https://www.facebook.com/932945906566207/

Rheumatology with Dr. Aravind Palraj, Chennai (2026)

$🚨 CAR-T JUST ENTERED RHEUMATOID ARTHRITIS.6 patients.Severe refractory ACPA+ RA.One infusion of CD19 CAR-T.At 36–52 week...$

31/08/2026

🚨 CAR-T JUST ENTERED RHEUMATOID ARTHRITIS.

6 patients.
Severe refractory ACPA+ RA.
One infusion of CD19 CAR-T.

At 36–52 weeks:

👉 All 6 improved
👉 3/6 achieved DAS28-CRP remission
👉 3/6 achieved ACR70

But the biology is even more interesting 👇🧵

1/

The therapy was mivocabtagene autoleucel (miv-cel) — an autologous, fully human CD19 CAR-T product.

Patients had severe RA despite multiple previous therapies.

All DMARDs were stopped before treatment.

They first received standard lymphodepletion.

2/ What happened clinically?

At latest follow-up:

📉 Median DAS28-CRP reduction: 34%

And:

3 of 6 patients
achieved BOTH:

• DAS28-CRP remission
• ACR70 response

Despite cessation of immunosuppressive treatment.

3/ But this wasn’t simply about swollen joints

CAR-T produced profound CD19+ B-cell depletion—

not only in blood…

but also in tissue.

That matters because autoreactive B cells can persist in tissue niches despite conventional B-cell-directed therapy.

4/ Then something fascinating happened to autoantibodies

Seroconversion occurred in:

🧬 4/6 for ACPA against mutated citrullinated vimentin

and

🧬 5/6 for rheumatoid-factor IgM.

That raises the bigger question:

Are we suppressing autoimmunity—or actually resetting it?

5/ Why CAR-T is different conceptually

Rituximab targets CD20+ B cells.

CD19-directed CAR-T can achieve extremely deep B-cell depletion across compartments.

Afterward, the B-cell compartment can repopulate.

The hope:

pathogenic immune memory ↓
healthier B-cell repertoire returns

An “immune reset.”

Still a hypothesis—not yet proof of cure.

6/ Now the safety reality

All 6/6 developed cytokine-release syndrome.

But it was limited to:

✅ Grade 1–2 CRS

There were:

✅ No ICANS events
✅ No serious adverse events

One patient had grade-3 transaminase elevation, which resolved without sequelae.

7/ Before anyone says “CAR-T cures RA”…

Absolutely not.

This was:

• 6 patients
• Phase 1
• Nonrandomized
• No control group
• Highly selected refractory ACPA+ RA
• Follow-up only 36–52 weeks
• Lymphodepletion required

The trial was primarily designed to evaluate safety, not prove efficacy.

8/ What makes this study important?

RA has traditionally been treated by continuously suppressing pathways:

TNF
IL-6
JAK
T-cell costimulation
B cells…

CAR-T introduces a radically different idea:

Instead of chronic suppression, can we deeply reset the autoimmune B-cell compartment?

That is the experiment.

9/ The take-home

CAR-T is nowhere near routine RA treatment.

But in severe refractory disease, these first prospective data are difficult to ignore:

1 infusion
→ deep B-cell depletion
→ autoantibodies fell
→ clinical improvement
→ 3/6 reached remission + ACR70

Now we need phase 2.

📌 Albach FN, et al. Nature Medicine. 2026.
DOI: 10.1038/s41591-026-04603-3

31/08/2026

🚨 PR3-ANCA positivity does NOT always mean GPA.

Infective endocarditis can be PR3-ANCA positive — with purpura, hematuria, AKI, even crescentic GN.

Before immunosuppression, ask:
Could this be infection?

A positive antibody ≠ a diagnosis.

31/08/2026

🚨 A high CRP in SLE does NOT automatically mean infection.

Infection must be excluded — but active serositis or arthritis can also raise CRP.

The trap: treating the lab instead of the phenotype.

30/08/2026

AI is NOT ChatGPT.

