Case Analysis Intra‑Cellular Therapies, Inc. v. Controller of Patents (Delhi High Court, 2026)

 I. Overview and Procedural Posture

The Delhi High Court, in an appeal under Section 117A of the Patents Act, 1970, upheld the Controller's rejection (order dated 27.04.2023) of the patent application (No. 201817033732) titled "Organic Compounds" and covering deuterated heterocycle-fused gamma-carboline compounds (Formulas I–IV). The claims fell into two clusters: Claims 1–3 & 5–10 (Formula I–III compounds) and Claims 4 & 5–7 (Formula IV compounds), with Claim 11 as a dependent pharmaceutical composition claim. The claimed invention was for the treatment of diseases involving 5-HT2A receptor, serotonin transporter (SERT) and/or pathways involving dopamine D1/D2 receptor signalling systems and also has application in the treatment of diseases/disorders like anxiety, psychosis, schizophrenia, sleep disorders, sexual disorders, migraine, etc. The rejection rested on lack of novelty (Section 2(1)(j)), obviousness (Section 2(1)(ja)), and non-patentability under Section 3(d). The Court sustained the novelty and Section 3(d) findings and, having done so, found it unnecessary to separately adjudicate inventive step.

II. THE IMPUGNED ORDER — CONTROLLER'S REASONING

A. On Novelty [Section 2(1)(j)]

1. The Controller treated D1 and D7 jointly as constituting the closest prior art for different claim clusters — D1 for Formula IV claims (4, 5–7), and D7 for Formula I–III claims (1–3, 5–10).
2. As to D1: Example 1.21 of D1 was read to disclose R1 and R6 as –H, X as –N(R5) where R5 is –CD3, and R3/R4 as –D — a combination stated to result directly in the Formula IV compounds of the subject application.
3. As to D7: Claim 1 read with paragraph [00012] disclosed a generic Formula I where R1 is CH3 or CD3; R2, R3, R4, R5 each independently H or D (provided not all H when R1 is CH3). Claim 7 of D7 (dependent) additionally claimed all of R2, R3, R4, R5 as D. The Controller held that selecting R1 as CD3/CH3 and R2/R3 as independently H or D — with R4, R5 as H — results in Formula I–III compounds identical to claims 1–3 and 5–10.
4. Conclusion: Claims 1–3 & 5–10 lack novelty over D7; Claims 4 & 5–7 lack novelty over D1; Claim 11 (composition claim), being parasitic on claims 1–10, falls with them.

B. On Inventive Step [Section 2(1)(ja)]

5. The Controller's operative reasoning (extracted verbatim in the judgment) was:

"D4-D6 discloses the 'principle of using deuterated alternatives of known drugs'; and from the disclosure of any one among D4-D6, it is obvious for a person skilled in the art to bring a 'deuterated alternative of known drugs' (as disclosed in D1 and/or D7)."

6. This reasoning was essentially syllogistic and generic: (i) D1/D7 disclose the base (non-deuterated or partially deuterated) compound; (ii) D4–D6 disclose the general principle that deuteration of known drugs is a known technique; therefore (iii) combining the "principle" with the base compound renders the claimed deuterated compound obvious.
7. The order did not independently analyse each cluster of claims against a specific combination of documents, nor did it articulate what would motivate a person skilled in the art to select the specific deuteration positions claimed, over other theoretically available positions.

C. On Non-Patentability [Section 3(d)]

8. First hearing notice (04.02.2022): Claims 1–10 held to be a "mere discovery of a new form of a known substance."
9. Second hearing notice (21.02.2023) and final order: Claims 1–11 held "not allowable under section 3(d)," on the basis that "compounds I-IV as claimed in claims 1-10 ... are same as disclosed in D1 & D7."
10. Notably, the Controller's 3(d) conclusion was textually derived from the novelty finding (i.e., "are same as disclosed in D1 & D7") rather than from an independent efficacy-based analysis identifying: (i) the known substance, (ii) how the claim is a derivative of it, and (iii) a comparative therapeutic efficacy assessment.
11. No express reference was made anywhere in the order to the co-inventor's (Dr. Peng Li's) affidavit, filed on three separate occasions (with the FER response, and with both sets of written submissions).

