Animal Models of Disc Degeneration Using Puncture Injury: A 20 Year Perspective

ABSTRACT

Background

Intervertebral disc (IVD) degeneration (IVDD) is a major cause of global disability. Three papers on puncture models of IVDD were published 20 years ago, transforming the application of preclinical animal models for pathophysiology and therapeutic screening studies.

Methods

Narrative review describing historic and current usage of preclinical puncture models of IVDD, documenting their introduction to induce slow, progressive IVDD and evolution to include many injury types broadly called “puncture models.” IVDD puncture models were reviewed for variability in species, needle gauge, puncture depth, IVD compartment, injectates, angle of puncture, motion of needle, and IVDD phenotype mimicked.

Results

IVD puncture models gained prominence following seminal 2005 publications describing needle puncture to induce slow, progressive IVDD for screening therapies. Specific details of puncture methods were described for controlling injury severity to induce IVDD phenotypes, including slow progressive IVDD, severe IVDD, chronic IVDD, disc herniation, and Modic‐like changes. Common measurements for characterizing IVDD were also described.

Conclusions

Surgically induced IVD puncture injury animal models have evolved over decades to include many variations simulating distinct clinical phenotypes of IVDD. To facilitate cross‐study comparisons, we recommend reporting a common set of injury features including needle gauge, puncture depth, injectates, puncture angle changes, needle motion, involvement of endplate and surrounding tissues, and phenotypes of IVDD mimicked. Surgically induced “outside‐in” puncture injury models are valuable tools to test specific hypotheses and screening therapies.

This review of surgically induced IVD puncture injury animal models presents an evolution over decades to include many variations simulating distinct clinical phenotypes of IVDD, and recommends reporting a common set of injury features to facilitate cross‐study comparison.

1. Introduction

Back pain is a leading cause of global disability, reducing healthy life expectancy and increasing disability‐adjusted life years [1]. Intervertebral disc (IVD) disorders, such as herniation and degeneration, contribute to both specific and nonspecific causes of disabling pain [2, 3, 4]. While IVD herniation can directly cause neuropathology [4, 5, 6, 7], IVD degeneration (IVDD) is often implicated in nonspecific back pain, where instability, stenosis, and chronic inflammation enhance nociception [2, 4, 8]. Human IVDD progresses over decades, and cell culture models fail to replicate immune and pain‐related responses [2, 3, 9, 10], necessitating preclinical animal models that accelerate IVDD to identify pathophysiological mechanisms and test therapeutic strategies.

Animal models simulating painful IVDD provide insight into complex interactions between degenerating IVDs, surrounding tissues, and the nervous system [11, 12, 13]. Many animal models simulate naturally occurring human IVDD conditions through biomechanical overload and instability models, genetic models, chemically induced models, surgically induced models, and hybrid techniques [14, 15, 16, 17, 18, 19, 20]. Puncture injury models typically involve needle insertion, scalpel laceration, or stab incision into the IVD to induce IVDD.

Twenty years ago in 2005, three papers were published applying syringe needle injuries that punctured the IVD to induce slowly progressive and reproducible models of IVDD [21, 22, 23]. These seminal studies established puncture models as the dominant surgically induced model of IVDD by describing methods for inducing IVD injury, characterizing the resulting pathological changes, and demonstrating the models' effectiveness in mimicking clinical IVDD in animal subjects. Notably, puncture injuries with syringes are more reproducible across labs than other surgically induced models. Syringe needle injuries also offer the advantage of enabling injections of substances to accelerate degeneration or promote repair. Concurrently, Carragee et al. demonstrated in clinical practice that discography procedures involving puncture and injection in healthy IVDs accelerated IVDD progression [24]. This highlighted that puncture injuries are not benign in humans and confirmed the translational relevance of animal puncture models. Due to their reproducibility, precision, and technical feasibility, puncture models have become the most widely used method for inducing IVDD in preclinical models, particularly in smaller species such as rabbits and rodents.

The widespread use of puncture models in IVDD research has resulted in variations in needle gauge, needle type, puncture depth, injectate, and other procedural factors that induce differing IVDD phenotypes. This review commemorates the 20th anniversary of seminal rabbit IVD puncture studies [21, 22, 23], and showcases the success of the puncture injury method in IVDD animal model research. It has three primary objectives: (1) to describe the evolution of surgically induced IVD puncture models, (2) to highlight the variability in “puncture” interventions, and (3) to propose standardized IVD puncture characteristics for improved reproducibility and comparison across studies.

