Why the GKV’s “Cannabis Reform” Cannot Stand

The legal case against stripping cannabis flower from statutory health insurance coverage is not complicated. The German government simply chose to ignore it

This story was originally published on Substack and is also available in German.

The GKV-Beitragssatzstabilisierungsgesetz, adopted on 10 July 2026, removes cannabis flower from statutory health insurance reimbursement as of January 1, 2027. The government has presented this as a routine cost-containment measure.

It is not — starting with the appalling decision to transfer a huge amount of money and potentially trial data to one Bavarian Billionaire. Not so conveniently, as it turns out, particularly as the lingering scandal of the COVID mask purchase mixes with recent memories of Jens Spahn’s removal, and the fact that Vertanical is also, unsurprisingly, a coalition government political donor. The fact that current Health Minister Nina Warken appears to have attempted to cover up Spahn’s knowledge and actions around a still swirling multi billion euro mask screw up won’t help matters.

Neither will the clearly documented political involvement of Markus Soder and other CDU party big wigs, with Clemens Fischer, founder of Vertanical, the pharma company behind a drug that has literally been put in the cat bird seat to make a huge amount of money out of a clearly mislabelled “budget bill” that also asks Germans far from cannabis patients, to shell out even more money for the shrinking services on offer. This giveaway includes being able to set his own price during the first six months of market introduction. Also conveniently coming soon this September.

While Georg Würth and I rarely agree on anything beyond basic cannabis reform, you have to hand it to DHV on this one. Namely, as he said on the day the government passed the bill, “It looks very much as if someone bought themselves a monopoly for their product from the political parties for half a million euros, to the detriment of the insured.”

However, beyond this, the bill is also clearly an attempt to legislate around a constitutional framework that has been settled for over twenty years, impose a mandatory clinical trial that violates the data protection rights of the patients it purports to benefit, and create a market architecture that routes those patients toward a single manufacturer’s product without their consent.

The legal problems are not marginal objections. They go to the foundation of the reform. Each stands independently. Together, they make this one of the most constitutionally exposed pieces of health legislation Germany has passed in recent memory.

The Foundation: Twenty Years of Constitutional Social Law the Amendment Does Not Engage

The starting point is not the 2017 cannabis law. It is the Nikolaus-Beschluss — the Federal Constitutional Court’s decision of 6 December 2005 (1 BvR 347/98). The case arose from an 18-year-old with Duchenne muscular dystrophy whose health insurer refused to cover a treatment outside the approved GKV catalogue. The court found the refusal unconstitutional. The principle it established: where a patient suffers from a life-threatening illness or one that lastingly impairs quality of life, and no adequate standard treatment is available, the GKV cannot deny coverage on the basis that the treatment falls outside the approved catalogue. That obligation derives from Articles 2(1) and 2(2) of the Grundgesetz, read together with the Sozialstaatsprinzip.

That ruling did not emerge from nowhere. It built on off-label coverage doctrine developed by the Federal Social Court and Administrative Courts running back to at least 2000, including the Sandoglobulin line of cases that established the definition of „schwerwiegende Erkrankung“ — serious illness — as a condition that is life-threatening or lastingly impairs quality of life, and that fixed off-label reimbursement on a documented absence of adequate therapeutic alternatives rather than on the regulatory status of the treatment.

When the Federal Administrative Court ruled in 2016 that the state must provide cannabis access to seriously ill patients who have no adequate and affordable alternative, it was applying this framework directly. The 2017 legislature, when it enacted § 31 Abs. 6 SGB V, was not inventing a new right. It was inserting cannabis-specific coverage into an already-settled body of doctrine that had existed for fifteen years.

The 2026 amendment does not argue that this framework is wrong. It does not engage the Nikolaus-Beschluss. It does not assert that individualised clinical assessment is the incorrect standard. It does not claim that the patients who obtained MDK approvals under that framework were wrongly approved. It simply removes the entire product category from the reimbursement system, without touching the constitutional obligations that put it there.

That is the critical distinction. A legislature can amend a statute. It cannot, by amending a statute, extinguish constitutional obligations that the statute was enacted to fulfil. A rule that displaces a doctrine older than the statute it amends, without engaging that doctrine at all, is a materially different and more serious legislative act than an ordinary policy adjustment. The courts will be asked whether the amendment discharges the constitutional obligation. The legislative record does not provide an answer.

The Article 3 Problem: Two Patient Classes, One Regulatory Distinction

The reform creates two categories of GKV patient.