III. APPELLANT'S SUBMISSIONS (Intra-Cellular Therapies, Inc.)

A. Procedural / Sequencing Objections

1. The novelty objection was raised only at the second hearing notice (21.02.2023); the first hearing notice (04.02.2022) raised only inventive step and Section 3(d) objections — suggesting the novelty ground was an afterthought rather than a considered, consistent position.

B. On Novelty — the "Multiple Selection" Argument

2. The invention is a species patent arising from Markush claims — six specific compounds within three formulas (Claims 1, 2, 3).
3. The Controller erred by treating multiple prior art documents (D1 to D7) as the "closest prior art" collectively; settled practice requires a single document to be assessed as the closest prior art for a novelty objection.
4. The obviousness objections in the first hearing notice were a verbatim reproduction of the European Search Opinion, which — considering the same documents D1–D7 — concluded that:

"claims 1-16 appears to be novel within the meaning of Article 54(1) and (2) of the EPC as no prior art document discloses the specified deuterated compounds presently claimed."

5. It was "intriguing" that the Indian Patent Office, examining the same D1–D7 set, reached the opposite conclusion without independent analysis or findings.
6. On D1 specifically: To arrive at the compound alleged by the Controller, a skilled person would need to make sequential, independent selections — first the base Formula I compound, then R6 as H (Formula 1.1), then Formula 1.21 — with the Controller's own first hearing notice having acknowledged that "multiple selections would be necessary to arrive at the presently claimed compounds."
7. On D7 specifically: The Controller referred only to the broadest embodiment of Formula I; multiple selections (R1, R2, R3, R4, R5) were needed to arrive at each of Claims 1–3, and the Controller had not shown why a skilled person would be motivated to make those specific selections (e.g., selecting R4 = H, R5 = H from Formula 1.4).
8. The inventor's subsequent research showed that compounds deuterated at the R4/R5 positions (as in D7) did not provide the expected metabolic benefits, whereas the compounds embraced by the amended claims did — indicating that the specific selection claimed was not an obvious or expected one.
9. Reliance was placed on Clause 8 of the Manual of Patent Office Practice and Procedure: a generic disclosure in prior art does not necessarily destroy the novelty of a specific disclosure.

C. On Inventive Step

10. Relying on paragraph 09.03.03.02, Chapter 9 of the Manual, the appellant argued that inventive step must assess the invention as a whole, and a claim cannot be held obvious merely because its individual parts, taken separately, are known.
11. The impugned order's reasoning was cryptic and unreasoned — it did not explain how or why a skilled person would be motivated to combine D1/D7 with D4–D6 to arrive at the claimed compounds; general reliance on "principles" disclosed in D4–D6 without "connecting the dots" rendered the order legally unsustainable.
12. Reliance placed on F. Hoffmann-La Roche Ltd. v. Cipla Ltd. (2015:DHC:9674-DB) and Agriboard International LLC v. Deputy Controller of Patents (2022:DHC:1206) — both requiring the Controller to analyse the existing knowledge and articulate how a skilled person would move from it to the claimed invention.

D. On the Co-Inventor's Affidavit

13. The affidavit of co-inventor Dr. Peng Li was filed on three separate occasions — with the FER response, with the first post-hearing written submissions, and with the second post-hearing written submissions — yet the impugned order contained no reference to it whatsoever.
14. This omission was said to be squarely violative of Milliken and Company v. Controller of Patents (2025:DHC:1782) and The Regents of the University of California v. Union of India (2019:DHC:2699), both holding that a Controller's failure to consider material evidence on record vitiates the order.

E. On Section 3(d)

15. For a Section 3(d) objection to be sustained, the Controller must identify the "known substance" relied upon — this was "conspicuous by its absence" in both hearing notices (the first merely alleging claims 1–10 were a "new form of a known substance"; the second merely repeating the statutory bar without identifying the substance).
16. Reliance on D.S. Biopharma Limited v. Controller of Patents (2022:DHC:3563) and Taiho Pharmaceutical Co. Ltd. v. Controller of Patents (2025:DHC:3777), the latter's para 13 “that in order to sustain an objection under Section 3(d) of the Act, the following factors have to be clearly identified by the Controller:

i) the ‘known substance’ with ‘known efficacy’;

ii) clear explanation as to how and why the claimed substance is a derivative or otherwise a new form of a ‘known substance’;

iii) an objective comparison between the therapeutic efficacy of the claimed invention and that of the known substance.”.