This narrative review captured the diversity of puncture injury models in preclinical IVDD research over the last 20 years following the publication of the three seminal papers in 2005. Literature searches occurred from December 13, 2024 through March 17, 2025 using three databases (PubMed, Scopus, Embase) and search terms “intervertebral disc degeneration,” “intervertebral disc puncture,” “animal models,” “puncture injury,” “puncture model,” “annular puncture,” and “annulus fibrosus injury.” Studies were included if they described puncture injury methodological details including surgical approach, number of punctures, needle gauge and motion, puncture depth (including percent of IVD height where reported), and the IVDD phenotype modeled. Studies were excluded if they lacked details on puncture injury methods, or if no conclusion could not be inferred based on impact of injury severity and phenotype. All papers were downloaded in Endnote and initial screening was conducted independently by two authors (C.J. and J.J.H.), with final inclusion decisions made through discussion and consensus with the senior author (J.C.I.). Key characteristics of animal models and puncture models were extracted including species used, spinal levels injured, IVDD phenotype mimicked, IVD compartment injured, needle gauge (including percent of IVD height where reported), puncture depth (including percent of IVD diameter where reported), number of punctures and motion, puncture approach, injectate and dwell time (if used). Papers describing puncture models in commonly used animal species (rabbit, rat, mouse, sheep, goat, cow, pig, dogs) were also included. Papers were further included in summary tables to demonstrate the range of variability in puncture injury models across species, IVDD phenotypes, and injury methods. Based on the data synthesized, we propose standardized puncture injury parameters, incorporating species selection, injury reporting practices, induction methods, and outcome measures.

2. Historical Perspective on Surgically Induced Animal Models of IVDD

Surgically induced animal models have been instrumental in studying the pathophysiology of IVDD and evaluating potential therapies. Preclinical models are essential to identify cellular, biochemical, and biomechanical mechanisms underlying IVDD. As surgical techniques, medical technologies, and injury models advance, there is a growing emphasis on translational research providing translational tools to screen therapies and bridge the gap between animal studies and human clinical applications.

2.1. Early Models: Mechanical Injury as a Tool for Inducing Degeneration

Early IVDD models focused on mechanically induced spinal injury, targeting the IVD through surgical incision. In 1948, Key and colleagues introduced a controlled method of inducing IVD injury using a scalpel, marking a pivotal beginning in experimental IVDD research [25]. Smith et al. refined this technique in 1951 by creating surgical incisions in rabbit IVDs to replicate degenerative changes observed in humans [26]. This study demonstrated that mechanical disruption of the annulus fibrosus (AF) could lead to progressive degeneration, highlighting the critical role of annular injury in both animal models and clinical IVDD.

2.2. Complete and Partial AF Stab Injury Models

The total annular stab model, developed by Lipson and Muir, involved a large stab incision (2–3 mm wide, 5 mm deep) in two noncontiguous IVDs, creating a full‐thickness, or “complete” AF injury with nucleus pulposus (NP) depressurization and rapid IVDD [27]. This paper reviewed existing animal models of IVD herniation, provided methods for surgically inducing IVDD in rabbits, and showed that stab injuries caused progressive IVDD with NP fibrosis and loss of aggregating proteoglycans (Figure 1). However, the severity of this model exceeded human pathology, limiting its relevance for use in screening regenerative therapies [27]. In contrast, Osti et al. developed a model with incomplete AF rupture (without NP depressurization) to assess whether surgically induced superficial AF tears could induce degenerative changes in the IVD and surrounding structures in a sheep model [29]. In this study, a 5 mm wide and 5 mm deep incision was created to induce rim lesions through partial AF damage. This demonstrated that AF defects progress to whole IVDD, affecting all compartments [30]. Although degeneration in this stab model developed more gradually compared to complete AF ruptures, the surgical complexity and cost of large animal models limited their broad application for screening novel therapeutics for IVDD [22].

FIGURE 1.

