylvania (Bcl-2, along with numerous other targets);

��National Institutes of Health (antisense technology).

Genta’s antisense patent portfolio covers compositions, methods of preparation, and methods of using antisense compounds that target Bcl-2 and its family members, including Genasense, both alone and in combination with other types of therapy for cancer and other diseases.

Genta also holds patents related to third-generation antisense compositions, based on its proprietary, chirally pure, methylphosphonate backbones. These linkages are highly resistant to degradation and enhance mRNA binding, thereby yielding more potent compounds. In addition, Genta has a patent estate of over 190 issued U.S. and foreign patents, covering in vivo methods of using antisense compounds for the selective down-regulation of mRNA.



Appendix 3: Antisense and the Case of Affinitak

Antisense Technology

Antisense drugs are small, chemically modified strands of DNA. These so-called oligonucleotides are engineered to binding specific mRNA. This mRNA-antisense duplex recruits an enzyme which degrades the mRNA portion of the duplex, inhibiting the production of the intended protein.

The advantages include a high specificity and selective down-regulation of single protein, reduced likelihood of adverse events because of specificity. They also may be the best approach for inhibition of difficult to reach proteins – e.g. perhaps no better way to get at a mitochondria membrane protein such as Bcl-2.

The drawbacks include limited numbers of mRNA targets and difficulty in identifying them and a large excess needed for efficacy, raising toxicity concerns. Further, they remain unproven clinically, and the stigma may be back again with the Affinitak failure.

Another Antisense Stumble

Prior to the phase III trial failure, Affinitak was thought to be the leading antisense molecule. Hopes were pinned on it to lead antisense into more important indications.

The FDA has approved only one antisense product, Vitravene from Isis. It is an inhibitor of ICAM-1 to treat cytomegalovirus (CMV) retinitis in AIDS patients intolerant of other therapies. There has been some controversy surrounding its mechanism. Some argue that it does not work by an antisense mechanism, but by immune stimulation through a non-specific binding event mediated by its ‘sticky’ phosphorothioate backbone.

PKC as a Target

Affinitak blocks production of a protein called protein kinase-C alpha (PKC-alpha). PKC-alpha is one of 11 PKC isoforms and part of a large network of signalling molecules that alter cell behaviour by modifying target molecules via the addition of phosphate groups.

Although PKC had been first recognised as a protein kinase, subsequent molecular cloning and biochemical studies have revealed that it belongs to a family of serine/threonine-specific protein kinases and is involved in diverse cellular functions, including cell proliferation, tumour promotion, differentiation, and apoptotic cell death and has been implicated in tumour growth and maintenance.

Many of the messages sent by PKC-alpha encourage growth of cancer cells. These same signals are used to a much lower extent by normal cells. Therefore, Affinitak hopes to reduce or eliminate the activity of PKC-alpha, inhibiting the growth of cancer cells while leaving other cells unharmed. However, the role of PKC-alpha is



not quite that simple. In some cell culture systems, PKC-alpha has both proliferation and anti-apoptotic functions (and in some cell lines, even apoptotic effects). In sum, its role in cancer is complicated and not entirely clear.

Results

In phase I and phase II clinical trials of patients with non-small cell lung carcinoma, patient survival time was increased significantly when Affinitak was added to standard chemotherapy. Further, only minimal side effects were caused by the antisense drug.

Isis’ antisense compound Affinitak, an inhibitor of PKC-alpha, was in a Phase III trial for non-small cell lung cancer. By targeting cells expressing abnormal levels of PKC-alpha Affinitak was thought to have therapeutic potential in a wide array of common cancers.

Affinitak for NSCLC failed in a phase III trial. It missed its primary endpoint badly, with only a 0.3 month improvement in survival (9.7 to 10 months, p=0.81). In other words, there is an 81% chance that it did nothing to help overall survival. Potential explanations include the placebo group surviving longer than expected. The median survival of the control group was longer than expected, in light of the fact that 87% of patients had Stage IV disease.

After an adjustment for duration of treatment and other predefined variables, the data was significantly better. It was almost significant for those who completed prescribed six cycles. Despite success in several secondary endpoints, Lilly has since pulled the plug on another phase III trial that was still ongoing.

Figure 26. Affinitak Phase III Results Primary Endpoint Result Number of Patients p value Primary log-rank analysis of the overall survival 0.3 month improvement; 10 months Affinitak vs. 9.7 months 616 in trial 0.81 Secondary Endpoint Result p value Adjusted log-rank analysis* Significat result, no futher details available 616 in trial 0.048 Completed Treatment subset survival analysis** 17.4 months Affinitak vs. 14.3 months control 256 in subset 0.054 Multiple other secondary endpoints Favored Affinitak, no further details available NA NA Source: Isis Pharmaceuticals; Bear, Stearns International Limited estimates. Notes: *Stratified log-rank statistical analysis that considered predefined variables, including duration of treatment. ** Survival subset analysis of 256 patients who completed treatment. The prescribed course of therapy was six cycles of treatment. Treatment groups in the trial were comparable with regard to all major prognostic factors.

Trial Details

Eli Lilly and Isis initiated a global Phase III trial in April 2002 to evaluate whether adding Affinitak to the already FDA-approved Gemzar-cisplatin combination helps patients with NSCLC to live longer. More than 1,000 patients with advanced, previously untreated NSCLC (no prior chemotherapy) were enrolled in the worldwide study. Participants in the randomised, open-label trial will receive either Gemzar and cisplatin or Gemzar and cisplatin in combination with Affinitak.

In addition to survival, Lilly evaluated that other measures of efficacy and safety are affected by the addition of Affinitak, including duration of progression-free survival,

AFFINITAK IS YET ANOTHER ANTISENSE FAILURE



time-to-treatment failure, tumour response rate, duration of tumour response and adverse events.

