Spring 2022 (Volume 32, Number 1)
The Future of Rheumatology
By Reza Mirza, MD, FRCPC
Download PDF
Rheumatology is among the most rapidly advancing
specialties. Though oncology rivals us, we reserve
the pleasure of routinely attaining remission. As we
approach the 75th year since the discovery of cortisone,
the Journal of the Canadian Rheumatology Association (CRAJ)
asked me to opine: What do the next 75 years hold?
Rheumatologists will secure the holy grail of personalized
therapy. We have the means to modulate specific
cellular, protein, or nuclear targets at will. The shot-gun,
“inhibit-everything” approach employed glucocorticoids
may work short-term but is not tenable in the long run. The
burning question is what to target and in whom.
We know clinical phenotypes are insufficient to predict
biochemical response. Small Japanese trials give us a
taste of how to predict, using lymphocyte flow cytometry
patterns.1 We must continue to stratify our patients into
more meaningful groups, as doctors have always done. We
have the resolution: long-read sequencing, -omics, mass
cytometry. We simply need to offer this data with high
quality longitudinal treatment-response data to the altar
of machine intelligence.
We are already experts at measuring and extrapolating
phenotypes associated with the B-cell receptor, particularly
in its soluble antibody form (we are ever grateful, Marv).
We must proceed beyond the anti-cellular antibodies that
define modern day rheumatology.
Our characterization of the innate arm is scanty, and we
have less than half the lymphocyte story. Like Alice in Wonderland,
we are blind to the wizard behind the curtain, the
T-lymphocyte and its receptor. Unlike B-cells, T-cells can
survey intracellular processes, can induce apoptosis where
they stand, and their repertoire is estimated to be an order
of magnitude more diverse.
The first hurdle we must overcome — to achieve the
phenotyping required for personalized therapy — is
characterization of the innate arm’s molecular signaling
patterns and the T-cell receptor. “Reactive lymphadenopathy”
will go the way of the dodo. Instead, embrace
this type of lab report: “Activated, proliferating cytotoxic
T-lymphocytes present. Sequencing of these T-cells reveal
a T-cell receptor with specificity for an unknown target.” It
will be so, until our knowledge is more refined.
Once we have completed our measuring, we can begin
predicting. Virtual simulations of biology (“in silico”) will
continue marching endlessly forward towards the asymptotic
perfect prediction. Many philosophers will give up
on the age-old conundrum of whether free will exists.
(Sadly, it does not exist beyond our own perception.) The
unknown, activated T-cells in your pathology sample will
be submitted to the clinical bioinformatist to have its receptor
topology predicted by AlphaFold, bellowing in the
server for his next puzzle.2 In silico evaluation will determine
the T-cell receptor’s target-based affinity matching
against all known and unknown proteins. Mind you, technically
this isn’t futuristic. All these tools exist; the future
simply holds application and refinement (Do reach out if
interested).
In 75 years, we will have the capacity to measure the
vast data contained in blood using multi-sensory physicochemical
properties, far beyond the two-dimensions
of mass-spectrometry (mass and quantity). Imagine "The
Array" (“The Matrix” was taken) where each constituent
is evaluated by its response to a distribution of nondestructive
wavelengths, allowing for unique fingerprinting.
This vast expanse of data would eclipse our working
memory by thousands of orders of magnitude. The machine
would interpret and draw us a cartoon of the immune
system with helpful red and bold lines at upregulated
pathways leading us to the most proximal source of
inflammation.
There will be no serotype-phenotype discordance, and
diagnostic labels will be redefined using serotype-first approaches.
Lines between specialties will blur when your
rheumatoid patient has a profound signal for peripheral
ulcerative keratitis or interstitial lung disease.
We will become well-read in genotypes, and their
non-coding grammar. Gene editing will continue to boom-and-bust. Its current boom-cycle will bust once off-target
effects are measured. The careful will advocate for reversible,
non-curative treatments that consistently work and,
crucially, do not spill over into gametes and forever perpetuate.
There will be a push for ex-vivo genetic editing of
cells, à la CAR-T cells.
The world will be starkly different in 75 years. Our
power to dispel disease will grow dramatically, paralleling
the existential threat we pose to ourselves. We have yet to
shed our beast.
Reza Mirza, MD, FRCPC
Clinical associate, Sunnybrook Hospital
Toronto, Ontario
Clinician-Investigator Program, McMaster University
Hamilton, Ontario
References:
1. Miyagawa I, Nakayamada S, Nakano K, et al. Precision medicine using different biological DMARDs
based on characteristic phenotypes of peripheral T helper cells in psoriatic arthritis. Rheumatology.
2019; 58(2):336-344. doi:10.1093/rheumatology/key069.
2. Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold.
Nature. 2021; 596(7873):583-589. doi:10.1038/s41586-021-03819-2.
|
