When Patent Appeals Survive but Patents Don’t: Taking a Look at the DHC Sulzer Decision

Two appeals. Two defeats. Yet, as Maneesha Gupta explains, Sulzer Mixpac AG v. Assistant Controller of Patents and Designs may be remembered less for the patent that failed than for the appeal that survived, adding another chapter to the debate over intra-court patent appeals. Maneesha is a fifth-year B.A. LL.B. (Hons.) student at NMIMS School of Law, Bengaluru, and a participant in the SpicyIP Summer School 2026.

When Patent Appeals Survive but Patents Don’t: Taking a Look at the DHC Sulzer Decision

By Maneesha Gupta

This is the story of an appellant’s pursuit of happiness (patenting their invention), troubled by prior art of its own, and an order refusing grant of patent application, which survived multiple rounds of appeals. In the High Court of Delhi, on one side stood the appellant Sulzer Mixpac AG, a Swiss global giant developer and manufacturer, regarded for its industrial dispensing systems. And on the other stood the respondent, Assistant Controller of Patents and Designs (ACPD), Mr. Kundan Kumar, an official known for writing robust, reasoned orders.

Sulzer, seeking a patent for its invention, ‘Static Mixer’, filed a patent application. The ACPD rejected it because it lacked an inventive step in light of prior art. Bitten by the order, Sulzer knocked at the door of a learned Single Judge of the Delhi High Court under Section 117A(2) of the Patents Act, 1970, only to find out that the Single Judge agreed with the Controller’s reasoning. Not giving up, it appealed again, filing a Letters Patent Appeal (LPA) before a Division Bench of the same court. However, even in the second appeal, Sulzer walked away with sweet nothing. 

Relying on Promoshirt SM SA v. Armasuisse (2023), the Division Bench rejected ACPD’s objection to the maintainability of Sulzer’s LPA, and in the same breath, dismissed the LPA by finding no error in the Single Judge’s judgement upholding ACPD’s order. In this blog, I examine the Court’s reasoning for allowing the maintainability of the LPA, review the grounds on which Sulzer’s patent application failed, and discuss what patent applicants appealing to High Courts under the Indian patent law can take away from this decision. With that, let’s first see what the claimed invention was and look at the specific grounds on which the Controller rejected it. 

 Subject Invention in Focus

 The subject invention was a plastic static mixer with multiple installation bodies connected via bar elements, withstanding internal tool pressures exceeding 1,000 bars without valve-breakage. Such a mixer would homogenize molten polymers in injection molding machines. 

In its application, Sulzer focused on comparing its new invention to EP’1426099B1 that utilized up to five installation bodies connected by a single bar, whereas the new one connected more than five installation bodies via a common bar element. Sulzer claimed this feature to be a novel technical advancement, enhancing mixer homogeneity and structural stiffness while enabling longer mixer lengths. 

However, in the Controller’s analysis, Sulzer’s Static Mixer also contested against four distinct prior art documents, D1 to D4, of which D1 was Sulzer’s own earlier European Patent [EP2181827A2]. The Controller rejected Sulzer’s application on the ground that the claimed invention lacked novelty, finding that D1 already disclosed the contested feature. Pointing to paragraph 75 of D1, he noted that a plastic static mixer made of 12 installation bodies connected by bar elements was already known in the art. The controller pointed out lack of inventive step as well. Extending a connecting bar across more than five installation bodies was an obvious modification for a Person Skilled in the Art (PSITA) when combining teachings from the four prior art documents, with no unexpected technical effect demonstrated.

APCD’s Home Run On Merits

The issue before the Court was whether Sulzer had invented enough beyond its own previous disclosure. The subject invention failed because the Court found it to be a mere modification based on the disclosures and teachings available in prior art documents, and thus sided with the Controller’s order. So why did the Court regard this as ‘a mere modification’?

A useful way to understand this is through Biswanath Prasad Radhey Shyam v. Hindustan Metal Industries (1978), wherein the Apex Court held that “to be patentable, an improvement on something known before, or a combination of different matters already known, should be something more than a mere workshop’s improvement, and must independently satisfy the test of invention or an inventive step.” This principle also prevents patent monopolies on any and every predictable advancement over a base invention.

What is mere modification and what is not? How would courts reason? To navigate this, Patent Law has provided us with the construct of a hypothetical PSITA. Building on Biswanath Prasad Radhey Shyam, the Delhi High Court in Roche vs Cipla (2012) explained that PSITA is someone possessing standard knowledge and skills in the relevant field. Obviousness is proved if a PSITA, having the prior art, could work upon the same in the workshop and achieve the desired result, leading to a patent which is under challenge. Further, while assessing inventive step, the Delhi High Court ruled in Agriboard International LLC vs Deputy Controller of Patents and Designs (2022) that a Controller must step into the shoes of the PSA to assess the invention. 