Some important preclinical models of IVDD prior to puncture model development. (A) Lipson and Muir applied a ventral stab incision causing NP herniation and progressive IVDD with fibrous NP and loss of aggregating proteoglycans [27]. (B) Hampton et al. showed the complete AF defect causes fibrous scarring in canines with limited healing potential highlighting the biological treatments for IVD repair [28]. (C) The Osti and Fraser sheep injury model [29] as reviewed by Melrose et al. [30] shows spontaneous repair response in the outer AF (asterisk) after 3 months and lack of repair of the defect internally with (a) focal collagenous remodeling of outer AF spanning the original defect site (Masson Trichrome); (b) loss of anionic proteoglycan along annular defect, (toluidine blue/fast green stain). (c) Collagen reorganization around the AF defect (asterisk), (picrosirius red stain viewed under polarized light). (d) Polarized light view of segment (b) with similar detail to segment (c) [29, 30].

2.3. AF Injury and IVDD Progression

In 1989, Hampton and colleagues evaluated healing in a non‐chondrodystrophic dog model following surgically induced injuries of different sizes [28]. Small lesions, created using smaller needles (e.g., 22‐ to 26‐gauge), involved minor punctures causing limited disruption to the IVD. These led to limited healing, enabling NP leakage for chronic durations. Larger injuries, created with bigger needles (e.g., > 18‐gauge), resulted in more significant damage, with inward bulging of the AF and lamellar disorganization. While these larger injuries healed with fibrous tissue that partially limited NP leakage, they did not fully restore IVD structure and function [28]. Overall, AF defects were shown to exhibit self‐repair, yet microscopic defects persisted even after 3 months, highlighting the chronicity of IVDD and the IVD's limited capacity for repair and regeneration. Otani and coauthors applied an 18‐gauge needle puncture with intradiscal saline injection to induce herniation in dogs. Their findings showed that neuropathy could be induced by the NP without causing visible nerve root compression on MRI [31]. This model allowed for the direct study of neuropathy, with less emphasis on degenerative changes induced by the puncture injury. Together, these findings demonstrated the importance of understanding AF defects and highlighted the need for preclinical models capable of testing biological therapies to enhance IVD repair and address neuropathy.

2.4. Discography Procedures Induce IVDD

Carragee et al. showed that modern provocative discography techniques using small‐gauge (22‐ and 25‐gauge) needles and limited pressurization caused accelerated IVDD, herniation, loss of IVD height, and endplate (EP) changes compared to control [24, 32]. These degenerative changes were attributed to the structural disruption caused by needle puncture and injection, leading to NP depressurization and AF delamination. Inadequate healing responses can lead to chronic inflammation, and some injectates are cytotoxic, resulting in cell death [33, 34, 35]. These studies heightened awareness that even small needle puncture injuries could result in degenerative changes in multiple species, an insight further reinforced by three pivotal 2005 papers on needle puncture injuries in rabbit models [21, 22, 23].

3. Seminal Papers on Surgically Induced IVDD Puncture Models From 2005

The introduction of needle puncture models marked a turning point in the evolution of surgically induced IVDD models. The publication of these three seminal studies in 2005 established simpler methods of inducing IVDD compared to earlier surgically induced injury methods [21, 22, 23]. Around this time, interest in injectable therapies was growing due to the chondrogenic potential of growth factors, like BMPs and TGF‐beta, among others [36, 37]. These models were thus developed to create repeatable IVDD conditions suitable for screening biological therapies. The design of these puncture models was likely inspired by the success of preclinical BMP studies in spinal fusion, which preceded the first human study of recombinant bone morphogenetic protein (rhBMP‐2) showing consistent bone formation and lumbar spinal fusion without the need for autograft or internal fixation [38]. In IVDD puncture models, a needle was inserted into the IVD with a stopper to control puncture depth, creating a complete AF tear with NP depressurization. Model severity could be adjusted by varying the needle gauge and puncture depth. This puncture injury approach was an important innovation, as it was easier to apply and more controlled than prior incision models. The nearly simultaneous publication of these three rabbit studies across different labs also reinforced the model's reproducibility. IVDD was thoroughly characterized using histology, X‐ray, and MRI, establishing parallels to human IVDD.

