Innate Possibilities

Might modulation of the innate immune
 response turn up the heat on "cold" tumours?

It's been (another) good year for cancer immunotherapy, marked by approval of the first autologous T cell therapies for otherwise untreatable haematological cancers, additional label indications for immune checkpoint inhibitors and even glimmers of hope around personalized cancer vaccines.

All of these advances exploit adaptive immunity- the body's capacity to recognise tumours (and invading pathogens) as "not self" and to programme the immune system to bring exquisitely specific antibodies and effector cells into the attack. A downside to this biological sophistication is that mounting an adaptive immune response takes time and can be deliberately misdirected. We rely on a more primitive and less selective innate immune response as a first defence against rapidly multiplying bacteria and viruses and to prime adaptive immunity.

Bringing the innate immune response into play as a means of increasing the efficacy of cancer immunotherapy is attractive. Non-responsiveness to immune checkpoint inhibitor therapy appears to correlate with tumour inflammation, rather than with immune checkpoint expression or the degree of tumour mutation. Turning “cold” (uninflamed) tumours “hot” (inflamed) might offer patients initially refractory to immunotherapy an additional treatment option.

A signature of tumour inflammation is the presence of IFN-β, a potent cytokine, the production of which is  triggered by STING ("stimulator of interferon genes") in response to molecules ("cyclic dinucleotides"- CDNs) that signal the presence of pathogen or host double-stranded DNA. CDNs produced in response to double-stranded DNA leaking from cancer cells are capable of activating STING and triggering IFN-β production, which in turn, stimulates cancer antigen-specific T cells.

STING activation has attracted the attention of Merck and BMS, rivals in the immune checkpoint inhibitor space. Two investigational STING activators (MK-1454: Merck and ADU-S100: Aduro Biotech/Novartis) are in early clinical evaluation. Both are synthetic CDNs, designed to be more potent than "natural" CDNs but, due to their chemistry, must be delivered directly into the tumour. Spring Bank Pharmaceuticals, BMS (through the acquisition of IFM Therapeutics), iTeos Therapeutics, Invivogen and GSK are also pursuing intratumoral CDN candidates, while Nimbus Therapeutics and CuraDev Pharma are developing small molecules that may allow oral dosing.

Other components of the innate immune system are of interest to immunotherapy developers.  "Toll-like receptors" (TLRs) recognise an array of bacterial and viral debris and are known to be expressed by various cancers. Past clinical studies with TLR-directed agents have been largely disappointing, although a TLR9 agonist, IMO-2125, developed by Idera Pharmaceuticals is under evaluation in combination with immune checkpoint inhibitors in melanoma patients who have previously failed immunotherapy.

"RIG 1 (retinoic acid inducible gene)-like receptors" (RLRs) sense viral infection and can eliminate infected cells, opening the possibility that RLR activation might be exploited to directly kill cancers. RLRs are largely activated by RNA and while perhaps not easily druggable, they offer a promising enough prospect for Merck to have acquired Rigontec, a pioneer in RLR research, for a headline figure of over $500 million.

Inflammasomes are multiprotein complexes of signalling molecules and enzymes that initiate and maintain inflammatory processes in infection and autoimmune disease. Inflammasomes are activated by various "NOD-like receptors" (NLRs), the most widely studied being NLRP3, a trigger sensitive to a wide range of microbial and "damage-associated" molecules, also environmental irritants (silica, asbestos) and amyloid-β, the hallmark protein of Alzheimer's disease. NLRP3 activation may prove to be a practical means of warming up cold tumours and BMS has ambitions to begin clinical studies with an NLRP3 activator in the next 12 months.

Inflammasome activation may not be without risk, being associated with both tumour promotion and suppression in different cancers. Certain tumour-expressed TLRs appears to contribute to development of a benign tumour microenvironment and promotion of metastasis and the STING signalling pathway can contribute to tumour development.  

"Pro-inflammatory" drug development faces the general challenge of achieving an effective degree of tumour inflammation without provoking potentially life-threatening "cytokine storms" or autoimmune adverse events. A better understanding of how the innate immune response might be modulated through the targeting of STING, TLRs, RLRs, inflammasomes or other elements could conceivably lead to novel or improved treatments for a spectrum of conditions that have chronic inflammation at the heart of their pathology. 

Photo credit: Rawich at FreeDigitalPhotos.net

Yet more on bugs and cancer

T-cells (red) on the attack

A Research Highlights piece in December’s Nature Reviews Cancer reports on another intriguing aspect of the interplay between our immune systems and the bugs we carry, namely how gut flora might influence the effectiveness of cancer immunotherapy.

Two international research groups set out to determine  whether the composition of the gut microbiome might influence the response to immunotherapy  directed against  PD-1, a so-called “immune checkpoint “ expressed by activated T cells and macrophages and which is exploited by cancer cells to switch off immune attack. Antibody-mediated blockade of the interaction between PD-1 and its ligand, PD-L1 can restore the anti-cancer response. The anti-PD-1 antibodies pembrolizumab and nivolumab (Opdivo® and Keytruda®, respectively) have proved their worth in the treatment of metastatic melanoma and a variety of other solid tumours.

Genetic analysis of faecal bacteria collected from cancer patients before and after anti-PD-1 immunotherapy found a correlation between gut bacteria diversity and the duration of progression-free survival in cancer patient after treatment.

A collaboration between US and French researchers found differences in the abundance of certain gut bacteria, with Faecalibacterium being enriched in melanoma patients responsive to anti‑PD1 therapy: Bacteroidales was enriched in those patients not responsive to immunotherapy. Differences were also found between responders and non-responders in regards to bacterial metabolism and the composition of immune cells found in the tumour microenvironment. Tumour-infiltrating “killer” T cells were more likely to be found in patients carrying an abundance of Faecalibacterium, while  immunosuppresive cells were more common in individuals carrying abundant Bacteroidales.