 

17. The Controller conflated the criteria for novelty/inventive step with those for Section 3(d) — the impugned order's finding that the compounds "are same as disclosed in D1 & D7" (a novelty-type finding) was used interchangeably as the basis for the 3(d) rejection, despite these being legally distinct enquiries.
18. Experimental data (Examples 5, 6 & 7 of the complete specification) was not appreciated by the Controller:
• Example 5 (mice, in vivo): Relative amide formation was significantly lower for Examples 1, 2, 3 (0.54, 0.38, 0.31) versus non-deuterated compound Q (0.79).
• Example 6 (rats): Comparative AUC data for parent compound and Metabolite Q-1 showed differing metabolic profiles between Formula Q and Example 2 (Formula I) across oral (PO) and intravenous (IV) routes.
• Example 7 (dogs): Deuterated Example 2 showed 72% higher parent-drug AUC than Formula Q on sublingual administration, and altered Q-1A metabolite ratios on subcutaneous administration — demonstrating that deuteration inhibits subsequent oxidation of the demethylated amine to its amide derivative.
19. These experimental comparisons, taken as a whole, were not even considered by the Controller in arriving at the impugned order.

IV. RESPONDENT'S ARGUMENTS (Controller of Patents)

A. Claim Structuring

1. The 11 claims divide into Part 1 (Claims 1–3, 5–10 — Formula I compounds) and Part 2 (Claims 4, 5–7 — Formula IV compounds); Claim 11 (composition) is entirely parasitic on the allowability of Claims 1–10.

B. On Novelty

2. D1 (Example 1.21) discloses R1 & R6 as –H, X as –N(R5) with R5 = –CD3, R3 & R4 = –D — which directly results in the Formula IV compounds of Claims 4 & 5–7; hence those claims lack novelty.
3. D7 (Claim 1) discloses a generic Formula I from which selecting R1 as CH3/CD3 with R2, R3 independently H or D, and R4/R5 as H provided that R2, R3, R4 and R5 are not all H when R1 is CH3, results in Formula I–III compounds identical to Claims 1–3 & 5–10.
4. The claimed invention is "nothing other than the generic compound of Formula I-III" already covered by D1 and D7 — the novelty objection "cannot be doubted or distinguished."
5. Since Claims 1–10 lack novelty, the composition Claim 11 necessarily also lacks novelty.

C. On Inventive Step

6. D4–D6 disclose the general principle of using deuterated alternatives of known drugs; since this principle applies to compounds disclosed in D1/D7 (which are in the "same field"), it would be obvious for a skilled person to arrive at a deuterated alternative — sustaining the Controller's combination-based obviousness conclusion.
7. Since Claims 1–10 already lack novelty, the inventive step requirement (which logically presupposes some element of departure from the prior art) is a fortiori not met.
8. "Serial patenting" / evergreening argument: Since the same appellant is the inventor/applicant behind the prior art documents (D1, D7) and the present claimed invention, the test of anticipation cannot be judged from the perspective of a "person ordinarily skilled in art" but must be judged from the perspective of the "person in the know" — relying on AstraZeneca AB v. Intas Pharmaceuticals Ltd. (2021 SCC OnLine Del 3746).

D. On Section 3(d)

9. The compounds claimed in Claims 1–10 (Formula I–IV) are the same as disclosed in D1 & D7, and are therefore a mere discovery of a new form of a known substance — hence non-patentable under Section 3(d).
10. Even assuming novelty could somehow be established, the claimed compounds would still fail Section 3(d) because the appellant has not established a technical advancement in "therapeutic effect" — relying on Novartis AG v. Union of India & Ors. (2013) 6 SCC 1.
11. The in vivo comparative data (mice and rat studies) relied upon by the appellant, even if considered, only shows better pharmacokinetic performance relative to the original Formula Q — it does not disclose a significant increase in therapeutic efficacy, and is therefore insufficient to overcome the Section 3(d) bar.
12. The appellant's reliance on international patent grants (EPO, etc.) is misplaced: patent rights are territorial, governed solely by the Patents Act, 1970, in India; there is no statutory obligation on the Controller to follow or rely upon foreign grants — relying on Communication Components Antenna Inc. v. Ace Technologies Corp. (2019 SCC OnLine Del 9123).

