CAR-T cell therapy involves the genetic engineering of a patient’s T cells to introduce a chimeric receptor that will redirect the T cell response specifically toward malignant cells (1). CAR-T cell therapy has exhibited remarkable success in certain haematological malignancies, where interaction between the CAR-T cells and cancer cells occurs easily in the blood vessels. Its successful application in solid tumours, however, still faces barriers resulting from reduced CAR-T cell trafficking in, and the immunosuppressing effect of, the tumour microenvironment (1).

There is a need for supporting modalities to investigate the homing of CAR-T cells within the tumour, including techniques like CAR-T cell imaging (2, 3). Photoacoustic imaging (PAI), an emerging imaging modality, offers the potential for 3D imaging at depths up to several centimetres for CAR-T cell tracking in vivo, with spatial resolutions of a few hundred microns.

PAI utilizes the photoacoustic effect, where the absorption of optical energy generates broadband acoustic waves via thermoelastic expansion. In PAI, light absorbers serve as the source of contrast, and in biological samples, intrinsic PAI contrast can be obtained from endogenous chromophores such as haemoglobin or melanin. Most cells lack the chromophores that would make them visible to PAI, making it necessary to label them for tracking (e.g., using a reporter gene that translates into a chromophore). Recently, reporter genes derived from bacterial phytochromes have provided a reversibly photo-switchable PAI signal, the temporal signature of which was used to enhance cell detectability in PAI, although not in CAR-T cells (4-6).

This study assesses the potential of photoacoustic imaging to support CAR-T cell therapy for solid tumours through the utilization of reversibly switchable photoacoustic reporter genes.

Initial experiments examining the photoacoustic characteristics of two reversibly photoswitchable proteins, ReBphP and DrBphP, using the multi-spectral optoacoustic tomography (MSOT) InVision-256™ PAI platform (iThera Medical GmBH, Germany) in bacteria and in a mammalian cell line, have corroborated findings reported in existing literature (5). Our data confirmed that both proteins exhibit a distinct decay in photoacoustic signal explained by the photoswitching rate which can be used to provide good discrimination against background signals through temporal unmixing (5). We have tested the detection limit, and linearity with cell concentration, of various simple temporal photoswitching features measured with our MSOT equipment using Jurkat, a T-cell leukaemia‑derived cell line, expressing ReBphP in agar-intralipid phantoms. This demonstrated the detection of 6250 ReBphP Jurkat cells/µL. The sensitivity of our imaging platform could be further enhanced by employing machine-learning based methods to combine multiple photoswitching features and other image features (5). In addition, we have proven detection of ReBphP positive Jurkat cells injected intratumorally in subcutaneous tumours in C57BL/6 mice. These findings pave the way for future studies to assess in vivo tracking of anti-mesothelin CAR-T cells expressing the photoacoustic reporters, which will be evaluated in a lung cancer subcutaneous tumour model in immunodeficient mice.