We conclude our look at AMNOG and cancer drugs with some insights from the benefit assessment process and focus on the use of outcomes beyond survival gain to show benefit.
As noted in part one of this short series, drugs coming to the German market no longer have free pricing and are instead reimbursed at the manufacturer’s price only during the first year on the market. During this time the Federal Joint Committee (G-BA) assesses the new product, or indication, for a benefit relative to comparator/s selected by the G-BA.
Usually the G-BA also seeks the advice of expert Health Technology Assessment (HTA) body the Institute for Quality and Efficiency in Healthcare (IQWIG) and the G-BA also rules on the strength of evidence of the added benefit.
Patient-relevant outcomes matter
Once a manufacturer has met the initial requirement of strong data against a G-BA accepted trial comparator, in a defined patient population, the clinical trial data is assessed to determine the level of added benefit. The G-BA assesses drugs on four criteria: mortality; morbidity (for example, symptom improvement); quality of life; and adverse events (toxicity and side effects).
The gold standard in oncology drug market access in Germany remains a clinically significant overall survival (OS) benefit (which fits in the ‘mortality’ criterion) in a head-to-head trial against a comparator, and achieving clinically significant survival gain is the simplest way to establish additional benefit. However, a drug may not have overall survival demonstrated for a number of reasons: a trial may have failed to achieve significance on this end-point, or – as in the case of Xalkori, below – the drug can achieve accelerated approval on the basis of an interim analysis or phase II data before OS is established. Another possibility is that OS is not a viable primary outcome of the clinical trial – for example, if a product is targeting approval at first line, where later therapy lines or crossover confound analysis of survival benefit ascribed to that product, or if the expected survival for a condition is several years and a clinical trial to establish survival benefit is not practical.
In the case where no overall survival benefit against a comparator has been demonstrated, the G-BA has demonstrated that it will consider ‘softer’ patient-relevant endpoints as sufficient to establish an added benefit.
Xalkori: Symptom improvement and quality of life gains add benefit
In the case of Xalkori (crizotinib), Pfizer’s product targeting a subsegment of non-small cell lung cancer (NSCLC) in which the tumour shows an excess of a particular biomarker, accelerated approval meant that no clinically significant OS was shown at the time of European approval and German assessment. There was evidence for progression-free survival, and the manufacturer had also collected data on symptom improvements and quality of life.
The initial IQWiG assessment found no benefit, because there was no phase III trial showing OS gain, and the quality of life gain was cancelled out by higher toxicity than the comparator chemotherapies – this is a useful methodological insight in itself, showing how the categories will be evaluated separately and then weighed against each other.
AMNOG encompasses re-assessment of drugs after some time on the market
However, the company was able to submit a re-analysis of morbidity and quality of life data. This led the G-BA to find a net benefit for Xalkori, despite the lack of overall survival benefit, because of benefits in morbidity (symptom improvement) and quality of life, not cancelled out by toxicity. Progression-free survival was found not to add benefit directly in this case, because it is a surrogate end-point and was trumped by the existence of direct measures (overall survival – though not with sufficient significance – and morbidity).
Show added benefit
Manufacturers that may rely on patient-relevant outcomes (PROs) will need to include these in phase III trial design. It is particularly important to be able to analyse them for patient subgroups (see above), which may require a larger trial or different planning – more patients will have mortality data than will fill out a quality of life questionnaire, for instance.
It is important to separate observable patient-relevant outcomes from surrogate measures based on biomarker analysis or pathological assessment, even though these may overlap: for example, in the case of Zytiga’s assessment in its second indication of chemotherapy-naïve prostate cancer, the G-BA found benefit in the morbidity component ‘time to onset of pain medication’, but not in ‘time to use of chemotherapy’ because this could be influenced by factors other than patient experience, such as a radiographic assessment of progression.
The AMNOG legislation encompasses re-assessment of drugs after some time on the market, and there is therefore potential for price adjustment later in a product’s life-cycle. Manufacturers should engage in post-marketing studies, especially around patient-relevant outcomes, to uphold product value and prevent price reduction.
| Example of non-mortality outcome use in early benefit assessments | ||
| Product | Details of benefit categories used | Outcome |
| Xalkori (crizotinib) in NSCLC | No OS benefit shown but positive data in morbidity (symptom improvement) and QoL, in one patient segment | Benefit in morbidity and mortality not cancelled out by some greater toxicity, allowing an overall benefit and price negotiation |
What does the benefit finding mean for price?
A benefit assessment is only one step along the path to a final, reimbursed drug price and German market access. A natural question to ask is how the benefit finding impacts the final price of a drug in Germany. The G-BA does not set the price – however, an initial positive benefit finding sends a product forward for price negotiation with the sickness funds. A ‘no benefit’ finding, on the other hand, means reference pricing and the possibility of an unacceptably low price.
It was thought at the start of the AMNOG era that discounts may show a simple correlation with benefit levels found – that is, higher added benefit, with stronger evidence, would be associated with a lower discount from launch price.
It is now clear that while the G-BA’s additional benefit finding is important for setting the context for pricing negotiations between a manufacturer and the sickness funds, the relationship between added benefit level and price discount as a percentage of launched price, is more complex.
Factors that affect a drug’s final reimbursed price include:
- Level of added benefit and strength of the evidence
- Segmentation and comparators. If a benefit is found only for a small subset of the label indication but the reimbursement is for the whole population, this may affect negotiations
- Company choice of launch price (which may be influenced by the manufacturer’s knowledge of previous German discount arrangements)
- Negotiation strategy and skill of the manufacturer and Germany’s statutory health insurance the GKV
- Contractual terms. Because the German price is used as a reference in other European markets and even further afield, manufacturers may look to more creative contractual terms to maximise list price. Although the negotiated discount is to be published for the setting of pharmacists’ and wholesalers’ drug distribution margins, there can be terms within a negotiated discount contract that affect the ‘real’ pricing and use of a drug. Manufacturers should consider contract options, including terms setting additional discounting for individual sickness funds, or a voluntary agreement to promote a drug only for the patient subset in which the G-BA found a benefit.
Overcoming the new challenges
With the advent of the AMNOG assessment, manufacturers face new challenges in gaining acceptable pricing for novel oncology drugs in Germany. A key step is a positive assessment by the G-BA of a drug’s added benefit relative to comparators. Analysis of decisions on cancer drugs over the past three years yields two major lessons for oncology drug manufacturers:
- As outlined in part one, manufacturers should prepare for segmentation of the target patient population and the clinical trial population, by identifying likely subgroups and comparators and deciding whether to amass evidence of added benefit in each group
- The traditional esteem for ‘hard’ overall survival data is still present, but drugs can survive without OS data if the manufacturer has planned the trial programme around patient-relevant outcomes: symptom improvement, quality of life and tolerability may all be enough to show benefit.
Click here to read Oncology drugs under AMNOG: part one