No difference was observed in a primary log-rank analysis of the overall survival of the two groups (p=0.81). Survival was the primary endpoint of the study. Patients receiving Affinitak plus the chemotherapy regimen of carboplatin and paclitaxel experienced a median survival of 10 months, compared to 9.7 months for patients receiving chemotherapy alone.

Other key findings from the trial of 616 chemotherapy naïve patients with Stage IIIb or Stage IV NSCLC are found in Figure 26.

Why it Doesn’t Matter so Much to Genasense

The key reason why Affinitak should not be lumped with Genasense is that the targets are substantially different. Bcl-2 promotes survival while PKC-alpha drives proliferation. Bcl-2 is a key and central mediator while PKC-alpha sits in one of many proliferation pathways. Bcl-2 is a good prognostic indicator while PKC-alpha is not.



Summary Financials

Figure 27. Aventis Pharmaceutical ProForma Sales, 2001-08E (€m) Brand Name Generic Name 2001A 2002A 2003E 2004E 2005E 2006E 2007E 2008ECardiovascular Lovenox enoxaparin 1,453 1,563 1,642 1,839 2,023 2,173 2,288 2,359Delix/ Tritace ramipril 709 923 1,018 1,018 1,018 866 737 663 cariporide 0 - - - 20 40 60 80Subtotal 2,162 2,486 2,660 2,857 3,061 3,079 3,084 3,103Respiratory Allegra / Telfast fexofenadine 1,761 2,030 2,048 1,863 1,077 781 692 661Nasacort triamcinolone

acetonide 266 329 330 348 367 387 406 427

Alvesco ciclesonide - - - 50 250 450 600 720Subtotal 2,027 2,359 2,378 2,261 1,694 1,618 1,699 1,808Oncology Taxotere docetaxel 1,002 1,261 1,513 1,833 2,175 2,564 2,962 3,346Campto irinotecan 202 241 265 292 321 353 388 427 flavopiridol - - - - - 20 40 60 taxoid 109881 - - - - - - 50 100Genasense - - - 5 20 40 60 80Subtotal 1,204 1,502 1,778 2,129 2,516 2,977 3,500 4,013CNS Copaxone glatiramer actetate 382 554 637 582 864 989 1,080 1,178Rilutek riluzole 118 118 50 50 50 50 50 50 100,907 - - - - - - 20 40Subtotal 500 672 687 632 914 1,039 1,150 1,268Diabetes & Endocrine Amaryl glimepiride 478 578 600 665 692 692 699 719Insuman insulin 170 172 163 155 147 140 133 126Lantus insulin glargine 94 299 477 642 771 865 972 1095Exubera inhaled insulin - - - - - - - - Insulin Glulisine - 80 200 350 450 525 DiaPep277 - - 20Subtotal 742 1,049 1,241 1,542 1,810 2,048 2,254 2,486Anti-Infectives Targocid teicoplanin 199 222 232 244 256 269 282 296Tavanic levofloxacin 192 257 239 287 315 347 382 420Ketek telithromycine 20 52 121 420 570 680 780 830Picovir pleconaril - - - - - - -Subtotal 411 530 592 950 1,141 1,296 1,444 1,546Arthritis & NSAIDS Arava leflunamide 258 271 286 339 380 423 478 506 pranalacasan - - - - - - - -Subtotal 258 271 286 339 380 423 478 506Total Strategic Products 7,304 8,870 9,623 10,712 11,518 12,479 13,609 14,729 Tail (other prescription drugs) 7,864 7,157 5,956 5,717 5,495 5,281 5,076 4,879 Total Prescription Pharmaceuticals 15,168 16,026 15,578 16,429 17,013 17,760 18,685 19,608 Vaccines 1,425 1,580 1,722 1,894 2,046 2,210 2,386 2,577 Elimination -17 -15.5 -15.5 -15.5 -15.5 -15.5 -15.5 -15.5 Total Core business 16,576 17,591 17,285 18,308 19,043 19,955 21,056 22,170Y-o-Y Growth 7.8% -1.7% 5.9% 4.0% 4.8% 5.5% 5.3% Non-core business 6,439 3,066 1,164 1,198 1,234 1,270 1,308 1,347 Elimination -74 -35 - - - - - - Total Group Sales 22,941 20,622 18,449 19,506 20,277 21,225 22,365 23,517 Y-o-Y Growth N a -10% -11% 6% 4% 5% 5% 5%1) 2000 is restated in accordance with new structure (i.e. ex-Behring) 2) Tail includes Refludan in 2000 and 2001 (€27 and €30, respectively) Source: Company data; Bear, Stearns International Limited estimates.