V. COURT'S FINDINGS

A. On Novelty [Section 2(1)(j)]

1. The Court accepted the Controller's mapping: D1 (Example 1.21) results in the Formula IV compounds of Claims 4 & 5–7; D7 (Claim 1 read with dependent Claim 7 and para [00012]) results in the Formula I–III compounds of Claims 1–3 & 5–10.
2. On the appellant's "multiple selections" argument, the Court held it is a "settled position that where a compound is disclosed under a genus patent (i.e. Prior art D1 and D7 here), specific disclosure is immaterial — rejecting the contention that sequential selection across independent variables defeats novelty.
3. This was grounded in AstraZeneca AB (DB), which emphasised that if a product is specifically "covered" in the claims of a Patent in question (i.e. prior patent), whether specific disclosure of that product (compound in this case) concerning the same has been made or not is immaterial, and in Boehringer Ingelheim Pharma GMBH v. Vee Excel Drugs (para 90), holding that "claimed," "covered," "encompassed," and "disclosed" are treated as functionally equivalent for this purpose.
4. The Court also invoked Novartis AG (paras 139, 156) on the danger of a "vast gap" between coverage and disclosure, cautioning against claim drafting that allows patent scope to exceed genuine technical disclosure.
5. Conclusion: The appellant's submissions on multiple selections in D1 "cannot be accepted"; the novelty objection under Section 2(1)(j) was upheld in respect of all claims.

B. On Non-Patentability [Section 3(d)]

6. The Court reiterated the Novartis interpretation of "efficacy" as meaning strictly therapeutic efficacy, not merely a beneficial physicochemical or pharmacokinetic property, and reproduced Novartis para 180 in full.
7. Compound Q — the known, non-deuterated congener — was found to be disclosed in D1 and D7 (paras [0096] and [0094] respectively), satisfying the "known substance" identification the appellant argued was missing (the Court effectively supplied this finding on its own analysis of the record).
8. On Example 7 (dog PK data): The Court accepted the appellant's own characterization of the results (72% higher parent-drug AUC on sublingual dosing; altered Q-1A metabolite ratios), but held that this shows only enhanced bioavailability/reduced metabolism, and that "enhanced bioavailability does not, by itself, lead to enhanced therapeutic efficacy" — the appellant needed to show, with research data, that improved bioavailability translates into a therapeutic benefit.
9. Relied on Natco Pharma v. Novartis AG [FAO(OS)(COMM) 178/2021, decided 24.04.2024] (Division Bench), particularly paras 74, 84, 86, & 87, holding that bioavailability is "one of the pharmacokinetic parameters and not a direct measure of therapeutic efficacy," and that enhanced bioavailability may sometimes even be undesirable (e.g., where side effects are dose-related).
10. On the co-inventor's affidavit (dated 12.02.2020): The Court did examine its contents (paras 17–20 of the affidavit), notwithstanding the Controller's silence on it, and found:
• Data under paras 17–18 (relative amide formation across compounds I, II, III; 40–50% reduction in circulating Metabolite X) replicated the same outcome as Example 7 and was, for the same reasons, insufficient to establish therapeutic efficacy.
• Data under para 19 (receptor-binding IC50 study) showed the deuterated compound and Formula Q have "substantially similar pharmacological activity" — this indicated the drug behaves the same way pharmacologically, which is not equivalent to proving it treats the disease better.
11. Reiterated Novartis paras 187 & 189: whether increased bioavailability enhances therapeutic efficacy "must be specifically claimed and established by research data" — and found no such specific claim or data on record.
12. Conclusion: The data submitted (whether in the specification or the affidavit) did not overcome the requirement of Section 3(d); the objection was upheld.