3.1. Masuda et al.—A Novel Rabbit Model of Mild, Reproducible Disc Degeneration by Anulus Needle Puncture: Correlation Between the Degree of Disc Injury and Radiological and Histological Appearances of Disc Degeneration [22]

Masuda and colleagues characterized the effects of needle diameter on IVDD severity in a rabbit model [22]. Lumbar IVDs (L2–L3 to L5–L6) were punctured using three needle sizes (16‐, 18‐, and 21‐gauge) and a controlled puncture depth of 5 mm with a stopper. Adjacent uninjured IVDs served as controls, and the rabbits were evaluated up to 8 weeks post‐injury (Figure 2). This puncture model induced slow, progressive IVDD and was more reproducible than earlier stab models (Section 2; [25, 26, 27, 29]). By varying the needle gauge, the severity of IVDD was carefully modulated, resulting in a gradual progression similar to human IVDD.

FIGURE 2.

Masuda et al. rabbit model of mild, reproducible IVDD by AF needle puncture [22]. Top left panel: illustration of the puncture method, including (A) stopper to control needle depth and (B) process for calculating radiographic disc height index (DHI). Bottom left panel: (A) the stab injury causes rapid loss of DHI and IVDD progression while (B) the puncture model induces slow, progressive loss of DHI and IVDD. Right Panel. (A) The needle gauge can be modified to control severity of IVDD with 18‐ and 16‐gauge needles causing moderate–severe IVDD after 8 weeks while 21‐gauge needle induced minimal IVDD.

In contrast, stab defects caused acute, rapid degeneration followed by a plateau. This mild, reproducible IVDD enabled pathophysiological studies of early‐stage degeneration and the mechanisms driving the transition from acute to chronic IVDD. The precise and repeatable puncture injury reduced variability compared to other models. The IVD height index developed by Masuda and colleagues has become a standard metric in IVD research, facilitating cross‐study and cross‐species comparisons (Figure 2) [22]. This model, using an 18‐gauge needle with interventions at 4 weeks post‐puncture, was rapidly adopted to screen growth factor therapies and continues to be applied for evaluating IVD pathophysiology and testing other biological treatments [22, 39, 40, 41, 42, 43, 44, 45, 46].

3.2. Sobajima et al.—A Slowly Progressive and Reproducible Animal Model of Intervertebral Disc Degeneration Characterized by MRI, X‐Ray, and Histology

Sobajima et al. applied similar methods to create a slow, progressive rabbit IVDD model (Figure 3) [23]. Their model induced an anterolateral AF injury using a 16‐gauge needle to mimic the gradual progression seen in human IVDD. Multimodal imaging techniques, including MRI, radiographs, and histology, tracked degeneration over 24 weeks, providing a robust and clinically relevant characterization of IVDD (Figure 3). Findings showed a progressive decrease in IVD height and MRI signal intensity in the NP starting at 3 weeks post‐injury, through 24 weeks. Histology showed loss of notochordal cells in the NP and replacement of normal tissue with fibrocartilage, indicating significant AF disruption. The model's use of a stopper (Figure 3) enabled a straightforward surgical procedure and ensured high repeatability for studying early‐stage degeneration, the phase most relevant to regenerative therapies. The slow, progressive rabbit IVDD model was subsequently applied to screen safety and efficacy of gene therapy methods for IVDD repair [47, 48, 49, 50], and highlighted potential complications of cell therapies, showing intradiscal mesenchymal stem cell injection could involve cell leakage with osteophyte formation [51].

FIGURE 3.

Sobajima et al. rabbit model of slowly progressive IVDD by AF needle puncture [23]. Rabbit lumbar IVDs were stabbed by (A) 16‐gauge hypodermic needle to a controlled depth of 5 mm in (B) left anterolateral AF. (C) Representative MRI, radiograph, and histologic views of different rabbit lumbar IVDs presurgery and 3, 6, 12, and 24 weeks after stab by 16‐gauge hypodermic needle, demonstrating progressive and reproducible changes like those seen in human IDD.