Figure 28. Aventis – Income Statement, 2000-08E (€m) 01-02E 02-03E 02-04E 02-05E 02-06E 02-07E 02-08EKey Growth Rates (CAGR) Sales -10.1% -10.5% -5.4% -0.6% 0.7% 1.6% 2.2%Pharma sales 5.7% -2.8% 5.5% 2.0% 2.6% 3.1% 3.4%EBITDA -10.8% 19.6% 34.5% 14.0% 11.4% 10.5% 9.9%EPS 37.9% 6.8% 34.8% 16.0% 13.9% 13.4% 13.1%Dividends 20.8% -0.2% 25.9% 9.9% 9.9% 9.8% 11.8% Group Business 2000A 2001R 2002A 2003E 2004E 2005E 2006E 2007E 2008ESales 22,304 22,941 20,622 18,449 19,506 20,277 21,225 22,365 23,517Y-o-Y increase 60% na -10% -11% 6% 4% 5% 5% 5%COGS (8,469) (7,943) (6,578) (4,965) (5,177) (5,488) (5,729) (5,971) (6,218)Gross Profit 13,835 14,998 14,044 13,484 14,329 14,789 15,496 16,393 17,300R&D (3,479) (3,481) (3,420) (3,119) (3,257) (3,290) (3,430) (3,617) (3,805)SG&A (8,987) (7,228) (6,773) (5,871) (6,121) (6,024) (6,377) (6,693) (6,975)Goodwill amortisation (752) (650) (1,021) (518) (518) (518) (518) (518) (518)Operating profit 617 3,639 2,830 3,976 4,433 4,958 5,172 5,566 6,001Operating profit before Goodwill amortisation 1,369 4,289 3,851 4,494 4,950 5,475 5,690 6,084 6,519% of sales 6% 19% 19% 24% 25% 27% 27% 27% 28%Y-O- Y increase -43% na -10% 17% 10% 11% 4% 7% 7%Equity Income 244 85 51 170 296 309 322 336 351EBITA 1,613 4,374 3,901 4,664 5,247 5,784 6,012 6,420 6,870% of sales 7% 19% 19% 25% 27% 29% 28% 29% 29%Y-O- Y increase -38% na -11% 20% 12% 10% 4% 7% 7%Depreciation & Amortisation of Intangibles (1,674) (1,425) (1,195) (1,042) (1,025) (1,018) (1,018) (1,025) (1,039)EBITDA 3,287 5,799 5,096 5,706 6,271 6,803 7,030 7,445 7,909% of sales 15% 25% 25% 31% 32% 34% 33% 33% 34%Y-O- Y increase -1% na -12% 12% 10% 8% 3% 6% 6%Interest (805) (704) (309) (203) (161) (74) 20 119 222Miscellaneous non operating income (expenditure) (81) (134) 1,120 (283) (84) (100) (100) (100) (100)Pretax income (25) 2,886 3,692 3,661 4,484 5,093 5,414 5,921 6,474Y-o-Y increase -62% na 33% -11% 20% 12% 6% 9% 9%tax rate 240% -38% -39% -35% -35% -34% -34% -34% -34%Taxes (60) (1,111) (1,430) (1,297) (1,552) (1,741) (1,841) (1,999) (2,170)Minorities 56 (142) (86) (54) (64) (74) (84) (94) (104)Preferred remuneration (118) (128) (85) (85) (85) (85) (85) (85) (85)Net income before exceptionals (147) 1,506 2,092 2,225 2,782 3,192 3,404 3,744 4,115Y-o-Y increase -121% na 39% 6% 25% 15% 7% 10% 10%No. of shares (average) 781 787 793 790 783 774 766 756 745EPS (0.19) 1.91 2.64 2.82 3.56 4.12 4.45 4.95 5.52Y-o-Y increase 38% 7% 26% 16% 8% 11% 12%Net income pre goodwill 605 2,156 3,113 2,742 3,300 3,710 3,922 4,261 4,632Amortisation-adjusted EPS 0.78 2.74 3.92 3.47 4.22 4.79 5.12 5.64 6.22Y-o-Y increase na na 43% -12% 21% 14% 7% 10% 10% Margin Analysis 2000A 2001R 2002A 2003E 2004E 2005E 2006E 2007E 2008EGross Profit 62.0% 65.4% 68.1% 73.1% 73.5% 72.9% 73.0% 73.3% 73.6%Selling and Distribution costs 40.3% 31.5% 32.8% 31.8% 31.4% 29.7% 30.0% 29.9% 29.7%R&D 15.6% 15.2% 16.6% 16.9% 16.7% 16.2% 16.2% 16.2% 16.2%Operating margin 6.1% 18.7% 18.7% 24.4% 25.4% 27.0% 26.8% 27.2% 27.7%EBITA 7.2% 19.1% 18.9% 25.3% 26.9% 28.5% 28.3% 28.7% 29.2%EBITDA 14.7% 25.3% 24.7% 30.9% 32.1% 33.5% 33.1% 33.3% 33.6%Tax Rate -240.0% 38.5% 38.7% 35.4% 34.6% 34.2% 34.0% 33.8% 33.5% Equity Income 2000A 2001A 2002A 2003E 2004E 2005E 2006E 2007E 2008EPrescription Drugs (23) 6 26 27 28 29 30 32 33Human Vaccines 64 39 25 26 27 28 29 30 32Corporate and Animal Health 122 169 157 151 157 163 170 176 183Total Core Equity Earnings 163 214 208 204 212 220 229 238 248% Annual growth Prescription Drugs 0% na 333% 4% 4% 4% 4% 4% 4%Human Vaccines 0% -39% -36% 4% 4% 4% 4% 4% 4%Corporate and Animal Health 0% 39% -7% -4% 4% 4% 4% 4% 4%Other Activities 81 (129) (157) (33) 84 89 93 98 103Total Non-Core Equity Earnings 81 (129) (157) (33) 84 89 93 98 103Total Equity Earnings 244 85 51 170 296 309 322 336 3511) 2001 figures restated to reflect new definition of ‘Core Business’ (excluding Aventis Behring). 2000 as previously reported for core business Source: Company data; Bear, Stearns International Limited estimates.



Figure 29. Aventis – Cashflow Statement, 2000-08E (€m) 2000A 2001R 2002A 2003E 2004E 2005E 2006E 2007E 2008ENet income (loss) before preferred remuneration (29) 1,634 2,177 2,310 2,867 3,277 3,489 3,829 4,200Depreciation & Amortisation 2,426 2,075 2,216 1,600 1,582 1,576 1,576 1,583 1,597Change in provision 176 (73) 1,053 (1,972) (514) (363) (252) (162) (87)Cash injected into pension fund - - - (289) (170) (170) (170) (170) (170)Other non-cash items - - - 61 30 - - - -Net capital gains/ (losses) from sale of assets (359) (545) (2,187) - - - - - -Equity in net income of affiliated companies, net of dividends

