Artificial intelligence is beginning to reshape research-based medicine. The shift is already visible in individual patient care, particularly in complex and advanced disease.

A few days ago, I watched a remarkable podcast featuring Sid Sijbrandij, who was diagnosed with a rare spinal chordoma. Faced with limited treatment options and a guarded prognosis, he used artificial intelligence tools to analyse published research and identify possible treatment strategies. These insights were then discussed with his treating doctors, who incorporated them into a medically supervised plan that included conventional therapies and highly customised gene therapies. He has now been in remission for over a year. The podcast is worth watching, because it offers a glimpse into how the practice of medicine may evolve in the near future.

Sid is not an isolated example. Doctors at Massachusetts General Hospital recently reported dramatic radiologic shrinkage of aggressive brain tumours after administering a personalised CAR T cell therapy called CARv3 TEAM E, a sophisticated evolution of traditional CAR T therapy. Some patients showed near complete responses within days. Several later relapsed, but the proof of short term efficacy of highly individualised immune therapies was unmistakable.

Another example comes from advanced esophageal cancer. A patient with metastatic disease and limited treatment options underwent tumour sequencing. A personalised mRNA neoantigen vaccine was designed using computational prediction tools and was combined with PD 1 immunotherapy. The patient achieved partial remission with disease control lasting more than one year. The therapy was engineered directly from the tumour biology. In early phase melanoma trials using the personalised vaccine platform EVX 01, designed using AI assisted neoantigen prediction, a majority of patients demonstrated tumour responses when the vaccine was combined with checkpoint inhibitors. Some achieved complete remission and several responses were durable over two years of follow up. A 70 year old man with metastatic gastric cancer progressed despite multiple chemotherapy regimens. He then received personalised peptide vaccines designed using a digital intelligence system along with ongoing conventional treatment. He was expected to survive less than six months but lived for fifteen months and demonstrated a measurable tumour specific immune response.

The key contribution of AI in these approaches is that it helps identify which mutations are immunologically meaningful and therefore actionable. There are also reports outside oncology. In one case, a patient with years of unexplained multisystem illness saw multiple specialists without a diagnosis. An AI system analysed the clinical history and medical records and suggested a unifying diagnosis that had previously been missed. The diagnosis was subsequently confirmed by treating doctors. The AI did not invent new knowledge. It connected existing dots more effectively.

These developments represent a shift in medical thinking. The central question is changing from ‘what is the best treatment for this disease’ to ‘what is the best treatment for this patient‘. The answer is increasingly coming from artificial intelligence, genomics, and deep individualised data analysis. This may be described as Science based AI guided Medicine, or SAM.

For the past century, randomized controlled trials have formed the backbone of evidence based medicine. These trials determine whether a treatment works on average in a defined population. By carefully controlling variables and minimising confounders, they generate reliable comparative data. Patients with multiple comorbidities are often excluded, interventions are standardised, and control and cohort groups are made as uniform as possible to reduce bias. Real world patients, however, rarely resemble trial populations. They may have interacting diseases, unique genetic profiles, and unpredictable treatment responses. A drug that works in the average patient may not work in a particular individual. Randomized trials answer population level questions. They do not always resolve individual level uncertainty. SAM does not replace randomized trials. It extends their findings. Instead of relying solely on population averages, clinicians can analyse tumour mutations, gene expression, prior treatment response, longitudinal laboratory trends, imaging, and pharmacologic interactions. Therapy can then be adapted to the biology of the individual patient. Treatment therefore evolves dynamically based on response, and is integrated with multiple data layers that predict outcomes more reliably than a one-size-fits-all approach.

This is already happening. AI’s real strength lies in integrating clinical history, imaging, laboratory data, genomics, and research literature into coherent science based treatment suggestions. It can identify existing drugs that may be repurposed for specific biological contexts. This is particularly useful in complex diseases and in patients who have exhausted standard options.

There are, of course, important caveats. For common diseases, conventional evidence-based treatments will continue to dominate. Access to AI tools is not a licence for patients to self diagnose or self treat without medical supervision. Individualised treatment often requires more investigations, which increases cost. Custom built therapies remain expensive and are currently used mainly in advanced or refractory disease. However, the economics are changing rapidly. Genome sequencing costs have fallen dramatically. mRNA and vector platforms have scaled globally. Access to research literature and computational tools is expanding. For rare diseases, where conventional research is often limited by small patient numbers, SAM may offer the most promising path forward.

As the business opportunity becomes clearer, companies are developing platforms that allow rapid customisation of treatment strategies. Indian companies such as Mapmygenome and MedGenome are already offering comprehensive genetic testing at increasingly competitive prices. Personalised cancer therapies for hematopoietic, colorectal, and skin malignancies are being explored in India, largely within clinical trials. Products such as APCEDEN and NEXCAR are being evaluated for advanced malignancies in carefully selected patients.

