Skip to content

First-in-Human Trial: Implanted Microdevice for Testing Patient Sensitivity to Multiple Chemotherapeutic Agents in NSCLC

Key findings

  • This first-in-human, phase 1 trial evaluated intraoperative implantation and postoperative retrieval of a novel implantable microdevice (IMD) in five patients who had non-small cell lung cancer resected
  • The IMD, which measures 0.7 by 6.5 mm, simultaneously administers microdoses of up to 20 chemotherapeutic agents into confined regions in a tumor, allowing evaluation of how an individual patient's tumor cells will respond to each drug
  • 14 IMDs were loaded with 12 FDA-approved chemotherapeutic agents or combinations, then implanted during open thoracotomy, video-assisted thoracoscopic surgery or robotic surgery
  • One of the first devices implanted could not be retrieved, but no serious adverse effects were observed in any patient and no trace of any chemotherapeutic agent loaded onto the IMD was detectable in patient blood samples after surgery
  • This trial resulted in a safe protocol for IMD implantation and a reliable, efficient method for retrieval in pathology; future work is needed to determine the value of the device for testing tumor sensitivity to multiple drugs

Adjuvant therapy improves survival in certain patients with non-small cell lung cancer (NSCLC), but how to identify candidates and choose their optimal regimen is largely unknown. This is a particularly pressing issue now that targeted therapies are being investigated in early-stage NSCLC and advanced disease.

Yolonda L. Colson, MD, PhD, chief of the Division of Thoracic Surgery at Massachusetts General Hospital, Lillian L. Tsai, MD, formerly a fellow in that division, and colleagues are developing a solution. In the Annals of Surgery, they describe the first clinical trial of a novel implantable microdevice (IMD) that serves as a tiny platform for individualized high-throughput drug screening—directly within the native tumor.

Background on the IMD

The IMD, which measures 0.7 mm by 6.5 mm, contains 20 reservoirs that simultaneously administer chemotherapeutic agents into confined, non-overlapping regions in a tumor. The device is designed to evaluate how an individual patient's tumor cells respond to microdoses of up to 20 agents simultaneously.

The thinking is that by keeping the tumor microenvironment and native physiology intact, the true in vivo response can be observed. The IMD has been validated in small-animal models of several solid tumors.


As the first step in clinical translation, this phase 1 trial evaluated whether it is safe and feasible to place IMDs into tumors of patients undergoing resection of NSCLC and retrieve them by pathologists.

The study participants were five adults with an average age of 77. To facilitate experimentation with the technique, all had lung lesions >4 cm in size located within 0.5 cm from the periphery/visceral pleura. Both lobes and three operative approaches were represented: open, video-assisted thoracoscopic surgery, and robotic surgery.

Each IMD was loaded with 12 FDA-approved chemotherapeutic agents or combinations, some that are commonly used to treat NSCLC and others used as controls. The total dose was 2 ng (less than 1/100,000 of a systemic dose), released over a local region about 300 microns in diameter. Polyethylene glycol was also used as a control.


14 IMDs were implanted intraoperatively and remained within the tumor for the duration of surgery, 60 to 350 minutes. No device became dislodged or migrated during an operation. No trace of any chemotherapeutic agent loaded onto the IMD was detectable in patient blood samples after surgery.

13 devices (93%) were successfully retrieved. The first IMD placed during robotic surgery was lost to view during insertion but has remained in place without movement or any adverse effects. The IMD is made of a biocompatible plastic approved for use in permanent implantable medical devices, even though it is not intended for permanent placement.

After refinement of the technique, IMDs were reliably inserted and retrieved in all three types of procedures. No severe adverse reactions were observed.

Different tumors had markedly different sensitivities to the drugs the IMD administered.

Next Steps

In their next trial, the researchers plan to place the IMDs into tumors in a minimally invasive manner 24–48 hours preoperatively, under interventional radiology imaging guidance. That will allow more uniform drug exposure time and more time for the agents to work. The team anticipates being able to monitor for differences in tumor response to various agents.

The ultimate goal is to insert the IMDs preoperatively, allow the drugs to interact with tumor tissue, surgically resect the tumor, and analyze the specimen to determine a precise adjuvant therapy regimen for each patient.

Future applications of the IMD might also include testing new investigational agents without exposing patients to the toxicity of systemic dosing and/or delivering locoregional drug therapy.

Explore research in the Division of Thoracic Surgery

Refer a patient to the Division of Thoracic Surgery

Related topics


Up to 5% of the patients who underwent prolonged invasive mechanical ventilation during treatment for COVID-19 later presented with post-intubation tracheal stenosis. This article discusses causes, identifiers and treatment of tracheal stenosis.


Massachusetts General Hospital researchers created a fully automated deep learning system to quantify and characterize muscle and adipose tissue on CT scans at multiple thoracic vertebral levels. The system is as accurate as human analysts on routine CT scans, regardless of intravenous contrast.