Another (again, predominantly American and French) research group found that the abundance of the gut bacterium Akkermansia muciniphila in non-small cell lung cancer and renal cancer patients correlated with a positive response to anti-PD-1 immunotherapy.

Both groups looked for possible mechanistic links between gut bacteria abundance and treatment response. When patient-derived gut bacteria were transplanted into germ-free mice, a variety of favourable effects on tumour growth and immune response were observed, including higher numbers of killer T-cells  and other, immune effector cells, along with changes in the expression of  T- cell receptors for key immune signalling molecules (“chemokines”).

The response to immunotherapy is difficult to predict and involves a variety of tumour factors (PD-L1 expression, tumour burden, degree of mutation) and host factors (immune system genetic makeup, T cell infiltration of the tumour). Analysis of the gut microbiome is unlikely to improve prediction of response, but preservation or manipulation of the gut microbiome through avoidance of antibiotic treatment prior to immunotherapy, or probiotic treatment to encourage “good” bacteria could conceivably translate into better and more sustainable response rates for at least some individuals.  

Photo credit : Rita Elena Serda.  National Cancer Institute \ Duncan Comprehensive Cancer Center at Baylor College of Medicine

More on bugs and cancer

A publication in the December issue of Cancer Research points towards another complicated relationship between bacteria and cancer risk.

A group headed by researchers at the Perlmutter Cancer Center (NYU Langone) looked at data gathered from more than 120,000 subjects already enrolled in an ongoing NCI-sponsored study looking at the link between nutrition and certain cancers.

The presence of a mouth-dwelling bacterium, Tannerella forsythia,  was associated with a 21% increase in the risk of oesophageal adenocarcinoma after adjusting for other known risk factors including smoking, drinking and body mass index.

In contrast, the presence of various Streptococcus and Neisseria species was associated with a 24% decrease in cancer risk. The presence of Porphyromonas gingivalis, a bacterium associated with gum disease, appeared to correlate with a higher risk of another form oesophageal cancer, oesophageal squamous cell carcinoma.

How mouth bacteria influence oesophageal cancer risk is not clear. An association between poor oral health and a higher risk of oesophageal cancer has been suggested in epidemiological studies. Neisseria are capable of partially detoxifying tobacco smoke, with lower numbers of Neisseria found in the mouths of smokers than in non-smokers.  Bacterial metabolism analysis hinted at an increase in oesophageal adenocarcinoma risk associated with some pathways but a lower risk with others. Certain metabolites produced by Neisseria sp correlated with the observed protective effect.

While cause and effect remains elusive, it’s possible that analysis of oral flora might eventually serve as a useful marker for oesophageal cancer risk and that manipulation of the oral flora could reduce occurrence in those already at higher risk through other behaviours.  What’s clear is that “local” microbiomes, whether mouth or gut, can have a profound effect on distant organs.

Bugs and cancer? The plot thickens...

An item from the New York Times gives me the chance to write about two great interests in the same blog piece: bacteria (once a microbiologist, always a microbiologist) and cancer.

Fusobacterium nucleatum, an
accomplice of colorectal cancer

That bacterial infection might cause or promote cancer was debated for most of the 20^th century, but with little solid evidence emerging to support the notions. During the 1980s, Barry Marshall and Robin Warren (the former famously swigging down a flask of culture broth to prove his hypothesis) established that Helicobacter pylori, a common corkscrew-shaped found in the stomach, was an undisputable cause of gastric inflammation and ulcers.

Epidemiological studies involving British, American and Japanese subjects confirmed that H.pylori carriage was indeed associated with an almost four-fold increase in the likelihood of developing gastric cancer and resulted in the WHO designating H.pylori as a Class I carcinogen.

Continuing research has established that the relationship between  H.pylori and cancer is not a simple one of cause and effect,  with H.pylori infection being a factor in some, but not all, forms of stomach cancer and that H.pylori  strains expressing a particular cytotoxin, “CagA”,  are more strongly associated with an elevated risk of cancer than are non-producing strains. Perversely, H.pylori infection appears to be associated with a lower risk of oesophageal cancer.

A more recently uncovered “smoking gun” is the presence of Fusobacterium nucleatum, a common mouth-dweller, found in higher numbers in around half of colorectal tumours than in the surrounding tissue. F.nucleatum- induced inflammation is cited as a plausible contributor to CRC initiation and progression.

But, as with the Helicobacter story, there is no clear-cut cause and effect between infection and cancer. Bacterial species are rarely solitary and the inhabitants of the local milieu or “microbiome” may be more important with respect to cancer initiation and/or progression than the presence of F.nucleatum alone.

CRC may spread to other organs and give rise to tumours in the liver. According to a recent Science publication, if F.nucleatum and its microbiome buddies are present in the original tumour, then they can accompany the metastasizing cancer and pitch up in the liver. CRC dwelling F.nucleatum remained associated with tumours even after their transplantation into mice. Moreover, dosing of tumour-bearing mice with an F.nucleatum-killing antibiotic slowed tumour growth.

Does this make a case for antibiotic therapy or vaccine development to reduce CRC rates? Well, not yet. Antibiotics therapy tends to ablate both the good and bad and, as is hinted at in immuno-oncology studies, certain gut bacteria might positively influence anti-cancer immune responses. And not all F.nucleatum strains might be bad guys. However, it’s feasible that getting a better handle on the mechanism(s) involved in the bacterial promotion of cancer might identify new interventions to improve outcomes or recurrence rates.

Photo credit: CDC Public Image Library