1990 — Andrew S. Bell / David Brown / Nicholas K. Terrett at Pfizer Sandwich UK Filed the 'Antianginal Agent' Pyrazolopyrimidinone Patent US5250534A. Reading the Prehistory of Sildenafil (later Viagra) from Primary Sources. Correcting the DB 'Peter Dunn / Albert Wood' Description

About excavation memos: "Excavation memos" in this series record candidate summaries at the stage where the primary-source URL has been confirmed. Full-text reading and verbatim verification of all claims have not been performed. Only confirmed facts are recorded; speculation is explicitly marked.

Why dig

Sildenafil (Viagra), approved by the FDA on March 27, 1998, became the starting point for erectile dysfunction (ED) treatment, 21st-century pharmaceutical marketing, men's health medicine, and the PDE5 inhibitor group (later tadalafil / vardenafil / avanafil). First-year sales exceeded $1 billion, and cumulative sales over 30 years exceed $20 billion. Since the 1998 approval, it has become the symbolic pharmaceutical of Pfizer's reputation.

The textbook narrative widely circulates the story "Pfizer developed sildenafil as an angina treatment, but in clinical trials, ED effects were observed by chance, and it was redeveloped as an ED treatment." But the patent-history details (the core substance patent's number, inventors, priority date, and original target indication) are rarely accurately referenced in general technology discourse.

The DB row PH-006 in ~/ai-archaeology/db/candidates.tsv lists "Viagra (sildenafil) patent," "Pfizer Inc., Peter Dunn / Albert Wood," and "1992." Comparing this with the primary source on Google Patents, three corrections are required:

1. Inventors: The DB description "Peter Dunn / Albert Wood" is wrong. The inventor field of US5250534A is Andrew S. Bell / David Brown / Nicholas K. Terrett three co-inventors. Peter Dunn was a medicinal chemist at Pfizer Sandwich and may be listed as inventor on separate patents (use patents or synthesis process improvement patents), but is absent from this patent. Same for Albert Wood.
2. Assignee: The DB description "Pfizer Inc." is too coarse. Original Assignee is Pfizer Corp SRL (Pfizer's UK subsidiary). The Pfizer Sandwich research site (Kent, UK) was the research base, and the patent was filed via the UK subsidiary.
3. Priority date: The DB description "1992" only reflects the year of U.S. filing date 1992-05-14, but the priority date is 1990-06-20 (likely via UK priority application GB9013750). The U.S. filing is a 2-year-later continuation filing.

In the consecutive DB-error correction series of Days 8/9/10/11, this is the 10th case of DB-error discovery and correction (feedback_db_meta_verify_primary). The description "Pfizer's Peter Dunn and Albert Wood invented" is a frequently seen arc-quotation in popular science books and Wikipedia; checking the patent primary source reveals three different people.

Patent basic information

• Patent number: US5250534A
• Title: Pyrazolopyrimidinone antianginal agents
• U.S. filing date: May 14, 1992
• Priority date: June 20, 1990 (presumed UK priority GB9013750)
• U.S. grant date: October 5, 1993
• Inventors: Andrew S. Bell / David Brown / Nicholas K. Terrett (three co-inventors, DB description is wrong)
• Original Assignee: Pfizer Corp SRL (Pfizer's UK subsidiary, Pfizer Sandwich research site, Kent, UK)
• Target indications: Specification explicitly lists "stable, unstable and variant (Prinzmetal) angina," "hypertension," "heart failure"—angina, hypertension, heart failure. ED treatment is not the indication of this patent (indication switch via later clinical trial side-effect observations)
• U.S. expiration: As substance patent, expired in 2010s (20 years from grant; may have been extended via Hatch-Waxman, etc.)
• Primary source: Google Patents (URL confirmed; title, three inventors, priority date, filing date, grant date, Original Assignee all obtained)

Core (Google Patents obtained information)

Claim 1 of US5250534A (summary) covers a class of compounds centered on the pyrazolo[4,3-d]pyrimidinone scaffold by claiming combinations of substituents R1 through R6 broadly. The specification includes as a specific example sildenafil (UK-92,480), recorded with InChIKey notation such as "BNRNXUUZRGQAQC-UHFFFAOYSA-N sildenafil." The patent explicitly states the action mechanism as cGMP-selective phosphodiesterase (cGMP-PDE) inhibitor and asserts selectivity over cAMP-PDE as "selective inhibitors of cGMP PDEs," not specifying PDE5 specificity.

