In the months before the formation of Bangladesh in 1971, millions of refugees fled to India to escape mass killings. Among them were four-year-old Pradip Majumder and his family, who left Kushtia district for one of the many refugee camps in West Bengal.

Sitting in his office in Bengaluru’s Electronic city, Majumder is a long way from that refugee camp in Hemnagar, but it’s an experience that has stayed with him and shaped him on his journey through All India Institute Of Medical Sciences, Harvard University, and the founding of cancer tech startup Mitra Biotech. “Everybody starts from zero, but I started from negative," he says with a smile.

Majumder has vivid memories of bombs exploding as his family trekked 100km to reach Hemnagar. Growing up, school came second as everyone worked to keep the family afloat. His older sister became a teacher and helped him through high school and college. “Each stage was a triumph. Every time I thought, ‘This is the last; let me do my best,’" he says.


The idea for Mitra Biotech came to Majumder while shuttling between Harvard and Merck. On the faculty at Harvard Medical School’s Dana-Farber Cancer Institute, he would shadow clinical fellows to observe their handling of patients. Many patients wouldn’t return, so he asked where they were. The answer was that the patients moved into hospices or had died. Pharma companies weren’t creating the right drugs, they explained.

Two blocks from the cancer institute, within the Harvard Medical School complex, Merck had set up a lab. Majumder offered to develop drugs better targeted at cancer. He built a team for Merck’s first initiative in oncology and created two drugs, but one “failed miserably". When he asked why, Merck told him doctors hadn’t recruited the right patients for his drug.

Majumder realized that matching drugs and patients had to be more personalized.


For 50 years, drugs have been tested on cancer cell lines taken from different tumours, cultured and stored. A drug has a 20-30% chance of reducing a tumour. Cancer cell lines in the lab are different from a tumour in the body. Immune cells, connective tissues and blood vessels around a tumour play a role in its growth. Majumder’s idea was to recreate that micro-environment in the lab to test which drug would work for a specific patient.

Harvard liked the idea but wanted him to get a grant. That was hard in the post-2008 business environment. But his friend Mallikarjun Sundaram, adjunct faculty member at Massachusetts Institute of Technology and co-founder of Momenta Pharmaceuticals, was interested. They co-founded Mitra Biotech in 2010 with half a million dollars in angel funding.

The first hurdle was to get live tumour samples to develop their technology and validate it. Biological materials from patients are expensive in the US, where companies and academic labs want them. So, Mitra Biotech moved to Bengaluru.

Early funding came from Accel, Karnataka Information Technology Venture Capital Fund (Kitven) for biotech startups, and India Innovation Fund (IIF). Kitven and IIF got an exit in 2016 when Sequoia Capital and Sands Capital co-led a series B round of $27.4 million. Last year, Northpond Ventures from Maryland, led a $40 million series C round. Northpond’s specialization in life sciences will help Mitra Biotech in its last lap to go commercial.

Starting up in India has advantages. “Getting access to doctors, patients and data is probably easier and more cost-effective here than in the US," says Anjana Sasidharan, principal at Sequoia Capital.

The difficulty was that Mitra was a startup with a new concept in an immature ecosystem like India. In the US, the relationship between oncologists, researchers and organizations is well established. In the initial days, biopsy samples trickled into its lab. Over time, Mitra has set up a state-of-the-art lab at a cost of $5 million and now gets 1,200 samples a month to test drugs and understand the biology of a tumour’s micro-environment.

At present, there’s no scientific basis for a doctor to choose a drug for a cancer. After three months, if one drug doesn’t work, another is tried. “As a population, every approved drug works, but as an individual you don’t know which will work," he says. A drug that works for 30% of patients may seem great, but can be devastating for 70%.

Mitra’s product, CANscript, can tell for eight types of cancer whether a particular patient will respond to a drug. A one-centimetre biopsy sample is split into 30 parts to test six to 10 drugs. Each part is immersed in a “cocktail" that mimics the stroma surrounding a tumour. Blood comes with the sample to “feed" the tumour in the lab just as it would in the body. Traditional cancer cell cultures, in contrast, use bovine serum.

The live sample is cultured and tested for three days. Data from multiple biomarkers is run through a machine learning system. What emerges is a report for an oncologist stating which drug works and which doesn’t for his patient.


Mitra is collating data to prove the efficacy of CANscript before releasing it in the market. Majumder says the plan is to sell it to hospitals and oncologists. With the maturing of its research work in Bengaluru, the company moved its headquarters back to Boston under the name Mitra RxDx. The first rollout will be in the US.

“The US is the gold standard," he says. “That’s not true, but that’s the mindset in healthcare. In India, they ask: ‘Woh bahar use hota hai kya (is it used abroad)?’"

The next step is to embed mini-labs in major hospitals across the US. Mitra recently entered a partnership with Brigham and Women’s Hospital in Boston to use CANscript for testing the use of viruses to treat brain tumours. Pharmaceutical companies are also using CANscript to test drugs that have not yet been released.

Majumder has had many personal encounters with patients who tried the product. For one doctor with breast cancer, the lab found that a drug not recommended for breast cancer worked. “She decided to try it," he says. “I got Diwali and New Year greetings from her for four years. The last message came two years ago." Majumder is reluctant to call, in case he finds out she is no more. Most patients who try a product yet to be launched have late stage cancer. He looks forward to the day when the diagnostic aid will be available in hospitals for patients to choose the right drug at an early stage.