312 89 114 121 128 136 144 153 162

Unrealised exchange differences (230) (111) (2) - - - - - -Minority interests (56) 142 86 54 64 74 84 94 104Deferred tax (772) 40 143 - - - - - -Change in Working Capital (197) (137) (1,740) (147) (197) (143) (176) (212) (214)Cashflow from Operations 1,271 3,114 1,860 1,737 3,791 4,387 4,694 5,115 5,592Purchase of PP&E (1,570) (1,245) (1,000) (983) (1,041) (1,083) (1,134) (1,197) (1,260)Other Capex (932) (486) (459) - - - - - -Proceeds from sale of assets 1,091 1,061 4,654 - - - - - -Net change in loans and short-term investments (30) (52) 44 (787) 153 - - - -Cashflow from Investing (1,441) (720) 3,239 (1,769) (887) (1,083) (1,134) (1,197) (1,260)New long-term debt 2,281 5,404 135 1,000 - - - - -Repayment of Long-term debt (2,019) (7,252) (2,931) - - (1,000) (1,000) (1,000) (500)Increase/ (decrease) of short-term debt (305) (284) (1,091) - - - - - -Issuance of ordinary shares / (share buyback) 367 429 199 (350) (486) (620) (735) (901) (1,105)Mandatory redeemable partnership interest - 279 - - - - - - -Repurchase of treasury shares - (137) (383) - - - - - -Amortisation of amortizable preferred securities (89) (85) (122) (85) (85) (85) (85) (85) (85)Purchase of minority interest (146) (5) (212) - - - - - -Dividends paid (453) (437) (490) (560) (554) (692) (794) (846) (930)Preferred remuneration paid (96) (109) (113) (85) (85) (85) (85) (85) (85)Cashflow from Financing (460) (2,197) (5,008) (80) (1,210) (2,482) (2,699) (2,917) (2,705)Net effect of exchange rate changes on cash - 15 (60) - - - - - -Increase/ (decrease) in net cash and equivs. - 211 31 (112) 1,694 822 861 1,001 1,626Net effect of consolidation changes on cash and equivs. - (58) (88) - - - - - - Cash at beginning of year 1,290 661 814 757 645 2,339 3,162 4,022 5,023Cash at year end 1,290 814 757 645 2,339 3,162 4,022 5,023 6,649Cash at year end incl. marketable securities 1,908 1,515 1,299 1,188 2,882 3,705 4,565 5,566 7,192Source: Company data; Bear, Stearns International Limited estimates.



Figure 30. Aventis – Balance Sheet, 2000-08E (€m) 2000A 2001R 2002A 2003E 2004E 2005E 2006E 2007E 2008E Cash & short-term deposits 661 814 756 645 2,339 3,162 4,022 5,023 6,649Marketable securities 618 701 543 543 543 543 543 543 543Accounts receivable 3,469 3,522 2,544 2,398 2,536 2,636 2,759 2,907 3,057Inventories 4,118 4,059 2,730 2,583 2,731 2,839 2,972 3,131 3,292Other current assets 5,146 3,689 3,073 2,767 2,926 3,042 3,184 3,355 3,528Total current assets 14,012 12,785 9,646 8,936 11,075 12,221 13,480 14,959 17,070 PP&E 7,498 5,740 4,455 4,633 4,854 5,092 5,356 5,653 5,981Intangibles 2,163 1,881 1,514 1,298 1,119 970 822 697 590Goodwill 12,659 12,383 9,630 9,112 8,595 8,077 7,559 7,041 6,524Investments 3,322 2,868 2,406 2,406 2,406 2,406 2,406 2,406 2,406Other long-term assets 2,529 3,577 3,422 3,422 3,422 3,422 3,422 3,422 3,422Total Assets 42,183 39,234 31,073 29,808 31,470 32,188 33,045 34,178 35,992 Short-term debt 6,197 6,058 2,965 2,965 2,965 2,965 1,965 965 465Accounts Payable 2,403 2,421 1,415 1,236 1,307 1,359 1,422 1,498 1,576Other current liabilities (ex-provisions) - 3,952 3,648 3,081 3,257 3,386 3,545 3,735 3,927Current provisions (ex-current pension liabilities) 6,511 1,508 2,450 - - - - - -Total current liabilities 15,111 13,939 10,478 7,282 7,529 7,710 6,932 6,198 5,968 Long-term Debt 8,216 4,652 1,787 1,787 1,787 787 787 787 787- new debt - - - 469 558 500 450 432 443Pension Provisions 3,242 3,350 3,328 3,548 3,674 3,811 3,957 4,115 4,284Restructuring provisions - 99 77 77 62 49 39 32 25Other provisions 2,714 2,682 2,556 2,556 2,045 1,636 1,309 1,047 838Deferred income tax 1,038 1,094 1,026 1,026 1,026 1,026 1,026 1,026 1,026Redeemable partnership interests - 284 238 238 238 238 238 238 238Minority interest 1,029 913 159 159 159 159 159 159 159Amortizable preferred securities 272 200 89 16 - - - - -Shareholders’ equity 10,561 12,021 11,335 12,650 14,392 16,273 18,147 20,144 22,224Total liabilities and shareholders’ equity 42,183 39,234 31,073 29,808 31,470 32,188 33,045 34,178 35,992Source: Company data; Bear, Stearns International Limited estimates.



Pipeline Summary

Figure 31. Aventis Pharmaceuticals Pipeline (ex-vaccines) Compounds (Brand & Generic Name)

Mechanism Indication Est. Launch

Comment

Filed Telithromycin (Ketek)

Inhibits bacterial reproduction via the inhibition of protein synthesis

Community-acquired pneumonia / acute bacterial exacerbations of chronic bronchitis / acute sinusitis / tonsillitis

4Q 03 Launched in Europe (Germany, Italy, Spain). Discussions with FDA about safety data and resistance claims, 24,000 pts study completed, submitted on 1 June 2002, accepted by FDA on 21 August 2002. Second approvable letter received on 24 January 2002, which asks for additional post-marketing data from Europe and a re-analysis of certain existing data. The FDA’s concern is surrounding the safety profile of the drug, with market observers suggesting, visual side effects, in particular.

