The pharmaceutical industry faces a persistent crisis in analgesic development, characterized by a high attrition rate where drugs effective in rodents fail in clinical trials [1]. Why does this translational gap persist despite decades of pain research? Here, the underlying causes of these failures will be discussed, focusing on the fundamental biological divergences between rodent and human. Next, a new perspective will be presented to overcome these limitations: a paradigm shift from surrogate animal models toward human-centric organoid platforms.

Clinical failures in analgesic development are largely due to the heavy reliance on mouse models, which assumes evolutionary conservation of pain pathways: from the transduction of noxious stimuli in the periphery and the processing of pain signals in the spinal dorsal horn and the sensory cortices. However, the dorsal root ganglion (DRG), the primary starting point of the nociceptive circuit, exhibits fundamental biological divergence between species [2]. In mice, nociceptors are segregated into distinct, non-overlapping populations—specifically, the peptidergic (CGRP+) and non-peptidergic (IB4+/P2X3+) neurons. This clear separation supports a “labeled line” theory, where specific neurons are responsible for specific types of pain, allowing for precise, targeted drug development in rodent models. However, human transcriptomic and histological studies reveal a fundamentally different architecture. In human DRGs, these markers exhibit significant overlap; for instance, approximately 33% of human sensory neurons co-express CGRP and P2X3, a phenotype virtually absent in mice. Furthermore, key ion channels such as Nav1.8 and TRPV1 are distributed much more broadly across human DRG neurons than in rodents [3]. This “molecular blurring” can hinder analgesic development because targeting specific neurons in humans, which possess functional redundancy and alternative signaling pathways, often leads to translational failure.

Consequently, there is an urgent imperative to shift the preclinical screening paradigm away from surrogate animal models toward human-centric platforms. The convergence of human induced pluripotent/embryonic stem cell (hiP/ESC) technology with high-density microelectrode arrays (HD-MEAs) offers a transformative opportunity to bridge this gap. This approach allows us to recapitulate the “electrophysiological topology” of the human DRG in vitro. However, realizing the potential of these platforms requires overcoming significant bioengineering and computational challenges, from the stochasticity of stem cell differentiation to the complex signal processing required to distinguish nociception (from nociceptive C-fibers) from non-noxious stimuli (from non-nociceptive afferents).

A major challenge in utilizing stem cell-derived neurons is the “batch effect”—the inherent variability in differentiation efficiency between experiments [4]. Stem cell differentiation is stochastic; one batch may yield 80% nociceptors while another yields only 50%, creating noise that can mask drug effects. To utilize these cells for industrial-scale screening, rigorous computational correction is required. For example, instead of simply measuring raw firing rates, to normalize the drug response against both negative (vehicle) and positive (toxic) controls within the same batch would be beneficial. This statistical framework corrects for variations in cell density and baseline excitability, allowing for reproducible comparisons of drug potency across different differentiation runs.

Despite the molecular blurring observed across species, the fundamental electrophysiological topology of sensory neurons remains conserved. This conservation allows the use of high-density microelectrode arrays (HD-MEAs) to functionally distinguish signals from pain-sensing neurons, creating a robust drug screening metric.

Waveform Classification: Nociceptive C-fibers are unmyelinated neurons that express high levels of “slow” voltage-gated sodium channels (Nav1.8, Nav1.9) [5,6]. This results in a broader action potential compared to non-nociceptive A-fibers, which rely on fast channels (Nav1.1, Nav1.6). On an MEA, C-fibers are identifiable by their broad action potentials and specific sensitivity to repetitive stimulation. In contrast, A-fibers (touch/proprioception) are myelinated, fast-conducting neurons that exhibit narrow action potentials and are generally non-nociceptive.

Activity-Dependent Slowing (ADS): HD-MEAs allow for the tracking of spike propagation along an axon. A definitive functional phenotype of C-fibers is activity-dependent slowing (ADS) [7]. Upon repetitive stimulation, the conduction velocity of C-fibers progressively decreases due to the slow recovery of Nav1.8 channels, whereas A-fibers do not exhibit this degree of slowing. This phenomenon is tightly linked to sodium channel inactivation and serves as a robust functional biomarker for identifying nociceptors in mixed cultures. Under pathologic conditions (e.g., chemotherapy-induced damage), human C-fibers develop ectopic discharges (spontaneous firing) and network synchrony [8]. A valid screening platform aims to identify drugs that selectively suppress these pathological signatures in the C-fiber population without silencing the physiological activity of A-fibers.

While human DRG models offer superior molecular precision over mouse models, they lack the systemic complexity of a living organism. In in vivo neuropathic pain states, the DRG is not an isolated island; it is an active site of neuro-immune interaction [9]. Satellite Glial Cells (SGCs) envelop the neuronal soma and, following injury, can form gap junctions that amplify pain signals [10]. Furthermore, nerve injury recruits macrophages and T-cells that release pro-inflammatory cytokines (TNF-α, IL-1β), which are potent sensitizers of nociceptors [11,12]. Most current in vitro screening platforms consist of pure neuronal cultures, missing these critical extrinsic modulators. Consequently, a drug that functions by dampening neuro-inflammation or blocking glial-neuron cross-talk would be missed in a neuron-only screen. Future platforms must advance toward co-culture systems to better mimic the in vivo nociceptive environment.

The ultimate goal is to consistently recapitulate the complex in vivo nociceptive microenvironment using in vitro systems. Human DRG-based platforms capable of reproducing this complexity will bring about true innovation in analgesic development, significantly improving translational success rates and delivering effective treatments to patients.