C. On the Co-Inventor's Affidavit / Natural Justice Objection

13. The Court noted the Controller's order was silent on the affidavit, but resolved the issue not by remand, but by independently considering the affidavit's contents on appeal and finding them technically unavailing (duplicative of Example 7 data already assessed).
14. Having examined the affidavit "on merits," the Court held that the precedents relied upon by the appellant on this point (Milliken, Regents of University of California) "need not be looked into" — effectively treating the natural-justice defect as cured by the appellate court's own substantive assessment rather than requiring the Controller to have engaged with it in the first instance.

D. On Inventive Step [Section 2(1)(ja)]

15. Having upheld both the novelty objection (Section 2(1)(j)) and the non-patentability objection (Section 3(d)), the Court held it "does not feel the requirement to address the objection on the ground of lack of inventive step."
16. The appellant's detailed submissions on the Controller's allegedly cryptic and unreasoned obviousness analysis (paras 9–12 of the submissions, invoking Hoffmann-La Roche and Agriboard) were, as a result, left unaddressed on the merits.

E. On the "Foreign Grant" / EPO Divergence Argument

17. The Court did not substantively engage with the appellant's contention that the EPO, examining the same D1–D7 prior art, had found the corresponding European claims novel. The point was implicitly subsumed within the territoriality principle (patents being governed purely by the Act in India), without a comparative analysis of why the EPO's selection-invention/individualization approach would or would not apply under Indian law.

F. Disposition

18. Result: The impugned order dated 27.04.2023 was upheld in its entirety on the grounds of lack of novelty (S. 2(1)(j)) and non-patentability (S. 3(d)); the appeal was dismissed, with no order as to costs.

VI. SUMMARY TABLE — ISSUE-WISE POSITIONS

Issue	Impugned Order	Appellant	Respondent	Court's Finding
Novelty (S. 2(1)(j))	D1 anticipates Claims 4, 5–7 (Formula IV); D7 anticipates Claims 1–3, 5–10 (Formula I–III)	Multiple independent selections needed from D1/D7; single-document rule violated; EPO found same art novel	Selection irrelevant once compound is "covered" by genus claims of D1/D7	Upheld — coverage = disclosure; specific disclosure immaterial for genus-covered compounds
Inventive Step (S. 2(1)(ja))	D4–D6 "principle" + D1/D7 base compound = obvious	Order cryptic; no motivation-to-combine analysis; invention must be viewed as a whole	Follows automatically once novelty is lost; "person in the know" standard given common inventorship	Not decided — rendered moot by novelty/3(d) findings
Section 3(d)	Claims 1–11 "same as disclosed" in D1/D7; non-patentable	Known substance never identified; novelty/3(d) criteria conflated; PK data (Examples 5–7) shows efficacy	PK data shows only better performance, not therapeutic efficacy; foreign grants irrelevant (territoriality)	Upheld — bioavailability/PK improvement ≠ therapeutic efficacy; no research data specifically establishing efficacy enhancement
Co-inventor's affidavit	No reference at all in the order	Complete non-consideration violates natural justice (Milliken, Regents of UC)	—	Affidavit examined by the Court itself; found duplicative/insufficient; natural justice objection treated as cured, not requiring remand
Foreign (EPO) grant	Not addressed	EPO found same D1–D7 art establishes novelty	Territoriality — foreign grants immaterial	Territoriality principle applied; no independent comparative reasoning on EPO's contrary finding

VII. CONCLUDING NOTE

The judgment ultimately turns on two doctrinal anchors — the "coverage-equals-disclosure" principle for novelty, and the strict "therapeutic efficacy" standard under Section 3(d) as laid down in Novartis and refined in Natco Pharma. Both grounds were sufficient, in the Court's view, to dispose of the appeal, resulting in the inventive-step objection — arguably the most rigorously contested issue on the appellant's side — being left unaddressed. The treatment of the co-inventor's affidavit, examined substantively on appeal rather than remanded for want of Controller consideration, reflects a pragmatic but doctrinally noteworthy departure from strict natural-justice remedy in patent prosecution appeals.