GA-EPO (Dynepo) Recombinant erythropoeitin - stimulates red blood cell production

Anaemia ?? Depending on outcome of patent litigation with Amgen.

Phase III Ciclesonide (Alvesco)

Anti-inflammatory glucocorticoid Asthma, COPD 2004 In-licensed for US market from Altana. Complete Phase III in 2002.

Cariporide NHE inhibitor - prevents myocardial cell from isochemic damage

CABG/AMI 2004

Genasense Anti-sense technology against bcl-2 mRNA

CLL / melanoma / myeloma (initially)

2004 Licensed from Genta. Filing of initial indications end-2002 / early 2003. In process of finalising P3 trials for NSCLC, NHL, breast cancer, colorectal cancer.

Insulin glulisine Fast-acting insulin analogue Diabetes, type I & II 2004 NDA / MAA submission planned for mid-2003. Inhaled insulin (Exubera)

Inhaled version of recombinant insulin

Diabetes type I & II 2006 At 2002 R&D day, preliminary data showed that the 1% reduction in FEV1 lung volumes, identified after six months, had stabilised after one year. Looking at expanding long-term safety database, time lines uncertain. On FY 2002 PFE conference call, PFE said it did not expect to file Exubera in 2004.

Phase IIb Flavopiridol (1275) Cyclin-dependent kinase inhibitor,

causes tumour cell cycle arrest and cell death

CLL/ NSCLC/ brain metastasis 2006 Data expected at ASCO 2002, no news for two years due to development of new administration regimes.

Taxoid 109881 / 106258

Kills taxoid-resistant tumour cell lines

Brain metastasis 2007 At June 2002 R&D day, it said response rates for 109881 are not stunning. 106258 is a back-up product.

DiaPep277 Prevents destruction of beta cells in pancreas

Type I and LADA diabetes ?? Licensed in from Peptor in July 2002. Have a Phase II trial complete in Type I – but additional trials ongoing. Still awaiting Phase IIa LADA. In discussions with FDA on Phase III endpoints.

100240 Dual inhibition of ACE and neutral endopeptidase

Hypertension/CHF ?? Phase IIb trials expected second-half 2002. Currently have back-up – AVE7688 in phase I.

Phase IIa 100907 5 - HT 2 A antagonist Insomnia 2007 Phase I data appears to show improvement in ‘quality of sleep’.

Previously under development for schizophrenia. Pralnasacan (3480/VX-740)

ICE inhibitor Rheumatoid arthritis > 2007 Vertex-collaboration

VLA 4-antagonist Prevents movement and inhibits activation of white blood cells in the lung

Asthma ?? Not confident of status for product. Failed to show Proof of Concept in Phase Iia

1426 Anti-obesity agent

Delays gastric emptying Anti-obesity agent ?? Start phase IIb in 2002. Back-up product 1954 currently being developed.

NV1FGF Induction of new blood vessel formation

Peripheral artery disease > 2007

Teriflunomide (1726) Multiple sclerosis > 2007 Enrollment recently completed Source: Company data.



Figure 32. Aventis Pharmaceuticals Pipeline (ex-vaccines) – (continued) Compounds

(Brand & Generic Name)

Mechanism Indication Est. Launch

Comment

Phase I IPL-512,602 Leukocyte suppressing anti-

inflammatory compound (LSAID) Asthma > 2007 Licensed from Inflazyme. Product is in P1b. The first-generation

product, IPL-576,092, is in PIIb – it showed non-statistically significant benefit in terms of lung FEV1 volumes. This second-generation compound had a better pre-clinical profile. However, clinical confirmation is required.

4011 / 0547 Non-corticosteroidal oral anti inflammatory agent

Asthma > 2007 Aventis has AVE-0547 as a back-up.

AVE8062 Cuts off blood supply to tumour Cancer > 2007 In-licensed from Ajinmoto & Co. Novel anti-tumour strategy. AVE 0545 Non-steroidal anti-inflammatory Rheumatism > 2007 Developed in collaboration with Millennium. AVE 5997 Dopamine D3 antagonist Schizophrenia > 2007 LIT 976 Prevents cancer cell division by

inhibiting cytoskeletal disassembly

Cancer (follow-up of taxotere) > 2007

AVE 6971 Topo IV inhibitor Broad spectrum antibiotic > 2007 P1 to start fourth-quarter 2003. 1766 Guanylate Cyclase Activator

Coronary vasodilator for chronic angina

Angina > 2007 First-in-class of NO-independent sGC activators. AVE 1069 is being developed as a back-up.

AVE-4579 CRF receptor blocker Stress / depression > 2007 P1 to start fourth-quarter 2002. HP-184 Spinal cord Injury > 2007 SERM 3471 Selective oestrogen receptor

modulator Treatment and prevention of post-menopausal osteoporosis

> 2007 Follow-on product from SERM 3399.

HMR 2371 Stroke > 2007 No information available recently. HMR 2906 Anti-thrombotic > 2007 No information available recently. Factor Xa Inhibitor Thrombolysis > 2007 No information available recently. Source: Company data.