Patients whose conditions respond to standardised extracts or finished pharmaceutical products retain coverage, subject to standard benefit-assessment processes. Patients for whom cannabis flower is the therapeutically appropriate form lose coverage entirely, unless and until they complete a mandatory six-month trial of finished pharmaceutical products.

The distinction between these two groups is not based on a clinical finding that flower is less effective. It is not based on a finding that flower is unsafe. It is not based on individual assessment of any patient’s condition. It is based on the regulatory pathway that applies to their medicine: flower has not been through centralised pharmaceutical marketing authorisation; extracts and finished products have.

Article 3 of the Grundgesetz requires that comparable situations be treated comparably, and that different treatment be justified by a difference that is material to the purpose of the regulation. The purpose of GKV reimbursement is to fund medically necessary treatments. The regulatory pathway through which a treatment was approved is not a clinically material distinction for patients whose medical necessity has already been individually adjudicated by the MDK. Those patients did not receive approval because of how their medicine was regulated. They received it because, under the state’s own review process, it was established that their medicine was medically necessary and that no adequate alternative existed.

Withdrawing coverage from patients on the basis of their medicine’s regulatory pathway, rather than its clinical performance, is precisely the kind of arbitrary classification Article 3 exists to prohibit. The government has not argued that patients who rely on flower are clinically distinguishable from those who rely on extracts. It has simply drawn a line based on regulatory status and called it cost containment.

The GDPR Problem: A Trial Nobody Can Refuse

The mandatory six-month trial of finished pharmaceutical products is presented as a clinical safeguard — an opportunity for patients to try available alternatives before the state decides their previous treatment is no longer covered. It is not a clinical trial in any legally meaningful sense. It is a compulsory data-generation exercise, and the legal problems it creates are significant.

Under the new framework, patients who wish to retain any prospect of GKV coverage must undergo a trial of finished pharmaceutical products — products that GKV flower patients have already demonstrated, in their own MDK approval files, that they cannot tolerate. The oils and alcohol carriers used in every existing extract were documented contraindications in the approval process these patients already completed. The trial is not testing whether an alternative might work for them. It is requiring them to fail a test they have already failed, on the record, for a manufacturer who had no part in the original assessment.

The data generated by that trial — tolerability, adverse reactions, efficacy outcomes across a large, documented, chronically ill population — flows to the manufacturer of the product being trialled. Vertanical/FUTRUE, whose product Exilby is the primary finished pharmaceutical cannabis product entering the market at precisely the moment flower coverage ends, is simultaneously conducting Phase 3 trials in the United States. Real-world tolerability and efficacy data from a large involuntary European patient population is exactly the kind of commercial asset that accelerates regulatory pathways in the US and UK markets. That asset is being generated by German law, not by voluntary participation in a research protocol.

This is where GDPR Article 9 becomes directly engaged. Health data is a special category under Article 9. Its processing requires explicit, freely given, informed consent — or a specific legal basis that satisfies the proportionality requirements of the Grundgesetz and European fundamental rights law. A patient who must participate in a trial or lose coverage has not given freely given consent. The element of compulsion — the loss of insurance coverage as the consequence of refusal — destroys the voluntariness that Article 9 consent requires.

The Federal Constitutional Court has recognised the right to informational self-determination as a fundamental constitutional right since its landmark Census Judgment of 1983, derived from Article 2(1) in conjunction with Article 1(1) GG. Patients are entitled to control the collection and use of their personal medical data except where constitutionally justified limitations apply. Being compelled to generate clinical data for a private company’s product under threat of losing insurance coverage does not meet that standard. It is closer to the opposite.

The pharmacy verification obligations compound this problem. Under the new framework, pharmacies must verify prescription compliance — determining whether a prescription originates from a first consultation, a repeat prescription, or an EU telemedicine provider operating under different rules. Legal experts who appeared during the parliamentary consultation stated directly that performing those checks requires access to confidential patient records beyond what dispensing ordinarily requires. That obligation cannot be reconciled with either GDPR or the constitutional right to informational self-determination. The legislation provides no answer to this tension.

The Exilby Problem: A Funnel, Not a Trial

The legal problems described above would exist even if the mandatory trial were designed fairly. It is not.

A genuine clinical trial of alternatives to cannabis flower would include flower as a control arm — the most obvious methodological requirement for a study designed to establish whether alternatives are equivalent. Flower is explicitly excluded from the trial. The trial tests alternatives to flower, while excluding flower as one of the alternatives being tested. That is not a clinical protocol. It is a funnel.