3.3. Kim et al.—Disc Degeneration in the Rabbit: A Biochemical and Radiological Comparison Between Four Disc Injury Models

Kim and colleagues presented IVDD models in rabbits using four IVD injury types that varied the extent of IVDD by applying different needle injectates and puncture injury severity [21]. Intradiscal camptothecin injection (an apoptotic agent), nucleus aspiration, and two annular puncture methods (21‐gauge triple puncture, 18‐gauge single puncture) were applied. IVDD was assessed through MRI, biochemical assays of water and sulfated glycosaminoglycan content, and histology. Results showed greater variability in the severity of IVDD induced by different injury types than by varying needle gauge alone. Both NP aspiration and AF punctures resulted in decreased glycosaminoglycan content and water in the IVD. However, the 21‐gauge triple puncture demonstrated the most reproducible changes and showed measurable decreases in IVD height and MRI grades of degeneration, making it more effective than the 18‐gauge single puncture model. Further, the 21‐gauge triple puncture model's reliability highlighted the importance of multiple punctures in causing IVDD progression and was therefore recommended for studies evaluating therapeutic interventions. This study's radiological and histological assessments, and the emphasis on direct biochemical measurements, provided additional details on IVDD pathophysiology and allowed comparisons with the human IVDD condition [21]. The regenerative potential of mesenchymal stem cells and the therapeutic effect of adenovirus‐mediated Wip1 following this IVDD injury in New Zealand White rabbits has also been assessed [52, 53].

Together, the slow, progressive rabbit IVDD models [21, 22, 23] inspired many subsequent papers on IVD pathophysiology and therapeutic screening of biological therapies for IVDD in numerous subsequent research papers. Lastly, these papers reinforced the clinical literature on discography, emphasizing that needle puncture injuries can cause IVDD and must be avoided in healthy control IVDs. Puncture models became increasingly common for inducing IVDD, with a rapidly increasing publication rate after 2005 (Figure 4).

FIGURE 4.

Publication trends for IVD puncture models rapidly increased after 2005. Scientific articles published using data compiled on a PubMed search for “Intervertebral degeneration puncture” conducted on March 03, 2025. All studies are presented, although filtering to exclude human studies and review articles reduced the total number of papers from 648 to 464 (~72% of the total).

4. Animal Species and Spinal Regions Where Puncture Models Are Used

The simplicity of the puncture model technique, along with its clinical relevance to human discography and local delivery of IVD therapeutics, has made it widely adopted across multiple species. Since the introduction of the rabbit puncture model as a reproducible and scalable method to induce localized IVDD [21, 22, 23, 52], Abe, Akeda, An, Aoki, Pichika, Muehleman, Kimura, and Masuda [54], usage of puncture models has expanded considerably, with numerous studies utilizing animal models to recapitulate degenerative pathophysiology. While rodents, rabbits, cows, and dogs have been studied, rodent and rabbit models are most used in part due to their relatively small sizes. The choice of species for modeling IVDD is contingent on factors including anatomical similarity to humans, surgical feasibility, cost, and ethical considerations, with each species offering distinct advantages and drawbacks in modeling IVDD, as described in multiple reviews [14, 15, 16].

4.1. Rabbit

The leporine model is widely used in puncture‐induced IVDD due to its intermediate IVD size and well‐characterized degenerative progression [21, 22, 23]. Rabbit IVDs are particularly amenable to precise surgical interventions, biologic injections, and advanced imaging, which likely contributed to their early adoption in puncture model development. These models have consistently induced structural and biochemical changes comparable to early IVDD [14, 55]. Recent puncture studies in rabbits have demonstrated progressive AF fissures and NP dehydration, with radiographic signs of IVDD becoming apparent within 4–6 weeks [56, 57, 58]. Rabbit puncture models commonly use 16‐gauge and 18‐gauge needles at 5 mm depth, though larger and smaller needle gauges have been applied (Table 1). The resulting IVDD phenotypes range from slow progressive degeneration to more severe conditions, including denucleation and radiculopathy.

TABLE 1.

Representative IVD puncture model studies with details on species and injury.