Companies mentioned:

Abbott (ABT – US$42.25) Outperform

Akzo Nobel (AZKOY – US$23.74)

AstraZeneca (AZN.L – GBp2491) Peer Perform

Avecia (AVGUP – US$22.29)

AVI BioPharma (AVII – US$5.51)

BASF (BASF – €37.00)

Baxter (BAX – US$22.52) Outperform

Bayer (BAY.DE – €16.91)

Bristol-Myers Squibb (BMY – US$25.59) Peer Perform

British Biotech (BBG.L – GBp58.50)

Ciba SC (CSB – US$35.12

Clariant (CLN.S – SFr13.90)

CSL (CSJ.DE – Eur8.15)

Elan (ELN – US$4.45)

Eli-Lilly (LLY – US$64.50) Underperform

Enzo Biochem (ENZ – US$15.25)

Genentech (DNA – US$36.25) Peer Perform

Genta (GNTA.O – US$8.32)

GlaxoSmithKline (GSK.L – GBp1325) Peer Perform

Hybridon (HYBN – US$0.91)

ImClone (IMCL.O – US$20.84)

Immune Response (IMNR – US$1.34)

Inex (IEX.CA – CA$4.14)



Introgen Theraupeutics (INGN – US$2.66)

Isis Pharmaceuticals (ISIS – US$4.95)

Lorus Therapeutics (LOR.CA – CA$1.42)

Medtronic (MDT – US$46.50) Peer Perform

Merck KgaA (MRK.DE – US$24.40)

MGI Pharma (MOGN – US$15.01) Outperform

NeoPharm (NEOL – US$15.10)

Novartis (NOVZn.VX – SFr53.20) Outperform

OSI Pharmaceuticals (OSIP – US$) Outperform

QLT (QLT.CA – CA$17.09)

Roche (ROCZg.VX – SFr85.60) Peer Perform

Sanofi-Synthélabo (SASY.PA – €54.60) Peer Perform

Shire Pharmaceuticals (SHP.L – GBp427.75) Outperform

Solvay (SOL.DE – Eur64.00)

Syngenta (SYN.S – SFr69.80)

Taisho (4535.JP – Yen1719)

Wyeth (WYE – US$42.50) Peer Perform

Xenova (XEN.L – GBp16.75) Not Rated

383 Madison Avenue

New York, New York 10179

US Atlanta, Georgia 3424 Peachtree Road N.E. Suite 1700 Atlanta, GA 30326 +404-842-4000

Boston, Massachusetts One Federal Street Boston, MA 02110 +617-654-2800

Chicago, Illinois Three First National Plaza Chicago, IL 60602 +312-580-4000

Dallas, Texas 300 Crescent Court Suite 200 Dallas, TX 75201 +214-979-7900

Denver, Colorado 3200 Cherry Creek South Drive Suite 260 Denver, CO 80209 +720-570-2327

Los Angeles, California 1999 Avenue of the Stars Los Angeles, CA 90067 +310-201-2600

San Francisco, California Citicorp Center One Sansome Street 41st Floor San Francisco, CA 94104 +415-772-2900

Hato Rey, Puerto Rico 270 Muñoz Rivera Avenue 5th Floor Hato Rey, Puerto Rico 00918 +787-753-2327

Europe Dublin, Ireland Bear Stearns Bank plc Block 8 Harcourt Centre Charlotte Way 3rd Floor Dublin 2 Ireland +353-1-402-6200

London, England Bear, Stearns International Limited One Canada Square London E14 5AD England +44-20-7516-6000

Lugano, Switzerland Bear, Stearns & Co. Inc. Corso Elvezia 14 P.O. Box 2168 6901 Lugano, Switzerland +41-91-911-7333

Milan, Italy Bear Stearns International Ltd. Via Pietro Verri 6 20121 Milano, Italy +39-02-3030-1700

Herzliya, Israel Bear, Stearns & Co. Inc. Representative Office Ackerstein Towers - East Wing 11 Hamenofim Street - 9th Floor Herzliya 46725 Israel +972-9-951-2770

Latin America São Paulo, Brazil Bear Stearns do Brasil Ltda. Rua Joaquim Floriano, 72 8 andar (floor) - cj 83-Itaim Bibi (area) 04534-000 São Paulo, Brazil +55-11-3457-3200

Asia

Beijing, China Bear, Stearns & Co. Inc. Beijing Representative Office Unit 1608, China World Tower 1 Jian Guo Men Wai Avenue Beijing 100004 People’s Republic of China +86-10-6505-5101

Hong Kong Bear Stearns Asia Limited 26th Floor, Citibank Tower Citibank Plaza 3 Garden Road, Hong Kong +852-2593-2700

Shanghai, China Bear, Stearns & Co. Inc. Shanghai Representative Office Shanghai Senmao International Building 101 Yin Cheng East Road 27th Floor Pudong, Shanghai 200120 People’s Republic of China +86-21-6219-2642

Singapore Bear Stearns Singapore Pte. Limited30 Raffles Place #21-00 Caltex House Singapore 048622 +65-6437-3300

Tokyo, Japan Bear Stearns (Japan) Ltd. Shiroyama JT Trust Tower 22/F 3-1 Toranomon 4-Chome, Minato-ku Tokyo 105-6022 Japan +813-3437-7800

This report has been prepared by Bear, Stearns & Co. Inc., Bear, Stearns International Limited or Bear Stearns Asia Limited (together with their affiliates, “Bear Stearns”), as indicated on the cover page hereof. This report has been adopted by Bear, Stearns & Co. Inc. for distribution in the United States. If you are a recipient of this publication in the United States, orders in any securities referred to herein should be placed with Bear, Stearns & Co. Inc. This report has been approved for publication in the United Kingdom by Bear, Stearns International Limited, which is regulated by the United Kingdom Financial Services Authority. Private Customers in the U.K. should contact their Bear, Stearns International Limited representatives about the investments concerned. This report is distributed in Hong Kong by Bear Stearns Asia Limited, which is regulated by the Securities and Futures Commission of Hong Kong. Additional information is available upon request.

Bear Stearns may be associated with the specialist that makes a market in the common stock or options of an issuer in this report, and Bear Stearns or such specialist may have a position (long or short) and may be on the opposite side of public orders in such common stock or options.

