Clinical Trial Activation in India: Where the Friction Actually Sits
India approves clinical trials faster than most sponsors expect. The New Drugs and Clinical Trials Rules, 2019 (G.S.R. 227(E)) set a decision window of 30 working days for drugs discovered and developed in India and intended to be manufactured and marketed there, with deemed approval if the regulator does not respond, and 90 working days for other new drugs, including those developed abroad. Measured against the regulatory clocks of most major markets, that is a predictable instrument.
Sponsors still lose quarters activating in India.
The distance between those two statements is the subject of this piece. It is not an argument that the regulator is slow. It is an account of what sits downstream of the regulator, and of one structural problem that shapes everything else: in India, a trial can be approved and still be difficult to find.
The regulatory clock is not the bottleneck
Start by removing the assumption that usually gets budgeted for. The NDCT Rules of 2019 legislated the decision windows above, and legislated deemed approval as the default outcome of regulatory silence on the domestic pathway, with notification to CDSCO. That is an unusually sponsor-favourable construction. It is also narrow. The window covers the regulatory decision and nothing else.
Everything else runs on separate clocks with separate owners. Ethics committee review sits with each registered committee, not with the regulator. Site contracting sits with the institution. Import licensing for investigational product runs its own path. None of those are governed by the 30 or 90 working-day windows, and in most India programmes the aggregate of those separate clocks exceeds the regulatory one.
A sponsor that plans an India arm against the published regulatory timeline is planning against the smallest component of its own critical path.
Approval and registration are not the same event
This is the part that no timeline captures, and it is the one that matters most operationally.
CDSCO, through the DCGI and its Subject Expert Committees, decides whether a trial may proceed, and those recommendations are published. Trial existence, separately, is recorded in the Clinical Trials Registry - India, maintained by ICMR and recognised as a WHO primary registry, and for multi-country protocols on ClinicalTrials.gov. These are different systems, maintained by different bodies, for different purposes. They do not reconcile automatically and they frequently do not reconcile at all.
The pattern is visible in public sources, with an important qualification about what can and cannot be concluded from it. Trade press reported in October 2025 that a Subject Expert Committee had recommended approval of a Phase 1/3 biosimilar trial for a major Indian sponsor. As of 5 August 2026 we were unable to locate a corresponding registration for that programme in either CTRI or ClinicalTrials.gov. In the same month the committee returned another sponsor's Phase 3 protocol for revision, and in April 2026 recommended approval of that sponsor's second Phase 3 programme; neither resolved to a registry identifier we could locate. Those are not-located findings against our source set on that date. Given that CTRI's search interface is CAPTCHA-gated and mirrors are incomplete, absence from our searches is not proof that no record exists.
Within our own dataset the fragmentation is direct rather than inferred. Across active India-site oncology studies as of 5 August 2026, five Indian sponsors appeared in one registry only and two appeared in the other only, and at least one pivotal programme was listed as actively enrolling under a registry identifier that a second registry attributes to an unrelated cardiac study. That last case is the one worth holding onto: it is not a gap, it is a collision, and it survives any amount of careful searching.
Site-level visibility is thinner still. In that same snapshot, roughly 40 percent of Indian site listings on industry-sponsored studies were masked as "Research Site" rather than named, with 70 masked entries in New Delhi, 47 in Kolkata, and 40 in Nashik. Every named-site count in India is therefore a floor.
Read that as an operational statement rather than a data quality complaint. A sponsor planning an India arm is planning against a picture that no single registry produces, and no intelligence feed built on a single registry produces it either. Reconciliation is not a refinement applied after site selection. It is the first gate, upstream of everything.
Ethics committee status is a live variable
The second structural fact is that review capacity in India is currently being withdrawn rather than added, and the withdrawal is documented.
The legal basis is not in question. Under the NDCT Rules of 2019, ethics committees must be registered with CDSCO, and the DCGI is empowered to suspend or cancel that registration for non-compliance. India runs a risk-based inspection programme on that basis.
In mid-2026, according to press reports, CDSCO suspended for 24 months the ethics committee registration at a large Indian oncology hospital network following a risk-based inspection. The findings behind that action have been described in press reporting rather than in a published inspection record. What is not in dispute is the suspension itself and its effect, which is that new study approvals at that site stop until 2028 while ongoing studies continue under enhanced safety reporting.