Species	Spinal levels	IVDD phenotype	Compartment	Needle G, (puncture % IVD Height)	Puncture depth, % of IVD diameter	# of punctures, needle motion	Puncture approach	Injectate, dwell time	Author, year
Rabbit	L2–3 to L5–6	Slow, progressive mild/moderate IVDD	NP, AF	16G, 18G, 21G	5 mm	1, static	Posterolateral, retroperitoneal	—	Masuda 2005 [22]
Rabbit	L2–3 to L4–5	Slow, progressive mild/moderate IVDD	NP, AF	16G, 18G, 21G	5 mm	1, static	Anterolateral	—	Sobajima 2005 [23]
Rabbit	L2–3 to L4–5	Slow, progressive mild/moderate IVDD	NP, AF	18G, 21G, 23G	—	1 (18G), 3 (21G), static	Retroperitoneal	8 mL NEP (21G), camptothecin (23G), 15 s	Kim 2005 [21]
Rabbit	L2–3 to L5–6	Severe IVDD (acute)	NP, AF	18G	5 mm	1, 360° rotation	Anterolateral, retroperitoneal	10 mL NEP	Moss 2013 [59]
Rabbit	L2–3 to L4–5	Severe IVDD (acute)	NP, AF	19G	—	1, static	Posterolateral at 30°–45°	10 mL NEP	Kim 2015 [57]
Rabbit	L2–3 to L5–6	Severe IVDD (acute)	NP, AF	18G	5 mm	1, 360° rotation	Anterior	5 mL NEP, 10 s	Chee 2016 [60]
Rabbit	L3–4 to L5–6	Slow, progressive mild/moderate IVDD	NP, AF	18G	5 mm	1, static	Anterior	5 s	Lei 2017 [61]
Rabbit	L5–6	Severe IVDD (acute)	NP, AF	18G	—	1, static	45° oblique‐cranial	10 mL NEP	Luo, 2018 [62]
Rabbit	L2–3 to L4–5	Slow, progressive mild/moderate IVDD	NP, AF	16G	5 mm	1, 360° rotation	Anterior	—	Ashinksy 2019 [56]
Rabbit	L2–3 and L4–5	Slow, progressive mild/moderate IVDD	NP, AF	18G, 26.5G	5 mm	1, static	—	PBS or GDF6	Miyazaki 2018 [44]
Rabbit	L2–3 to L4–5	Slow, progressive mild/moderate IVDD	NP, AF	21G	—	1, 360° rotation	Anterolateral	30 s	Ogunlade 2019 [63]
Rabbit	L3–4 to L5–6	Slow, progressive mild/moderate IVDD	NP, AF	14G, 16G, 21G	—	1, 180° rotation	Anterior at 15°–25° bilaterally	5 s	Wang 2020 [64]
Rabbit	L5–6, L6–7	Severe IVDD (acute)	NP, AF	16G, 26G	5 mm	1, static	Anterior	PBS or ChABC, 30 s	Yang 2022 [65]
Rat	L4–5	Herniation	NP, AF	25G, 27G	—	1, static	Posterior	0.3 mL PP	Onda 2005 [66]
Rat	c5–6, c7–8	Slow, progressive mild/moderate IVDD	NP, AF	18G, 21G	5 mm	1, 180° rotation	—	5 s	Zhang 2009 [67]
Rat	c6–7, c8–9	Severe IVDD (acute)	NP, AF	18G, 22G, 25G	2 mm	1, static	—	—	Hsieh 2009 [68]
Rat	c6–7, c8–9	Severe IVDD (acute)	NP, AF	18G, 20G, 22G	5 mm	1, static	—	—	Keorochana 2010 [69]
Rat	L4–5	Severe IVDD (acute), herniation	NP, AF	0.4 mm	—	1, static	Anterior	—	Nilsson 2011 [70]
Rat	c6–7, c8–9	Severe IVDD (acute)	NP, AF	20G	—	1, 360° rotation ×2	Anterior	30 s	Issy 2013 [71]
Rat	L4–5, L5–6	Slow, progressive mild/moderate IVDD	NP, AF	27G	—	1, static	Posterior	—	Li 2014 [72]
Rat	Caudal	Severe IVDD (acute)	NP, AF	20G	—	1, static	Posterolateral	0.5 mL NEP	Stannard 2016 [73]
Rat	L3–4, L4–5, L5–6	Chronic IVDD and severe IVDD (acute)	NP, AF	26G	1.5, 3 mm	1, static	—	Saline, TNFα, or NGF and VEGF	Lai 2016 [74]
Rat	c5–6, c7–8, c9–10	Slow, progressive mild/moderate IVDD	NP, AF	18G, 21G, 25G	—	1, 180° rotation	—	5 s	Chen 2018 [75]
Rat	c6–7 to c8–9	Slow, progressive mild/moderate IVDD	NP, AF	18G, 21G, 23G, 25G, 27G, 29G	5 mm	1, static	—	20 s	Hu 2018 [76]
Rat	L3–4 to L5–6	Chronic IVDD and Severe IVDD (acute)	NP, AF	26G	1.5, 3 mm	1 static, 3 static	Anterior	TNFa	Mosley 2019 [77]
Rat	c7–8 to c9–10	Herniation, Severe IVDD (acute)	NP, AF	16G, 18G, 26G	—	1, 360° rotation	—	30 s	Qian 2019 [78]
Rat	L5–6	Chronic IVDD and severe IVDD (acute)	—	27G	2 mm	10, static	Left posterolateral	0.3 mL air	Leimer 2019 [79]
Rat	c5–6 to c7–8	Slow, progressive mild/moderate IVDD	NP, AF	21G	3 mm	1, 360° rotation	—	30 s	Tian 2021 [80]
Rat	L5–6	Slow, progressive mild/moderate IVDD	—	0.5 mm	3 mm	1, 90° rotation	—	—	Lillyman 2022 [81]