Bear Stearns and its employees, officers and directors may have positions and deal as principal in transactions involving the securities referred to herein (or options or other instruments related thereto), including positions and transactions contrary to any recommendations contained herein. Bear Stearns and its employees may also have engaged in transactions with issuers identified herein.

This publication does not constitute an offer or solicitation of any transaction in any securities referred to herein. Any recommendation contained herein may not be suitable for all investors. Although the information contained herein has been obtained from sources we believe to be reliable, its accuracy and completeness cannot be guaranteed. This publication and any recommendation contained herein speak only as of the date hereof and are subject to change without notice. Bear Stearns and its affiliated companies and employees shall have no obligation to update or amend any information contained herein.

This publication is being furnished to you for informational purposes only and on the condition that it will not form a primary basis for any investment decision. Each investor must make its own determination of the appropriateness of an investment in any securities referred to herein based on the legal, tax and accounting considerations applicable to such investor and its own investment strategy. By virtue of this publication, none of Bear Stearns or any of its employees shall be responsible for any investment decision. © 2003. All rights reserved by Bear Stearns.

This report may discuss numerous securities, some of which may not be qualified for sale in certain states and may therefore not be offered to investors in such states.

NOTE TO ACCOUNT EXECUTIVES: For securities that are not listed on the NYSE, AMEX or NASDAQ National Market System, check the Compliance page of the Bear Stearns Intranet site for State Blue Sky data prior to soliciting or accepting orders from clients.

Disclosures Ratings for Stocks (vs. analyst coverage universe)

Outperform (O) — Stock is projected to outperform analyst's industry coverage universe over the next 12 months.

Peer Perform (P) — Stock is projected to perform approximately in line with analyst's industry coverage universe over the next 12 months.

Underperform (U) — Stock is projected to underperform analyst's industry coverage universe over the next 12 months.

Ratings for Sectors (vs. regional broader market index)

Market Overweight (MO) — Expect the industry to perform better than the primary market index for the region over the next 12 months.

Market Weight (MW) — Expect the industry to perform approximately in line with the primary market index for the region over the next 12 months.

Market Underweight (MU) — Expect the industry to underperform the primary market index for the region over the next 12 months.

Bear, Stearns & Co. Ratings Distribution as of April 14, 2003: (%Rated companies / %Banking client in the last 12 months) Outperform: 37.0 / 24.8 Peer Perform: 44.7 / 17.7 Underperform: 17.3 / 9.6 Not Rated: 1.0 / 20.0

IMNR; ISIS: Bear, Stearns & Co. Inc. is a market maker in this company's equity securities. BMY: The analyst(s) that prepared this report or a member of the analyst's household has a long position in the equity securities of this company. ELN, ABT: Bear Stearns is associated with the specialist that makes a market in the common stock of this issuer, and such specialist may have a position (long or short) and may be on the opposite side of public orders in such common stock. GSK.L; MDT: Bear Stearns is associated with the specialist that makes a market in the options of this issuer, and such specialist may have a position (long or short) and may be on the opposite side of public orders in such options. The costs and expenses of Equity Research, including the compensation of the analyst(s) that prepared this report, are paid out of the Firm’s total revenues, a portion of which is generated through investment banking activities.

The price chart may be found on page 2 of this report.

The Research Analyst(s) who prepared the research report hereby certify that the views expressed in this research report accurately reflect the analyst(s) personal views about the subject companies and their securities. The Research Analyst(s) also certify that the Analyst(s) have not been, are not, and will not be receiving direct or indirect compensation for expressing the specific recommendation(s) or view(s) in this report.

Aventis (AVE.PA)

OutperformBuy

30

50

70

90

110

130

Aug-01 Nov-01 Feb-02 May-02 Aug-02 Nov-02 Feb-03Rating & target data valid through April 11, 2003.

Stock Prices: FactSet Research Systems; Recommendations: Bear, Stearns Equity Research

Eu

ro

Close Rating 12 month target

DIRECTOR OF GLOBAL RESEARCH Kay Booth (212-272-4205)

MACRO Accounting & Tax Policy Pat McConnell, CPA (212-272-4193) Janet Pegg, CPA (212-272-4191)

Convertible Research Rao Aisola (212-272-5988) Thomas Sugiura (212-272-9194) John P. Wright (212-272-9522)

Economics, Global David Malpass (212-272-4293) Tim Kearney, Ph.D. (212-272-5981) Economics, Market John Ryding (212-272-4221) Conrad DeQuadros (212-272-4026) Emerging Markets Portfolio Strategy Thierry Wizman, Ph.D. (212-272-7527) Equity Derivatives Wing C. Chow (212-272-5150) Charles G. Galligan (212-272-4106) Gary J. Semeraro (212-272-4800) Investment Strategy François Trahan (212-272-2103)

Risk Arbitrage David Freedman, CFA (212-272-4209) John Lammers (212-272-6411) Scott A. Schefrin (212-272-8280)

Small Companies Peter J. Barry (212-272-6090)

U.S. RESEARCH

BASIC INDUSTRY Aerospace & Defense Electronics Steve Binder, CFA (212-272-4235) Airlines David Strine (212-272-7869) Airfreight & Surface Transportation Edward M. Wolfe (212-272-7048)

Auto Domenic D. Martilotti (212-272-9100) Chemicals/Commodity Frank J. Mitsch (212-272-5600) Engineering & Construction/ Precious Metals Michael S. Dudas, CFA (212-272-2289) Environmental Services Steve Binder, CFA (212-272-4235) Industrial Manufacturing Scott Graham (212-272-6691) Metals & Mining Anthony B. Rizzuto, Jr. (212-272-4269) Railroads Thomas R. Wadewitz (212-272-3169)

CONSUMER Beverages Carlos A. Laboy (212-272-3203) Discretionary Consumer Products Joseph J. Yurman (212-272-2519) Food Terry Bivens (212-272-6782) Hard Line Retailers Dana L. Telsey (212-272-6052) Hotels & Gaming Mark Abramson (212-272-4305) Leisure R. Glen Reid (212-272-6636) Restaurants Joseph T. Buckley, CFA (212-272-4263)