Then it widened. In July 2026, the Drugs Controller General of India published a consolidated notice cancelling the registration certificates of seven further ethics committees, in Kolkata, New Delhi, Navi Mumbai, Ahmedabad, and Pune, citing serious deficiencies and non-compliance under the NDCT Rules 2019 and applicable GCP guidance. Per that notice, the cancellations themselves had occurred over the preceding twelve months rather than on the notice date. The notice did not create them. It made them visible in one document.
That sequence is the point. India is inspecting registered ethics committees systematically, and the result is that review capacity is being removed from a market where capacity was already concentrated. One administrative action took a large oncology network out of new-study supply. The July notice took out seven more review bodies, several of them independent committees serving multiple sites.
For a sponsor holding a site allocation at any of them, that is a re-routing problem this quarter. For everyone else it is a forecast. Ethics committee status is not static infrastructure to be confirmed once at feasibility and assumed thereafter. There is a documented mechanism by which it fails, and confirming it belongs on a recurring cycle.
Early-phase capacity is concentrated in a handful of institutions
The third constraint is physical. In our landscape assessment as of 5 August 2026, verified Phase 1 oncology capability in India sits in approximately four institutions.
Tata Memorial Centre operates a dedicated Clinical Pharmacology and Phase I Trials department in oncology at ACTREC, with a first-in-human oncology track record and a place in the ICMR Phase 1 clinical trial unit network. We did not locate a second dedicated oncology Phase I department elsewhere in India, which is a not-located finding rather than a claim of exclusivity. CMC Vellore has completed point-of-care CAR-T Phase 1 work. Tata Medical Center Kolkata carries named early-phase dose-escalation site experience. PGIMER Chandigarh operates a designated Phase 1 clinical trial unit, though we did not locate evidence of oncology first-in-human hosting there specifically. A further seven institutions, on our classification, show credible early-phase participation without a verifiable dedicated unit. Those are our gradings, not official designations.
One further observation belongs here, stated as an observation. In our 5 August 2026 snapshot we did not identify any healthy-volunteer early-phase oncology pharmacology studies registered in India. That is consistent with global oncology practice, where first-in-human work runs in patients rather than healthy volunteers by design, and it means early-phase oncology molecules in India go directly to patients at that handful of centres.
The constraint here is not patients. India carries roughly 1.4 to 1.5 million new cancer cases a year. The constraint is activation-ready slots, and a sponsor cannot resolve a slot constraint by adding recruitment budget.
Market structure as of 5 August 2026
The counts in this section come from our own snapshot of the ClinicalTrials.gov API v2 and secondary CTRI mirrors on 5 August 2026, not from official aggregate statistics. Registry positions move daily. Re-check before relying on them.
In that snapshot, 326 active oncology studies listed India sites. Of those, 182 were led by foreign commercial sponsors across 38 lead-sponsor groups after de-duplicating by corporate group. One sponsor accounted for 65 of them. The top four held 115 of 182, roughly 63 percent of foreign-sponsored India-site activity. At Phase 3, India functions as a recruitment geography inside a global protocol.
At Phase 1 the ratio inverts. We enumerated thirteen active pure Phase 1 oncology trials, and twenty-four including Phase 1/2 and 1b/2 designs. Nine of the thirteen pure Phase 1 sponsors were India-headquartered. The multinational presence at first-in-human was three trials. The same companies running dozens of Phase 3 studies in India do not bring dose-escalation work there.
Six of those twenty-four early-phase trials were cell therapies. A quarter of India's early-phase portfolio, on that count, sits in a modality that barely existed there five years ago, run largely by organisations navigating CDSCO for the first time without in-house regulatory infrastructure.
Ahead of that, we identified ten organisations that have publicly committed to oncology trials in India for which we could not locate a registration. Six of the ten are cell and gene therapy programmes. Four carry explicit Indian capital or capacity behind them. Two are already inside CDSCO's process.
The direction of activation demand in India is domestic, cell-therapy-led, and funded.
What this means for activation architecture
Three observations, stated at the rung the evidence supports.
The bottleneck is downstream of the regulator. The regulatory clock is legislated and comparatively predictable. Ethics committee status, site qualification, and the reconciliation work a sponsor must complete before it can even name its options are none of those things. An activation plan that does not model them separately is not modelling the critical path.
The friction is verifiable and timestamped. The suspension has a date. The cancellations have a notice. The protocol revision has a version number. Activation friction in India is not anecdotal, which means it is addressable by architecture rather than only by relationships.
When a regulator has approved a trial and no registry shows it, the answer is not a better guess. It is an attested record of what was checked, against which source, on which date, and by whom, including an explicit NOT LOCATED where the search came back empty. The counts above are usable because they carry that provenance, including the places where provenance runs out. Fifteen leads in the underlying landscape work are recorded as unresolved rather than estimated, because a stated UNKNOWN is defensible and a guessed answer is not.