Important points in the specification:

• Target indications: The specification opens with "treatment of various cardiovascular disorders such as angina, hypertension, heart failure," explicitly listing "stable, unstable and variant (Prinzmetal) angina." It claims a broad indication targeting all three angina pectoris subtypes.
• Action mechanism: The specification claims a mechanism of relaxing vascular smooth muscle and increasing coronary blood flow via cGMP-selective PDE inhibition to ameliorate angina. PDE5 isoform specificity was not established at the time, and the patent claims at the broader cGMP-PDE level.
• Absence of ED indication: References to erectile dysfunction (ED) / impotence / penile erection are absent in the specification. The indication switch to ED treatment came via small Phase I clinical trial side-effect observations in 1992-1993, and was filed in a separate-series use patent (later US5346901A, etc.).

The inventor problem: Peter Dunn and Nicholas Terrett

The DB description "Peter Dunn / Albert Wood" is wrong. The Pfizer Sandwich sildenafil development team had multiple contributors, and patent inventors do not necessarily match historical contributors, so there are multiple inventor narratives:

• Nicholas K. Terrett (medicinal chemist, Pfizer Sandwich): One of the three co-inventors of US5250534A. Widely recognized as the leader of sildenafil molecular design, often called "father of Viagra" in media. Later left Pfizer and went independent at Ensemble Therapeutics (Massachusetts), etc.
• Andrew S. Bell / David Brown (medicinal chemists, Pfizer Sandwich): Two of the three co-inventors of US5250534A. Implementation of chemical synthesis.
• Peter Dunn (medicinal chemist, Pfizer Sandwich): Absent from the inventor field of this patent, but likely contributed to mass-synthesis process improvements for sildenafil. May be listed as inventor in Pfizer's internal Viagra manufacturing process patents from 1998 onward, but unconfirmed in this article.
• Albert Wood: Absent from the inventor field of this patent. Likely affiliated with the Pfizer Sandwich analytical chemistry team, but unconfirmed in this article.

The DB description "Peter Dunn / Albert Wood" likely originates from popular science books and Pfizer press releases after the 1998 Viagra launch, where two scientists were arbitrarily selected from "multiple scientists who contributed to sildenafil development." Checking the patent primary source (the inventor field of US5250534A), Peter Dunn and Albert Wood are absent, and the three co-inventors are Bell / Brown / Terrett.

Indication switch to ED treatment

US5250534A was filed and examined in 1990-1992 as an antianginal agent. Pfizer Sandwich conducted Phase I clinical trials of UK-92,480 (later sildenafil) in angina patients in 1989-1991, but angina efficacy was limited, and Phase II/III progression was nearly abandoned.

In 1992, Pfizer's clinical trial director Ian Osterloh re-analyzed Phase I trial side-effect reports and observed penile erection side-effect reports in multiple male subjects. He reconsidered this as a possibility for ED treatment, and Phase II clinical trials in ED patients started in 1993. In 1996, Pfizer separately filed use patent US5346901A as "use of pyrazolopyrimidinones for impotence." On March 27, 1998, the FDA approved Viagra (sildenafil 25/50/100mg) as an ED treatment.

This path is widely told in pharmaceutical history as "indication switch from angina to ED via serendipity," but actually requires a 4-step process: (a) side-effect observation → hypothesis formulation → Phase II/III clinical trials → FDA approval—a continuous observation-based rational judgment, not pure chance. Osterloh himself later reflected, "rather than serendipity, observation-driven hypothesis testing" (Pfizer official retrospective, source unconfirmed in this article).

Connection to modern (includes speculation)

Notes on reading the table:

Row 1 is identical. Since sildenafil was approved as an ED treatment in 1998, it became the starting point of the PDE5 inhibitor group: Cialis (tadalafil, Eli Lilly 2003 approved), Levitra (vardenafil, Bayer/GSK 2003 approved), Stendra (avanafil, Vivus 2012 approved). The same compound has been used for 25 years, exceeding the substance patent term (U.S. 20 years) while the clinical value of the compound continues.

Row 2 is identical. The indication switch from "angina to ED" is an early example of repurposing drug discovery. Modern examples include (a) thalidomide → multiple myeloma, (b) ranolazine → chronic fatigue syndrome (in clinical trials), (c) metformin → anti-aging (TAME trial), etc. This patent is a rare case that went through an indication switch process from 1992 to 1998.

Row 3 is identical. The rational design of cGMP-selective PDE inhibitors is an early example in the lineage of PDE isoform-specific inhibitor design that continues with later PDE3 inhibitors (cilostazol), PDE4 inhibitors (roflumilast, apremilast), and PDE9 inhibitors (in clinical trials).

Row 4 is similar. The 1998 Viagra approval is a symbolic case of the medicalization of male health—"restoring male sexual function with pharmaceuticals." It overlaps with modern testosterone replacement therapy and andropause medicine in problem-consciousness, but the targets differ (erection vs. male hormones), so we keep it at "similar."