CONSUMER (CONT’D) Retailing/Department Stores & Broadlines Christine K. Augustine (212-272-2276) Specialty Stores Dana L. Telsey (212-272-6052)

ENERGY Coal Michael S. Dudas, CFA (212-272-2289) Electric Utilities David Fore (212-272-8700) Integrated Oil Frederick P. Leuffer, CFA (212-272-6344) Oil & Gas E&P Ellen K. Hannan (212-272-4199) Oil Service & Equipment Robin Shoemaker (212-272-6475)

FINANCIAL SERVICES Banks/Large-Cap David Hilder (212-272-7507) Banks/Mid-Cap Salvatore J. DiMartino (212-272-2219) Brokers & Asset Managers Daniel C. Goldberg (212-272-2945) REITs Ross L. Smotrich (212-272-8046) Specialty Finance/Mortgage Finance David Hochstim, CFA (212-272-4243)

HEALTH CARE Biotechnology Ronald C. Renaud (212-272-9918) Akhtar Samad, M.D., Ph.D. (212-272-2722) Health Care Technology & Distribution Raymond G. Falci (212-272-4245) Major Pharmaceuticals Joseph P. Riccardo (212-272-4253) Scott J. Shevick, CFA (212-272-4301) Managed Care John F. Rex (212-272-4242) Medical Supplies & Devices Frederick A. Wise, CFA (212-272-4265) Pharmaceutical Services Stephen Unger, CFA (212-272-5173)

MEDIA Advertising & Marketing Services Alexia Quadrani (212-272-2149) Cable Raymond Lee Katz (212-272-6857) Education Services Jennifer L. Childe, CFA (212-272-7051) Entertainment Raymond Lee Katz (212-272-6857) Interactive Entertainment Jeff Vilensky (212-272-5251) Publishing & Information Services Kevin R. Gruneich, CFA (212-272-5867) Radio & TV Broadcasting Victor B. Miller IV (212-272-4233) Satellite Communications Robert S. Peck, CFA (212-272-6665)

TELECOMMUNICATIONS Satellite Communications Robert S. Peck, CFA (212-272-6665) Wireless Equipment Wojtek Uzdelewicz (212-272-2139)

Wireline Equipment Wojtek Uzdelewicz (212-272-2139)

TELECOMMUNICATIONS (CONT’D) Wireless Services Phil Cusick (212-272-9078) Wireless Transmission Towers Jim Ballan (212-272-8024) Wireline Services Robert Fagin (212-272-4321)

TECHNOLOGY Business/Professional Services Andrew Steinerman (212-272-5557) Business/Professional Services James F. Kissane (212-272-2832) Computer Services & IT Consulting James F. Kissane (212-272-2832) Electronics Manufacturing Services Thomas Hopkins (212-272-3719) Enterprise Hardware Andrew J. Neff (212-272-4247) Internet Jeffrey Fieler, CFA (212-272-9629) PC Hardware Andrew J. Neff (212-272-4247) Semiconductors Brian Wu (415) 772-3028 Supply-Chain Logistics Software Edward M. Wolfe (212-272-7048) Philip Alling (212-272-6905)

EAST ASIA RESEARCH Airlines/Infrastructure David Strine (212-272-7869) Economics/Strategy Michael Kurtz [852] 2593-2736 Hotels Mark Abramson (212) 272-4305 Oil & Gas Adam Clarke [852] 2593-2745 Property/Conglomerates Vincent Luk [852] 2593-2747 Technology Gurinder Kalra [852] 2593-2742 Telecom Services Evan Erlanson, CFA [852] 2593-2785 Utilities Bradley Burch [852] 2593-2734

EUROPEAN RESEARCH Director of European Research Roger Hirst (44-20) 7516-6026

CONSUMER Hotels & Gaming Anna Barnfather (44-20) 7516-6767 Mark Abramson (44-20) 7516-6002

FINANCIAL SERVICES Insurance/Multi-Line Paul Goodhind (44-20) 7516-5192 Reinsurance David Cavallanti (44-20) 7516-5360 Specialty & Other Financials Tony Cummings (44-20) 7516-6360

HEALTH CARE Biotechnology Simon Conway (44-20) 7516-5234

EUROPEAN RESEARCH (CONT’D)

HEALTH CARE Pharmaceuticals Alexandra Hauber (44-20) 7516-5347

MEDIA European Broadcast/Consumer Nicholas Bell (44-20) 7516-6977

TECHNOLOGY Hardware Axel Funhoff (44-20) 7516-5236 Software & Services Toby McCullagh (44-20) 7516-5381

TELECOMMUNICATIONS Wireless Fanos Hira (44-20) 7516-6791 Wireline Jonathan Dann (44-20) 7516-6367

LATIN AMERICA RESEARCH Director of LatAm Research F. Rowe Michels, CFA (212-272-2429)

LatAm Portfolio Strategy Thierry Wizman, Ph.D. (212-272-7527)

LatAm Economics/Currencies Tim Kearney, Ph.D. (212-272-5981)

Banking/Financial Services Jason Mollin (212-272-2005)

Beverages Carlos A. Laboy (212-272-3203)

Cable & Media Christopher Recouso (212-272-6541)

Oil & Conglomerates Marc McCarthy, CFA (212-272-7822)

Paper and Forest Products Carl Weaver (55-11) 3457-3204

Retail Daniel Parker (212-272-9441)

Steel, Mining, & Cement Daniel C. Altman, CFA (212-272-5929)

Telecommunications Rizwan Ali (212-272-3513)

Utilities F. Rowe Michels, CFA (212-272-2429)

…and a ray of light: genasense

Documents