Label-Free 3D Photoacoustic Imaging of Tumor Organoids for Volumetric Drug Screening

Abstract

As one of the most advanced in vitro drug screening platforms, tumor organoids require accurate pharmacosensitivity assessments to ensure reliable results. However, achieving accurate volume assessment of these 3D organoid models remains a challenge in traditional drug screening processes. Here, a label-free organoids photoacoustic imaging (LFOPI) system is introduced for high-resolution 3D imaging of tumor organoids. The capabilities of the LFOPI system are evaluated by monitoring structural transformations in melanoma organoids. The LFOPI system is further employed for volumetric drug screening of melanoma tumor organoids. Drug screening with cisplatin and temozolomide reveal that LFOPI-based organoid volume data correlates more strongly with viability trends (0.8627 and 0.9069) than traditional diameter-based methods (0.8190 and 0.7849). Additionally, immunotherapy drug screening demonstrates the capability of the system to precisely assess the 3D volumes of irregularly shaped organoids. This LFOPI system provides a promising method for label-free drug screening and personalized treatment.

Keywords: label free; photoacoustic imaging; precision medicine; tumor organoids; volumetric drug screening.

Figure 1

LFOPI system for non‐invasive assessment of volume changes in tumor organoids. A) The workflow of individualized precision treatment for tumor patients. B) The comparison of conventional drug screening and our LFOPI‐based volumetric drug screening. C) Schematic of the LFOPI system for 3D photoacoustic imaging of tumor organoids. D) Principles of volume calculation in the LFOPI system. A laser pulse generates ultrasound signals that form an A‐line. Linear scanning of the sample produces a B‐scan, while planar scanning results in a 3D C‐scan. Subsequent thresholding segments the sample, and volume is calculated based on pixel counts. E) Representative 3D photoacoustic images of melanoma organoids.

Figure 2

Spatiotemporal characterization of the LFOPI system. A) Illustration of the imaging mode in LFOPI. B) Calculation of lateral resolution of the system. C) Calculation of longitudinal resolution of the system. D) The 3D photoacoustic image of four fiber in agar. E) Maximum x‐projections of the 3D volume in D. F) A set of 3D reconstructions of two fibers at different z‐positions. G) Linear correlation plot comparing the measured z positions from LFOPI to the actual z positions.

Figure 3

3D label‐free photoacoustic imaging of melanoma organoids. A) Workflow of the 3D photoacoustic imaging process for melanoma organoids. B) 3D reconstructions of melanoma organoids by the LFOPI system. C,F) Bright field microscopy images of melanoma organoids before photoacoustic imaging. D) and G 3D photoacoustic images of two melanoma organoid samples. E,H) Series of z‐stack images showing the depth‐resolved imaging capability of the LFOPI system for group 1 and group 2, with normalization of photoacoustic amplitude from 0 to 1. Scale bars, 100 µm. I,L) The melanoma organoids surface profiles of group 1 and group 2, respectively. J,M) Volume histograms for samples a‐c and d‐f, respectively. K,N) Diameter histograms for samples a–c and d–f, respectively.

Figure 4

Volumetric chemotherapy drug screening with LFOPI. A) Workflow illustrating the process of 3D tumor organoid chemotherapy drug screening using LFOPI. B) Maximum z‐projections of the 3D volumes of organoids treated with cisplatin across four treatment groups: 0 µm (control), 1 µm, 3.25 µm, and 10 µm drug concentration groups. C) Maximum z‐projections of the 3D volumes of organoids treated with temozolomide across the same four treatment groups. The tumor organoids highlighted by the yellow boxes in the control and 1 µm groups exhibit similar 2D projected areas (35775 µm² and 36825 µm², respectively), yet their volumes differ markedly (3171375 µm³ and 2341875 µm³, respectively). D,E) Detailed 3D volume and cross‐sectional views (indicated by white dots) of the organoid groups treated with cisplatin and temozolomide, respectively. F,G) Sequential z‐position slices from the 3D reconstructions shown in (D,E), demonstrating depth‐resolved imaging capabilities. Scale bars, 100 µm. H) Dynamic volumetric responses across drug dosages. Data are presented as mean ± s.d. (n = 3). I) Cell viability of tumor organoids treatment with varying concentrations of cisplatin and temozolomide. Data are presented as mean ± s.d. (n = 3). J,K) Correlation between organoid volumes estimated by LFOPI and traditional methods with cell viability after treatment with cisplatin and temozolomide.

Figure 5

Volumetric immunotherapy drug screening using the LFOPI system. A) Workflow of 3D tumor organoids immunotherapy drug screening via LFOPI. B) Maximum z‐projections of 3D volume from control group. C) 3D visualization of the organoid from (B). D) Cross‐section of organoid shown in (C). E) Different z‐position slices through the organoid in (C). F) Maximum z‐projections of 3D volume from CD45⁺ T cell group. G) 3D visualization of the organoid from (F). H) Cross‐section of organoid shown in (G). I) Different z‐position slices through the organoid in (F). J) Bright field microscopy images of melanoma organoids co‐cultured with CD45⁺ T cells. K) Viability comparison between control and CD45⁺ T cell group. Data are presented as mean ± s.d. (n = 3). L) Volumetric comparison between control and CD45⁺ T cell group. Data are presented as mean ± s.d. (n = 3). M) Correlation between organoid volumes and viability in co‐cultured models. Scale bars in (E,I) are set at 50 µm.