Excitement around the new wave of T-cell therapies against cancer is unlikely to abate anytime soon with the FDA approval of the first two products: Kymriah (Novartis) and Yescarta (Gilead – formerly Kite). Novartis will soon compete with Gilead’s Yescarta in DLBCL – a subset of non-Hodgkin’s lymphoma. This illustrates the dilemma of the first set of T-cell therapies – they all target B-cell antigens, and often CD-19, so potentially all compete in a limited set of patients. The September Edison report on T-cell Cancer Therapies examines the following: refractory and relapsed B-cell cancers cover 1.4% of U.S. cancers and about 1% of deaths, at least 6,500 cases. The major T-cell therapy opportunities are in multiple myeloma (MM), acute myeloid leukemia (AML) and major solid cancers with over 1.2 million new U.S. cases and 450,000 deaths a year.
 
Any T-cell therapy needs an antigen target that is abundant on cancer cells and not found on normal cells. This is a very rare set of attributes. CAR-T competes in MM and perhaps in AML but lacks the antigens to attack solid cancers. However, Celyad’s NKR CAR T-cell therapy targets stress antigens with multi-indication potential in AML, MM and solid cancers. It has shown a clinically significant response in AML with two initial responses in colorectal cancer. The T-cell receptor approach has high specificity and versatility but with specific patient segmentation. Non-cellular therapies (bi-specific T-cell Engagers (BiTEs) and already approved checkpoint inhibitors) could be synergistic.
 
One of the most interesting developments is pricing. In ALL, Kymriah is $475,000 but Novartis offers money back if no complete response (CR) is seen. This is targeted for pediatric use and offers good value with 83% CR rates. Yescata in DLBCL has a 51% complete response rate and Gilead has set a lower price of $373,000. This is an older target population (60+) so in the U.S., it remains to be seen whether Medicare will cover this. In addition, there are also a lot of associated hospital costs to CAR-T therapies, as these are risky treatments.
 
Another unresolved question is how these living therapies will be manufactured and delivered. They are all currently autologous – that is white cells are harvested from the patient by Leukapheresis, sent to a central facility, genetically transformed by virus and, after extensive QC, sent back for infusion into the patient. This is a key logistical chain as shipping must be under correct conditions and the right cells given to the right patient. A further factor is the variable quality of the patient cells as they arrive at the processing facility. The T-cell transformation process can be standardized, but the product cannot be. In addition, in use, the current therapies (but not Celyad’s NKR CAR T-cell approach) use pre-conditioning to expand the cells perhaps a thousand-fold inside the patient. As a result, current CAR-T therapies are highly variable products.
 
Novartis claims there is a 22-day turnaround time on this process. It has two manufacturing facilities in New Jersey and Europe. Segregating the facilities will be an important management issue as any contamination could close the entire plant. Gilead has a 17-day process turnaround and uses a 4,000-5,000-dose plant near Los Angeles airport. This plant could generate $1.5 billion of sales a year. Costs will be high but mainly consist of virus, fixed overhead and staff.  International shipping is feasible but these therapies cannot be held up by bureaucratic delays.
A factor not often considered in public debate, but of intense industry interest, is scale up. Each batch currently is for one patient. The cell culture processes are little changed from the research lab, so clean room space governs capacity. Automated systems are under development, as from Miltenyi Biotec amongst others, and will be needed if T-cell therapies are to be more than a minor niche.
 
The same applies to retroviruses, which integrate the CAR genes into the T-cell genome. There are two types used. One is lentivirus, which works in non-dividing cells. Lentivirus is complicated to make. Oxford Biomedica (who supplies virus for Kymriah) is working on high-capacity fermenters but currently floor space determines capacity. The other system uses gamma retroviruses, as in Yescarta. This needs to be used in growing T-cells. All GMP grade viruses are very expensive and order times are currently long. 
 
For the long-term, allogeneic T-cell therapies need to be developed. The first (UCART123, Cellectis) entered a trial in 2017 but ran into toxicity issues. Allogeneic technology, to avoid Graft versus Host disease, has been patented by Celyad and licensed by Novartis. By making large batches of possibly semi-tissue matched CAR or TCR T-cells, scale economies might be achieved, time to administration reduced to hours or days, and costs lowered.
 
At the current prices and limited capacity, it is hard to see mass CAR or TCR use against mainstream solid cancers (assuming that they work) where the majority of clinical need occurs. Yet they could revolutionize cancer therapy. While, both clinical and process improvements are needed, this whole area is likely to be a focus for much innovation and investment.

---

Dr. John Savin is an analyst working on biotech, pharma, medical device and diagnostics companies. As founder CEO of Physiomics plc, he devised the strategy, raised funds and took the company to AIM in 2004. At Greig Middleton, John was director in charge of the pharma and biotech analyst team and worked with corporate finance on fund-raising, IPOs and corporate restructuring. He has an industry background in sales and marketing with GE Healthcare and AstraZeneca and is a co-author on a number of scientific publications. He has a Ph.D. in organic chemistry as well as MBA degrees.