That is the principle behind Trial Activation Intelligence, and the reason NexTrial builds against verification rather than prediction. The Three-Gate Verification System is the verification architecture underneath it. As of August 2026, Gate 2 has shipped and is in first validation with one sponsor, and a Lean 4 formal verification layer is in build and not yet live. Evidence, not substitution. The human decides.
FAQ
How long does CDSCO take to approve a clinical trial in India?
The New Drugs and Clinical Trials Rules, 2019 (G.S.R. 227(E)) set a decision window of 30 working days for drugs discovered and developed in India and intended to be manufactured and marketed in India, with deemed approval if the regulator does not respond within that period, and 90 working days for other new drugs, including those developed outside India. Those windows cover the regulatory decision only. They do not cover ethics committee review, site contracting, or import licensing, which run separately and are where most activation time is actually spent.
What is the difference between CDSCO approval and CTRI registration?
They are two separate events with two separate owners. CDSCO, through the DCGI and its Subject Expert Committees, decides whether a trial may proceed. The Clinical Trials Registry - India, maintained by ICMR, is a WHO primary registry and the public record that a trial exists. A trial can hold a documented regulatory recommendation and have no public registration that a sponsor or competitor can locate. Planning against either source alone produces an incomplete picture.
Do clinical trials in India need to be registered on both CTRI and ClinicalTrials.gov?
Indian rules expect CTRI registration for trials conducted in India. Global sponsors running multi-country protocols commonly register on ClinicalTrials.gov as well, but that is practice rather than a domestic legal requirement, and dual registration is inconsistent. Sponsors appear in one registry only in both directions, and identifiers do not always reconcile across the two. Any India trial count drawn from a single registry should be treated as a floor rather than a total.
Which institutions in India can run Phase 1 oncology trials?
Verified dedicated early-phase oncology capability is concentrated in a small number of institutions. Tata Memorial Centre operates a dedicated Clinical Pharmacology and Phase I Trials department in oncology at ACTREC, and is part of the ICMR Phase 1 clinical trial unit network. CMC Vellore and Tata Medical Center Kolkata carry documented early-phase execution. PGIMER Chandigarh operates a designated Phase 1 unit, though we did not locate evidence of oncology first-in-human hosting there specifically. In our own landscape assessment as of 5 August 2026, roughly seven further institutions show credible early-phase participation without a verifiable dedicated unit. These are our classifications, not official designations.
Why are ethics committee registrations being cancelled in India?
Under the New Drugs and Clinical Trials Rules, 2019 and applicable GCP guidance, ethics committees must register with CDSCO, and the DCGI is empowered to suspend or cancel that registration for non-compliance. India runs a risk-based inspection programme against registered committees on that basis. In July 2026, per a consolidated DCGI notice and contemporaneous press reporting, the registration certificates of seven ethics committees across five cities were cancelled for serious violations identified in inspections. The practical consequence for sponsors is that ethics committee status is a live variable with a documented failure mechanism, not fixed infrastructure to be confirmed once at feasibility.
What is Trial Activation Intelligence?
Trial Activation Intelligence is a verification category covering the work required to move a protocol from final draft to First Patient In: regulatory packet construction, jurisdiction-specific conformance checking, ethics committee and site qualification, and the evidence trail supporting each decision. It produces evidence for a human reviewer rather than replacing the reviewer, which is why the output of a Trial Activation Intelligence system is an attested record of what was checked, against which source, and on what date.
Method and provenance
Every claim in the underlying landscape work carries one of three grades. VERIFIED means confirmed against a primary source. REPORTED means single-source or trade-press attributed. INFERRED means reasoned from adjacent evidence and labelled as such.
Negative findings are handled separately and deliberately, because two different claims are easy to confuse. A verified absence means a source that should contain a record does not contain it. A not-located finding means our search of a stated source set on a stated date returned nothing. Every negative statement in this piece is the second kind unless it says otherwise. Not located is not the same as does not exist, and this piece does not use the first to mean the second.
One limitation is stated rather than absorbed. The CTRI search interface is CAPTCHA-gated, so CTRI-only entries were enumerated through registry mirrors and secondary sources. Those entries are a floor, not a ceiling. Academic trials registered on CTRI without press coverage may exist beyond them.
Counts and capability gradings are current as of 5 August 2026 and should be re-checked before they are relied on. Product status statements are current as of August 2026.