And ChatGPT is not “all of AI.”

If you work in medicine, you need to understand 5 terms:

AI
Machine Learning
Deep Learning
Generative AI
LLMs

Here’s the 3-minute explanation every doctor should know.

AI ROUNDS #1 🧵

2/ First: Artificial Intelligence — AI

AI is the umbrella term.

It refers broadly to computer systems designed to perform tasks involving capabilities such as:

• recognition
• prediction
• language processing
• decision support

AI ≠ consciousness.
AI ≠ a robot doctor.

Think of AI as the large umbrella.

3/ Next: Machine Learning — ML

Traditional programming:

Rules + Data → Answer

Machine learning:

Data + Examples → Model → Prediction

Instead of explicitly programming every rule, an algorithm learns statistical patterns from data.

That distinction is fundamental.

4/ Then comes Deep Learning — DL

Deep learning is a subset of machine learning built using multilayer neural networks.

It is particularly powerful for complex data such as:

• Images
• Text
• Audio
• Signals

That is why deep learning became so important in medical imaging and biomedical AI.

5/ Now the term everyone hears:

Generative AI

Many traditional AI systems primarily classify or predict:

“What is this?”

“How likely is this outcome?”

Generative AI can generate new output:

“Write this.”
“Summarize this.”
“Create this image.”
“Explain this.”

That is a major shift.

6/ And what exactly is an LLM?

LLM = Large Language Model

An LLM is trained on enormous amounts of text to learn statistical relationships in language.

Given a sequence of tokens, it estimates what should come next.

Repeated again and again…

→ paragraphs
→ explanations
→ summaries
→ conversations

emerge.

7/ This explains something extremely important:

An LLM can produce an answer that sounds confident, fluent and medically sophisticated…

…and still be wrong.

Fluency ≠ factual accuracy.

This is one of the most important principles for doctors using generative AI.

8/ AI in medicine is also MUCH bigger than chatbots.

AI-enabled systems are already used or being developed for areas including:

• Medical imaging
• ECG and physiologic signal analysis
• Clinical prediction
• Workflow automation
• Monitoring
• Natural-language processing

So:

ChatGPT is one application of AI — not the definition of AI.

9/ Why should rheumatologists care?

Rheumatology produces unusually diverse data:

Symptoms

* examination
* serology
* imaging
* disease activity
* longitudinal records
* patient-reported outcomes
* omics

That makes rheumatology an important field for AI research.

Applications are being explored in imaging, disease classification, outcome prediction, treatment-response prediction, digital monitoring and LLM-assisted workflows.

10/ But the safest mindset is:

AI should augment clinical reasoning — not bypass it.

Before trusting an AI output, ask:

1. What data produced this?
2. What exactly was the model designed to do?
3. Was it properly validated?
4. Does it apply to this patient?
5. Can I independently verify the answer?
6. What is the consequence if it is wrong?

11/ If doctors are going to use AI, we should understand it — not just use it.

That is what AI ROUNDS will do.

One thread every week.
From first principles → practical clinical use.

Next:
How does an LLM actually generate an answer?

Tokens. Transformers. Context windows. Hallucinations.

Without the computer-science headache.

AI ROUNDS #1

Dr. Aravind Palraj

29/08/2026

🚨 Methotrexate does NOT appear to increase the risk of RA-associated ILD.

That old fear needs nuance.

MTX can rarely cause acute pneumonitis - but that is not the same as chronic RA-ILD.

Recent systematic reviews show no increased risk of developing or progressing RA-ILD with MTX. (PubMed)

MTX pneumonitis ≠ RA-ILD.

A distinction that can completely change treatment decisions.

26/08/2026

🚨 A WEIGHT-LOSS DRUG JUST IMPROVED PsA OUTCOMES—ON TOP OF A BIOLOGIC.

And this wasn’t an observational study.

A 2026 randomized trial tested:

Ixekizumab + tirzepatide
vs
Ixekizumab alone

in active psoriatic arthritis.

