Cancer drug discovery: unravelling the mysteries of halichondrin

Chris Lo 5 November 2019 (Last Updated October 31st, 2019 16:09)

A joint Harvard-Eisai project has achieved a breakthrough by fully synthesising the highly potent anti-cancer molecule halichondrin, sourced from marine sponges, at scale. It has taken 30 years of research to reach this point, so what has made halichondrin synthesis so difficult, and what are the implications for the treatment of cancer now that the molecule can be synthesised for clinical testing?

Cancer drug discovery: unravelling the mysteries of halichondrin
A close-up shot of halichondria panicea, a marine sponge similar to the origin point for the halichondrin molecule discovery in the late 80s. Credit: Minette Layne

In June, the Kishi Lab at Harvard University’s Department of Chemistry and Chemical Biology made an important announcement. A team at the lab, in collaboration with researchers from Japanese pharma company Eisai and led by Morris Loeb professor of chemistry, emeritus, Yoshito Kishi, had achieved the world’s first total synthesis of the halichondrin molecule class.

Understanding the significance of the Kishi Lab’s breakthrough requires winding the clock back to 1986, when Japanese scientists Hirata and Uemura first isolated halichondrin B from the marine sponge halichondria okadai. The researchers reported that the natural product exhibited powerful anti-cancer activity in mouse models, fuelling hope that halichondrin could be developed into a powerful new class of cancer therapeutics.

The main problem was a lack of natural availability of halichondrin, and the sheer complexity of this molecular class made it impossible, at the time, to synthesise it at any meaningful scale to start clinical testing.

“The structure of the complete halichondrin molecule is particularly challenging to replicate because it has 32 chiral centres, asymmetrical points that must each be correctly oriented,” explains Dr Takashi Owa, chief discovery officer for Eisai’s oncology business unit. “In other words, there are at least 4.2 billion ways to get it wrong.”

Small-scale halichondrin synthesis in the 90s

Kishi, a pioneer in chemical synthesis and one of the minds behind the Nozaki-Hiyama-Kishi reaction, was the first to synthesise the halichondrin B molecule in 1992. The sequence to get there was arduous and inefficient – Owa notes the project required “more than 100 chemical reactions and produced less than a 1% overall yield” – but the discovery’s implications were still significant.

The process of halichondrin B synthesis was licensed to Eisai in 1993, and the company quickly initiated a related drug discovery programme at Eisai Research Laboratories in Andover, Massachusetts. For the rest of the decade, Eisai scientists synthesised around 200 right-half halichondrin analogues (“None of the left half intermediates submitted for testing were found to be active,” Owa says) for biological evaluation.

Those efforts led to the discovery of drug candidate eribulin, which Eisai developed into an effective therapy, approved by the US Food and Drug Administration (FDA) in 2010 for advanced metastatic breast cancer, and approved in 2016 for soft tissue sarcoma. Eribulin, marketed as Halaven, is an important pillar of Eisai’s oncology portfolio and still drives drug development programmes today.

“We are still exploring combination regimens of eribulin with other oncology drugs such as the anti-PD-1 monoclonal antibody pembrolizumab [Keytruda] and the CXCR4 antagonist Balixafortide in the breast cancer space,” says Owa.

Synthesis at scale drives the next step for halichondrin in oncology

Now that the Kishi Lab has achieved total synthesis of halichondrin at scale – 11.5g of drug candidate E7130 at 99.81% purity – Eisai is moving fast to translate the discovery into a new clinical product. A Phase I clinical trial assessing E7130 against solid tumours has been ongoing since February 2018, and the company also plans to incorporate the new candidate into two studies of eribulin.

“From our preclinical data, E7130 does not appear to be a simple anti-tubulin agent but can target the tumour microenvironment [TME] in a quite unique fashion, through the increase in intratumoral CD31-positive endothelial cells (vascular remodeling effect) and the decrease in alpha-SMA (smooth muscle actin)-positive cancer-associated fibroblasts (anti-CAF effect),” Owa explains.

“Both of these effects may be connected with the amelioration of the TME, resulting in the enhancement of combination drugs’ anti-tumour activity in preclinical animal models. Based on these observations, we would expect clinical activity of E7130 not only in monotherapy but also in combination with other oncology drugs, including immuno-oncology agents.”

The total synthesis of halichondrin at scale may have been more than three decades in the making, but Eisai’s proactive clinical development programme is clearly aiming to make up for lost time. With today’s focus on the TME as a driver of drug sensitivity and resistance in cancers, E7130’s TME-targeting mechanism of action could be an important facilitator of more effective therapies.

“Given the fact that the immunosuppressive tumour microenvironment is one of the

key resistant factors for the currently available immuno-oncology therapies, we would expect the TME-ameliorative effect of E7130 may contribute to a cure for cancer,” says Owa.

If E7130 does prove central to a major breakthrough in the tumour-killing power of immuno-oncology treatments, the world will owe a debt to the teams at Eisai and the Kishi Lab, as well as the humble sea sponge.