Were the Roche Five Steps Mandatory?

In assessing whether an invention is obvious, courts often times turn to the landmark framework laid down in Roche vs Cipla (2012). The Roche five-step test guides the evaluation of inventive step by requiring the court or Controller to:

1. Identify the inventive concept of the claimed patent.
2. Define the hypothetical PSITA and their general technical knowledge.
3. Identify the state of existing prior art.
4. Pinpoint the structural or functional differences between the prior art and the invention.
5. Assess whether reaching those differences from the prior art would have been obvious to the PSITA without hindsight.

Sulzer contended that the Controller did not apply the five steps outlined in Roche vs Cipla (2012) while examining the obviousness of the patent under consideration from the prior art. The High Court held that “such steps merely provide guidance as to how, in a particular case, the court has to proceed while examining the aspect of inventiveness or obviousness of an invention, which is sought to be patented, vis-à-vis prior art.”  Simply put, the court held that mere absence of rigorous application of the Roche vs Cipla’s 5 steps would not, by itself, invalidate the Controller’s order. They cannot be regarded as ‘commandments cast in stone’.

Intriguingly, the Delhi High Court’s Division Bench, inclusive of Justice C. Hari Shankar, who is one of the Judges for the present case as well, in Tapas Chatterjee vs Assistant Controller of Patents & Designs & Anr (2025)subtly departed from the current decision. When the applicant similarly challenged a patent application rejection for lack of inventive step under Section 2(1)(ja), the Division Bench there took on a stricter line, setting aside the Single Bench’s order for incorrect application of the 5 steps. Not treating the Roche steps as mere guidelines, the Bench insisted that explicit and structured application of those five steps would tick off the proper evaluation of obviousness. And because neither the Assistant Controller nor the Single Judge had systematically worked through the Roche framework, the Division Bench set aside the rejection order. It remanded the matter back for fresh consideration. 

The contrast seen between the two reasonings reflects an uncertainty- whether Controllers are required to adhere to a rigorous five-step analysis or whether just a bird’s-eye assessment of obviousness will survive an appeal.

Appeal after Appeal

The maintainability of the LPA hinged on whether the ACPD can be deemed to be a Court. Section 100A of the CPC strictly prohibits a second appeal where a Single Judge of a High Court has determined an appeal arising from an order or decree of a civil court. The Division Bench, drawing on its decision in Promoshirt, held that since the ACPD performs quasi-judicial functions, it cannot be deemed to be a civil court within the meaning of the CPC. The Court also observed that even if ACPD were to be seen through the lens of the doctrine of ‘trappings of the court’ that assess authorities having court-like characteristics, ACPD still won’t be seen as a civil court equivalent. The Single Judge’s decision was not then on an appeal from the original order of a civil court, rendering Section 100A inapplicable and affirming that the LPA was maintainable.

In Promoshirt, the Court also pointed out that Section 100A of CPC only steps in to block an LPA if the governing special statute expressly bars it or adopts the CPC’s restrictive appeal provisions. While reading the Patents Act, nowhere in Section 117A(2) of the Patents Act or the Act as a whole is an LPA expressly barred. Since the abolition of IPAB, all the appeals under Section 117A(2) were transferred to the High Courts, and the provision has ever since remained silent on the maintainability of an intra-court appeal. Since the Patents Act also does not adopt the restrictive rules of CPC,  Section 100A does not get attached to appeals under its Section 117(A) of the Patents Act.

While the Delhi High Court ultimately stood firm on the merits, it deciding to entertain the LPA under Section 117A(2) makes me think of the growing procedural split among the High Courts.

In my recent blog post, I noted: High Courts hold equal footing to one another, and so their judgements are not binding precedents, but hold persuasive value. However, on similar issues, other High Courts such as the Madras High Court and Calcutta High Court have taken the opposing view, holding that LPAs are not maintainable.

The fork between the Delhi High Court permitting LPAs and the Madras and Calcutta High Courts barring them can be a headache for the patent applicants. Following the abolition of the IPAB, Parliament left a key question unanswered: Would intra-court appeals remain available? So, until either the Supreme Court steps in and resolves this, or the Patents Act is fixed- the one holding the patent would base the appealability against a rejected patent on where their appeal is filed.

This was all about how Sulzer’s long pursuit of happiness came to a bittersweet end. This saga kept the Delhi High Court’s door open for future intra-court appeals, even as it reconfirmed ACPD’s reputation for drafting meticulous rejections to patent applications.