Row 5 is similar. Pfizer Sandwich research site (Kent, UK) was established in 1953 and was an important base producing sildenafil, azithromycin (Zithromax), doxazosin (Cardura), etc. In 2011, Pfizer scaled down R&D at Sandwich, and currently it functions as a base for Servier UK and small biotechs. We read it as an early example of modern multinational pharma geographic research division (Boston/Cambridge in the U.S., Shanghai/Beijing in China, Basel/Cambridge in Europe, Hyderabad in India).

Row 6 is identical. The dual enclosure of substance patent US5250534A (angina indication) and use patent US5346901A (ED indication) is an early example of the modern pharmaceutical patent strategy standard (multi-layer structure of substance + use + formulation + crystal form + manufacturing process).

Row 7 is similar. Phase I side-effect observation → indication switch resonates with modern AI-driven side-effect prediction (drug side-effect signal detection from FDA FAERS data, electronic medical records, social media mentions) and drug repurposing AI in problem-consciousness. In 1992, Pfizer's Osterloh manually re-analyzed Phase I trial side-effect reports, but today similar signal detection from large-scale data is possible with AI.

Why this is worth digging

Reason 1: 10th DB error correction

In the consecutive DB-error correction series of Days 8/9/10/11, this is the 10th case (IC-009/011/012, PH-001-005/007, FH-001/002). The arc-quotation "Pfizer's Peter Dunn and Albert Wood invented" circulates in Wikipedia and popular science books, but checking the patent primary source reveals different three people (Bell / Brown / Terrett). We record it as a phenomenon that arc-quotation errors in patent metadata frequently occur in pharmaceutical patents.

Reason 2: Precise narrative of "from angina to ED by chance"

The story "sildenafil was developed as an angina treatment but happened to work for ED" is widely circulated in popular science, but accurately tracing the 5-step process—(a) 1990-1992 filed and examined as antianginal agent, (b) 1992 Phase I side-effect observation, (c) 1993-1996 Phase II/III in ED patients, (d) 1996 use patent US5346901A separately filed, (e) 1998 FDA approval—shows 6 years of continuous rational judgment, not pure chance. The distinction between "serendipity" and "observation-driven hypothesis testing" is important.

Reason 3: Substance patent vs. use patent dual strategy

The dual enclosure of US5250534A (substance patent, 1990-2010s) and US5346901A (ED use patent, 1996-2010s) is an early example of the modern pharmaceutical patent strategy standard format. Through the patent term extension system (Hatch-Waxman, etc.), a structure where substantial extension is possible via use patents after substance patent expiration is shown by this patent group.

Pitfalls

Pitfall 1: "Peter Dunn and Albert Wood are inventors" is a popular science error

Wikipedia's English Sildenafil article and Pfizer's official history pages sometimes describe Peter Dunn and Albert Wood as "contributors to sildenafil development," but they are absent from the inventor field of US5250534A. Peter Dunn may be listed as inventor on separate-series synthesis process improvement patents, but the inventors of substance patent US5250534A are Bell / Brown / Terrett three. "Pfizer's overall contributors" and "patent inventors" are separate concepts.

Pitfall 2: "Failed as antianginal" is oversimplification

The angina efficacy in Phase I clinical trials was limited, but not a complete failure—it was a stage where the effect size fell below expected values. Phase II/III progression was abandoned due to (a) effect size insufficient compared to existing nitrates, calcium channel blockers, β-blockers, (b) side-effect profile (penile erection, headache, facial flushing) hard to tolerate for angina patients—a relative judgment, not absolute failure. "From failure to chance success" is an oversimplified narrative.

Pitfall 3: "PDE5-specific inhibition" was established post-1998

US5250534A was filed and examined in 1990-1993 as a "cGMP-selective PDE inhibitor," but at the time, PDE enzymes were at a stage where the 5 isoform classification PDE1-PDE5 was being established, and sildenafil's PDE5 specificity (weak cross-activity to PDE6 was later reported as a cause of cyanopsia side effect) was clarified after the 1998 approval. Claim 1 of this patent does not assert PDE5 specificity, so the current "PDE5 inhibitor" category is narrower than the patent's claim scope.

Pitfall 4: "Viagra = sildenafil" is confusion of trademark and substance name

Viagra is Pfizer's trademark; sildenafil is the compound name (INN, international nonproprietary name). Substance patent US5250534A covers the sildenafil compound itself, and the Viagra trademark was registered in the U.S. in 1998. After substance patent expiration (2010s), generic sildenafil circulated in the market, but Pfizer continues to hold the Viagra trademark. Equating "Viagra = sildenafil" confuses trademark and substance name.