The funnel routes patients toward Exilby, Vertanical/FUTRUE’s product, which received federal marketing authorization on June 9, 2026 — one month before the Bundestag voted to remove flower coverage on July 10. Exilby is the first finished cannabis medicine in Europe specifically approved for chronic pain. It will enter a market from which its cheaper, more accessible competitors are being simultaneously removed. For the first six to eighteen months of market entry, its manufacturer is free to set its own price under the standard AMNOG new-product pricing rules — there are no caps.

Vertanical’s own regulatory filings state explicitly that the clinical evidence supporting Exilby cannot be extrapolated to other cannabis extracts or to prescribed flower. The government points to Exilby as evidence that equally effective alternatives now exist for flower patients. The manufacturer says its own evidence does not support that claim. The government uses the product’s existence regardless.

The result is a mandatory trial that: excludes the product whose coverage is being withdrawn from its own control arm; routes patients toward a single manufacturer’s product; generates involuntary real-world clinical data for that manufacturer’s US and UK regulatory submissions; allows the manufacturer to set its own price during the trial period; and rests on a justification — that Exilby demonstrates adequate alternatives exist — that Exilby’s own manufacturer has contradicted in its regulatory filings.

None of this is incidental. Taken together, it describes a legislative arrangement designed not to protect patients but to channel them.

What the Government Has Not Demonstrated

The constitutional obligation established by the Nikolaus-Beschluss, confirmed by the 2016 Bundesverwaltungsgericht ruling, and implemented through § 31 Abs. 6 SGB V has a specific content: the state cannot withdraw coverage from seriously ill patients who have no equally effective and affordable alternative.

To discharge that obligation through the current reform, the government would need to demonstrate — not assert, demonstrate — that equally effective and affordable alternatives exist for the 65,000 patients who currently rely on cannabis flower. The legislative record does not contain that demonstration. It contains an assertion that Exilby has been approved and that extracts remain available. It does not address Exilby’s own statement that its evidence cannot be generalised to flower patients. It does not address the six-month trial requirement that must precede extract coverage. It does not address the documented contraindications — oil and alcohol intolerance — that are on file for the majority of GKV flower patients. It does not address the price point at which Exilby will enter the market, or whether that price satisfies the „affordable“ element of the constitutional standard.

Constitutional obligations are discharged through evidence. In fact, it appears that the current GKV bill ignored it’s own findings and adopted, almost wholesale, AOK’s lobbying language (AOK is one of the largest public health insurance companies) beyond the Exilby funnel.

What the government has produced is not a fiscal savings bill. Instead, it is a product approval pipeline mislabeled as a budget bill — and worse one that is grotesquely dishonest. Those are not the same thing as evidence that 65,000 individually adjudicated medical necessity findings have been superseded by the availability of a newly approved pharmaceutical whose own manufacturer says it is not equivalent to what those patients currently use.

Hopefully the taxpayers will not be on the hook for a multi-billion euro settlement on this one either.

Where This Goes Next

Opposition parties have already filed legal challenges to procedural aspects of the reform. The substantive constitutional questions will follow through the courts, through administrative proceedings, and through individual patient cases. The 2016 Bundesverwaltungsgericht ruling that created the current framework was itself the product of a single patient’s administrative challenge that ran for years. The patients who built that jurisprudence did not do so quickly. But they did so successfully.

The reform proceeds on the assumption that the constitutional questions it avoids will not be asked. They will be. The Nikolaus-Beschluss established that the German constitutional order does not permit the state to leave seriously ill patients without access to treatments that work for them when no adequate alternative exists, simply because those treatments are inconvenient to the budget.

That principle has not been overturned. It has been engineered away. Unfortunately for the Bundestag, it has also created a constitutionally unacceptable, highly discriminatory bill that targets those historically least able to fight back. Given the pending lawsuits now being scribed across the country, this time they will be taught, yet again, that this community doesn’t give up. This is, after all, just a blip in the long history of cannabis reform. There are millions of patients bodies along that trail. Hopefully, this time, that undue suffering can be prevented before it claims a single life.

The next steps are also very clear for those now in the trenches. The courts will be asked whether this may stand — even if for now it is one patient’s emergency motion and hauptklage (master law suit) at a time.

I write about AI, regulation, constitutional governance, and who gets to decide how emerging technologies — and the laws surrounding them — shape society.