The results deserve attention 👇🧵

1/

TOGETHER-PsA is a phase 3b, randomized, multicenter, open-label trial.

271 patients with active PsA were randomized.

They had either:

• Obesity: BMI ≥30
OR
• BMI 27–

25/08/2026

A historic shift in IgG4-related disease.

In 2025, inebilizumab became the first FDA-approved treatment for adults with IgG4-RD.

In the MITIGATE trial:

• Disease flare: 10% vs 60% with placebo
• 87% relative reduction in flare risk
• HR 0.13; P

25/08/2026

🚨 “IT IMPROVED WITH STEROIDS” IS NOT A DIAGNOSIS

A common reasoning trap in rheumatology:

“The symptoms responded to prednisolone, so it must be autoimmune.”

Not necessarily.

Here are 7 situations where steroid response can mislead you 👇🧵

1/ Crystal arthritis

Gout and CPPD often respond rapidly to glucocorticoids.

So a dramatic response does not distinguish crystal arthritis from autoimmune inflammatory arthritis.

👉 Aspirate the joint when indicated.

2/ Polymyalgia rheumatica

Rapid improvement with glucocorticoids is characteristic of PMR—

but it is not diagnostic by itself.

PMR mimics include:

• Elderly-onset RA
• Infection
• Malignancy
• Mechanical shoulder disorders

Phenotype + exclusions matter.

3/ Malignancy

Some malignancy-associated inflammatory symptoms may improve after glucocorticoids.

And in certain lymphomas, steroids can modify clinical or pathological findings.

🚨 If unexplained lymphadenopathy or another strong malignancy signal is present, obtain appropriate diagnostic tissue before steroids whenever clinically feasible.

4/ Infection

Glucocorticoids suppress the host inflammatory response.

Fever, pain and inflammatory markers may improve—

while the underlying infection persists.

Think particularly about:

• Tuberculosis
• Septic arthritis
• Endocarditis
• Deep-seated infection

Symptomatic improvement ≠ infection excluded.

5/ Mechanical musculoskeletal disorders

Pain from conditions such as:

• Rotator-cuff disease
• Bursitis
• Radicular pain
• Osteoarthritis

may also improve after local or systemic glucocorticoid exposure.

So:

Pain relief ≠ proof of inflammatory arthritis.

6/ Granulomatous/inflammatory disorders

Sarcoidosis and several other inflammatory conditions are glucocorticoid-responsive.

But steroid responsiveness does not identify the underlying disease.

Always consider the relevant differential:

👉 Infection
👉 Malignancy
👉 Drug-related disease
👉 Immune-mediated inflammation

7/ The “diagnostic steroid trial” trap

Giving glucocorticoids before establishing the phenotype may modify:

• Fever
• Synovitis
• Rash
• CRP/ESR
• Imaging findings
• Histopathologic findings

Sometimes treatment is urgent and cannot wait.

But when the patient is stable:

document the phenotype and obtain key investigations first.

The clinical rule:

Response to glucocorticoids is therapeutically useful—but diagnostically nonspecific.

It does NOT prove:

❌ Rheumatoid arthritis
❌ PMR
❌ Vasculitis
❌ Connective-tissue disease
❌ Autoimmunity

Ask instead:

What is the phenotype?
What else can produce it?
What objective evidence supports my diagnosis?

Treat the disease, not the steroid response.

24/08/2026

🧵 ACUTE-PHASE REACTANTS: 8 rules every rheumatologist should know

CRP ↑
Ferritin ↑
Fibrinogen ↑
Albumin ↓

And ESR?

🚨 ESR is NOT an acute-phase reactant.

It is an indirect marker of inflammation.

High-yield pearls 👇

1/ What exactly is an acute-phase reactant?

An acute-phase reactant is a plasma protein whose concentration changes during inflammation.

Positive APRs ↑
• CRP
• Serum amyloid A
• Fibrinogen
• Ferritin
• Hepcidin
• Complement proteins

Negative APRs ↓
• Albumin
• Transferrin
• Transthyretin

2/ ESR is NOT an acute-phase reactant

ESR measures how rapidly RBCs sediment.