Rat	c6–7 to c8–9	Severe IVDD (acute)	NP, AF	21G, 25G, 29G	5 mm	1, 360° rotation ×2	—	30 s	Elmounedi 2022 [82]
Rat	L4–5, L5–6	Chronic IVDD and severe IVDD (acute) with Modic‐like changes	NP, AF, EP	0.6 mm (Kirschner Wire)	3 mm	1, static	Anterior	TNFα or PBS	Wang 2023 [83]
Rat	c8–9	Severe IVDD (acute)	NP, AF	20G	4 mm	1360° rotation	—	30 s	Zhu 2023 [84]
Rat	c5–6 to c7–8	Slow, progressive mild/moderate IVDD	NP, AF	21G	—	1, static	—	10 μL of NEP	Ambrosio 2024 [85]
Rat	L3–4 to L5–6	Severe IVDD (acute)	NP, AF	26G	3 mm	3, twisting and side‐to‐side probing	Anterolateral (1) + posterolateral (2)	PBS	Lai 2024 [13]
Rat	c3–4 to c6–7	Chronic IVDD	NP, AF	33G (< 25%)	4 mm	1, static	—	PBS or LPS, 1 min	Lisiewski 2024 [86]
Mouse	c4–5	Slow, progressive mild/moderate IVDD	AF	31G	1 mm	1, static	—	—	Yang 2009 [87]
Mouse	c6–7, c8–9	Severe IVDD (acute)	NP, AF	26G (90%), 29G (65%)	1.75 mm	1, static	Posterior	—	Martin 2013 [88]
Mouse	L4–5	Severe IVDD (acute)	NP, AF	33G, 35G	—	1, static	Right posterolateral	—	Ohnishi 2016 [89]
Mouse	c7–8, c9–10	Slow, progressive mild/moderate IVDD	NP	27G, 29G, 31G	—	1, static	—	—	Piazza 2018 [90]
Mouse	c4–5, c6–7	Herniation	NP, AF	Adults: 26G (80%), Neonates: 31G (80%)	50% IVD diameter	1, static	Posterolateral	30 s	Torre 2018 [91]
Mouse	c4–5, c6–7	Severe IVDD (acute)	NP, AF	26G (90%)	50% IVD diameter	1, static	Posterior	—	Zhang 2019 [92]
Mouse	L4–5, L5–6, L6–S1	Slow, progressive mild/moderate IVDD	NP, AF	30G	1 mm	1, static	—	2 μL of saline	Tang 2021 [93]
Mouse	L6–S1	Severe IVDD (acute), herniation	NP, AF	Scalpel no. 11, 33G	0.3 mm, 100% of IVD diameter	1, static	Retroperitoneal	—	Walk 2022 [94]
Mouse	L3–4 to L5–6	Severe IVDD (acute)	NP, AF	26G (70%)	1 mm	1, static	Anterior	5 min	Zhang 2022 [95]
Mouse	c3–4, c5–6, c7–8	Herniation	NP, AF	80%	50% IVD diameter	1, static	—	—	D'Erminio 2024 [96]
Sheep	Caudal	Severe IVDD (acute)	AF	—	5 mm	1, static	Left anterolateral	—	Osti 1990 [29]
Sheep	Cervical 2–3 to 5–6	Herniation	NP, AF	2 mm (punch)	4 mm	1, static	Anterolateral	—	Long 2020 [97]
Sheep	L1–2, L2–3, L3–4	Progressive, severe IVDD, herniation	NP, AF	Scalpel no. 11	8 mm	1, static	Lateral retroperitoneal at 45°	PEGDA, FibGen, C‐MC, FibGen + C‐MC	Panebianco 2023 [98]
Goat	L1–2 to L4–5	Slow, progressive mild/moderate IVDD	AF	22G	—	1, static	Left lateral, retroperitoneal	ChABC	Gullbrand 2016 [99]
Cow	Caudal	Severe IVDD (acute)	NP, AF	14G, 25G	—	1, static	Posterolateral	—	Korecki 2008 [100]
Cow	Caudal	Severe IVDD (acute)	NP, AF	21G	—	1, static	Rotated clockwise	LPS or IL1β, 30 s	Teixeria 2016 [101]
Pig	L1–2 to L3–4	Severe IVDD (acute) with Modic‐like changes	EP	18G	4 mm	1, static	—	—	Sheyn 2019 [102]
Pig	L1–6	Severe IVDD (acute)	AF	3.2 mm (trephine)	2–3 mm	1, static	—	—	Yoon 2008 [103]
Pig	L2–3, L3–4	Severe IVDD (acute)	NP, AF	16G	—	1, rotation	Anterolateral, retroperitoneal	60 s	Omlor 2010 [104]
Pig	L1–2 to L4–5	Severe IVDD (acute)	NP, AF	Scalpel no. 11	—	1, static	—	—	Acosta 2011 [105]
Pig	L1–2, L2–3	Slow, progressive mild/moderate IVDD	AF	3.5–4 mm (trephine)	3 mm	1, static	—	—	Shi 2016 [106]
Dog	L6–7	Herniation	NP, AF	18G	—	1, static	Posterolateral	0.01 mL saline	Otani 1997 [31]
Dog	L1–2, L3–4, L5–6	Severe IVDD (acute)	NP, AF	16G, 18G, 21G	7 mm	1, 180° rotation	Posterolateral, retroperitoneal	5 s	Gu 2017 [107]
Dog	L1–2, L3–4, L5–6	Slow, progressive mild/moderate IVDD	NP, AF	32G	—	1, static	—	NTG‐101 or PBS	Matta 2018 [108]
Dog	L1–2, L3–4, L5–6	Severe IVDD (acute)	NP, AF	20G	—	1, static	—	NTG‐101 or PBS	Matta 2022 [109]