Strictly speaking

Confirmed facts

From Google Patents: US5250534A / Title "Pyrazolopyrimidinone antianginal agents" / Inventors "Andrew S. Bell / David Brown / Nicholas K. Terrett" three co-inventors (mismatch with DB description "Peter Dunn / Albert Wood" = DB error) / Original Assignee "Pfizer Corp SRL" (Pfizer UK subsidiary, DB "Pfizer Inc." description is coarse) / Priority date 1990-06-20 (DB "1992" description only reflects year of U.S. filing date 1992-05-14) / Filing date 1992-05-14 / Grant date 1993-10-05 / Target indications "stable, unstable and variant (Prinzmetal) angina," "hypertension," "heart failure"—angina, hypertension, heart failure (explicitly stated in specification) / References to ED treatment absent in this patent specification / March 27, 1998 FDA Viagra (sildenafil) ED treatment approval (Pfizer press release confirmed, original not consulted).

Author's interpretation

"Peter Dunn may be listed as inventor on separate-series patents," "Albert Wood may be affiliated with Pfizer Sandwich analytical chemistry team," "Nicholas Terrett led molecular design," "Ian Osterloh re-analyzed Phase I side-effect reports," "Phase I angina efficacy was limited," "PDE5 specificity was established post-1998"—these are author's interpretations / secondary-source-derived information. Pfizer internal records, Sandwich research site notebooks, and Osterloh's retrospectives are unconfirmed within the scope of this article.

Metaphors / analogies

Row 4 of the table (1998 Viagra approval vs. modern men's health medicine) is similar. We arrange them as the medicalization of male health, but the targets (erection vs. male hormones) differ, so we keep it at "similar." Row 5 (Pfizer Sandwich vs. modern multinational pharma geographic research division) is similar. We read as continuity of geographic research division, but era differs. Row 7 (serendipity vs. AI-driven side-effect prediction) is similar. We arrange as observation-data-to-hypothesis-generation problem-consciousness, but technical means differ.

Unconfirmed

Verbatim verification of the full Description text of US5250534A / complete acquisition of Claim 1 verbatim (only summary obtained) / full text of Claims 2 and onward / original UK priority application GB9013750 / use patent US5346901A (ED indication) inventors, filing date, Claim 1 / Pfizer internal UK-92,480 development records / original Ian Osterloh 1992 Phase I side-effect report re-analysis / 1998 FDA Viagra approval document (Approval Package) / circumstances of Nicholas Terrett's Pfizer departure and Ensemble Therapeutics founding records / Peter Dunn and Albert Wood's Pfizer-era affiliated teams and patent groups / sildenafil synthesis process improvement patent group (Pfizer Manufacturing Process patents) / Ian Osterloh 1996 British Journal of Clinical Pharmacology retrospective (bibliographic info confirmed, original not consulted) / 1992 Phase I trial protocols, subject numbers, side-effect reports / comparison with Eli Lilly tadalafil (Cialis, US5859006, 1995 priority) and Bayer vardenafil (Levitra, US6362178, 1998 priority).

Where this comparison breaks down

US5250534A is a pyrazolopyrimidinone antianginal substance patent with priority date 1990, and the modern PDE5 inhibitor ED treatment market differs fundamentally in regulatory environment, target indications, and PDE isoform specificity understanding. Connecting "substance patent = modern ED treatment" directly invites multiple objections from experts: (1) the patent's target indication is in the specification angina, hypertension, heart failure with ED absent, (2) the indication switch to ED treatment came via use patent US5346901A in 1996, (3) the 1998 FDA approval is a separate review path as ED treatment, (4) PDE5 specificity was established post-1998, with this patent's Claim 1 at the broader cGMP-PDE level. The DB description "Peter Dunn / Albert Wood" is an arc-quotation circulating in Wikipedia and popular science books; checking the patent primary source (US5250534A inventor field) reveals three different people (Bell / Brown / Terrett). In the consecutive DB-error correction series of Days 8/9/10/11, this is the 10th case, reconfirming that the DB note column descriptions are arc-quotations from popular science books that mix in errors. The excavation memo is limited to Google Patents range confirmation; complete acquisition of Claim 1 verbatim, confirmation of use patent US5346901A, Pfizer internal records, original Osterloh retrospective, and FDA Approval Package are not obtained—these limits are stated explicitly.

References:

• Original patent: US5250534A on Google Patents
• Related excavation note (Pharmaceutical Patents #4): Syntex Djerassi/Miramontes/Rosenkranz norethindrone US2744122A (1951, Mexico)
• Related excavation note (Pharmaceutical Patents #3): Squibb Ondetti & Cushman ACE inhibitor (captopril) two-series patents
• Related excavation memo #3 (Pharmaceutical Patents): ICI propranolol US3337628A + derivative US3408387A (1962-1968)
• Related excavation memo #1 (Pharmaceutical Patents): Akira Endo statin US4049495A (1974, Sankyo)
• Related excavation memo #2 (Pharmaceutical Patents): Banting/Best/Collip insulin US1469994 (1923)