Inflammation → fibrinogen ↑
→ rouleaux formation ↑
→ ESR ↑

So ESR reflects the acute-phase response indirectly.

It is a test—not an acute-phase protein.

3/ CRP moves FAST

CRP production is driven predominantly by:

👉 IL-6 → hepatocyte → CRP

CRP rises within hours and peaks around 24–48 h.

Plasma half-life ≈ 19 hours.

Therefore CRP usually tracks changing inflammatory activity relatively quickly.

4/ ESR moves SLOW—and has many confounders

ESR can increase with:

↑ Age
↑ Anemia
↑ Pregnancy
↑ Fibrinogen
↑ Immunoglobulins

And decrease with:

↓ Polycythemia
↓ Abnormal RBC morphology

So:

High ESR ≠ automatically active inflammation.

5/ ESR ↑↑ + CRP relatively low? Think SLE

A classic rheumatology pattern.

Why?

👉 Type-I interferon biology can blunt CRP production.

Meanwhile:

Hypergammaglobulinemia + fibrinogen can push ESR upward.

This ESR–CRP discordance can itself be diagnostically useful.

6/ Ferritin is much more than an iron-storage protein

Ferritin is a positive acute-phase reactant.

In a sick rheumatology patient:

🔥 Persistent fever
📈 Rapid ferritin rise
📉 Platelets
📈 AST
📉 Fibrinogen
📈 Triglycerides

Think:

🚨 MAS / HLH

The trend often matters more than one cutoff.

7/ Fibrinogen gives an important paradox

Most inflammation:

Fibrinogen ↑

This also contributes to:

ESR ↑

But in severe MAS/HLH or DIC:

🚨 Fibrinogen may FALL

So:

Ferritin ↑ + platelets ↓ + fibrinogen ↓

should immediately raise concern for macrophage activation/coagulopathy.

8/ Positive vs negative acute-phase reactants matters clinically

Inflammation drives:

📈 CRP
📈 SAA
📈 Ferritin
📈 Fibrinogen
📈 Hepcidin

while:

📉 Albumin
📉 Transferrin

Hepcidin ↑ also explains an important rheumatology phenomenon:

👉 Iron sequestration → anemia of inflammation

Remember the distinction:

CRP = acute-phase reactant
Ferritin = acute-phase reactant
Fibrinogen = acute-phase reactant
SAA = acute-phase reactant
Albumin = negative acute-phase reactant

ESR = NOT an acute-phase reactant

It is an indirect measure of the acute-phase response.

That distinction is a favorite viva question.

23/08/2026

More than 30,000 people across platforms.

But what moves me most is not the number - it is the trust behind it.

In November 2024, I began sharing what I was learning in rheumatology, one post at a time. I never imagined that, 21 months later, this small effort would grow into such a wonderful community.

Every follow, message, question, correction, share, and word of encouragement has meant more to me than you may realize. Thank you for learning with me, supporting me, and giving my work a purpose beyond the classroom and clinic.

Teaching has always been my passion. I am still a student myself, and I hope I never stop being one.

I will continue to learn, simplify, share, and teach - with the same curiosity and sincerity with which I began.

From the bottom of my heart, thank you for being part of this journey. ❤️

— Dr. Aravind Palraj

Connect with me:

Facebook
facebook.com/share/1BdiUGPx…

LinkedIn
linkedin.com/in/aravind-pal…

X
x.com/rheumat_aravin…

Instagram
instagram.com/rheumat_aravin…

Threads
threads.com/…

Whatsapp channel
whatsapp.com/channel/0029Vb…

Rheumatology with Dr. Aravind Palraj

31/08/2026

31/08/2026

31/08/2026

30/08/2026

29/08/2026

26/08/2026

25/08/2026

25/08/2026

24/08/2026

23/08/2026

Address

Website

Alerts

Shortcuts

Share

Category

Listing takedown request