The Indian government recognised this opportunity relatively early. In 2019, the ICMR issued guidelines for gene therapy research. The Gene Therapy Advisory and Evaluation Committee was constituted by the Department of Health Research under the Ministry of Health and Family Welfare of the Government of India. A Centre for Advanced Genomics and Precision Medicine was established at AIIMS Jammu in 2025. Researchers at CSIR-IGIB have also developed novel gene editing approaches that have improved on the standard CRISPR-based gene editing technology and overcome patent hurdles. CSIR-IGIB licensed Indian biotechnology giant Meril for using its technology to develop better diagnostics for tuberculosis and gene therapies. In November 2025, CSIR-IGIB also entered into an agreement with Serum Institute India to license India’s first gene therapy for sickle cell anemia, named BIRSA 101, to bring this treatment to the masses at an affordable cost. Sickle cell anaemia is endemic in large segments of India’s tribal population. The development of gene therapy for sickle cell disease is part of the government’s National Sickle Cell Eradication Mission launched in 2023 for elimination of the disease by 2047. Clinical trials with participants from Madhya Pradesh, Chhattisgarh, and Jharkhand are set to begin this year in a collaborative effort of CSIR-IGIB, SII, and All India Institute of Medical Sciences (AIIMS), Delhi. SII is also one of the collaborators of CSIR-IGIB for thalassemia gene therapy development which is being done in parallel with the Sickle Cell gene therapy program.

Gene therapy is available at private centers in India but is expensive at 25 to 50 lakhs per treatment though only a fraction of the 25 crore cost in the USA. Five Indian patients of Haemophilia A received gene therapy at a trial conducted by the Christian Medical College, Vellore in 2024. None of them had any bleeding episode during the 14 months of follow-up, suggesting a cure. Aided by grants from the India Alliance for Rare Diseases, Dr Mohankumar Murugesan at CMC Vellore is also developing gene-based therapy for thalassemia and Dr Jayandharan Giridhara Rao at the IIT-Kanpur has developed a potential genetic treatment for Duchenne Muscular Dystrophy using an innovative adeno-associated virus vector. Dr Rao’s therapy has now entered pre-clinical trials and he posted recently on his LinkedIn that the results were very encouraging. Gene therapy for DMD is already available and USFDA-approved under the brand name of ELEVIDYS, but its use is limited due to its toxicity. If Dr Murugesan and Dr Rao’s goals of bringing these treatments to fruition are successful, they can potentially benefit close to a million children and adults who suffer from these incurable diseases in our country. Apart from this, biotech startups in India such as Cellogen Therapeutics have also brought gene therapy for thalassemia major and sickle cell anemia to pre-clinical trial. It is estimated that the treatment would cost about 40 lakh rupees in India compared to about Rs.30 crores in Western countries. Cellogen was also the collaborator with CMC Vellore for their groundbreaking CAR-T cell therapy trials in 2024.

The advanced CAR-T cell therapy for various types of cancers is now freely available in major metro cities at a fraction of the cost in the West. At a starting price of one lakh rupees, it is upto 400 times cheaper than its cost of USD 400,000 in the United States. Indigenous CAR-T therapy was developed by a collaborative effort of the Tata Memorial Hospital and IIT-Bombay headed by Dr Alka Dwivedi, who trained at the National Institute of Health to refine the techniques and overcome patent hurdles. The end-product was put through clinical trials. US researchers noted that outcomes of the Indian CAR-T modification were on par with their own, but without some of the severe side-effects seen in the US version of CAR-T cell therapy. President Draupadi Murmu formally launched the therapy at the Tata Memorial Hospital at a cost 90% less than that in the USA.

Far from replacing clinicians, SAM is likely to re-empower doctors who over the decades have been effectively reduced to implementing guidelines made by researchers with little freedom to customise treatments to the individual patient. As they gain access to unprecedented open-source research that combines with individual patient factors to give tailor-made solutions, the responsibilities of doctors will take on a whole new role. We are entering a new era where patients at the end of conventional therapies for their life-threatening condition are making completely rational and scientific decisions under medical supervision, and building cures for their own diseases in collaboration with research labs. Randomized trials will continue to define what generally works. AI guided precision approaches will help determine what works best for a particular patient. Doctors will also need to reinvent themselves to keep up with the changing times. Their depth of knowledge and the ability to use AI tools to be better informed and deliver superior outcomes for their patients will soon be put to the test.

References

1. Haendel MA, Chute CG, Robinson PN. Classification, ontology, and precision medicine. N Engl J Med. 2018;379(15):1452–1462.
2. Zhang C, et al. Personalized mRNA neoantigen vaccine combined with PD-1 inhibitor in advanced esophageal squamous cell carcinoma: a case report. Front Immunol. 2024;15:1396843.
3. Ott PA, Hu Z, Keskin DB, et al. An immunogenic personal neoantigen vaccine for patients with melanoma. Nature. 2017;547:217–221. (and subsequent AI-guided neoantigen vaccine studies including EVX-01 early-phase trials)
4. Keskin DB, et al. Neoantigen vaccine generates intratumoral T cell responses in metastatic cancer. Nature. 2019;565:234–239. (conceptual basis; includes computational neoantigen selection; supplemented by VERDI-based case reports in metastatic gastric cancer)
5. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25:44–56.
6. Press Release Page | Press Information Bureau
7. The role of data revolution in rare disease research
8. guidelines_GTP.pdf
9. Indian scientists build breakthrough gene-editor – The Hindu
10. Birsa-101: India’s Path Towards Affordable CRISPR Therapies
11. India’s First Homegrown CAR T-Cell Therapy – NCI
12. India launches gene therapy for cancer treatment | DD India News Hour
13. ImmunoACT | CAR-T Cell Therapy & Cancer Treatment in India
14. Press Release Page | Press Information Bureau
15. About Cellogen Therapeutics |Innovators in Cell & Gene Therapy