Abbreviations: AF, annulus fibrosus; c, coccygeal; ChABC, chondroitinase ABC; C‐MC, carboxymethylcellulose‐methylcellulose; EP, endplate; FibGen, genipin‐crosslinked fibrin; G, gauge; IVD, intervertebral disc; IVDD, intervertebral disc degeneration; L, lumbar; LPS, lipopolysaccharide; min, minute; mm, millimeter; NGF, nerve growth factor; NEP, negative pressure; NP, nucleus pulposus; NTG‐101, rhTGF‐β1 and rhCTGF proteins suspended in an excipient solution; PBS, phosphate‐buffered saline; PEGDA, poly(ethylene glycol) diacrylate; PP, positive pressure; s, second; TNFα, tumor necrosis factor alpha; VEGF, vascular endothelial growth factor.

In rabbits, notochordal cells persist in IVDs throughout adulthood and contrast with human IVDs where they are replaced by chondrocytic cells during adolescence [110]. The notochordal cell‐rich characteristic of the rabbit IVD is thought to contribute to its strong innate healing capacity [111, 112, 113]. A well‐healing model is a strength for early‐stage screening of IVDD therapies where identifying its potential efficacy is a more important factor than determining its potential translatability to human IVDs. For example, if a therapy shows limited or no benefit treating rabbit IVDs with their favorable reparative capacity, then the therapy is unlikely to be effective treating human patients. The strong healing potential of rabbits and other species with notochordal cell retaining IVDs has inspired an expansive literature, and precise methods for notochordal cell isolation, culture, and characterization are established [114].

A more recent study on the therapeutic potential of growth differentiation factor‐6 (GDF6) for IVDD‐related pain involved the novel application of rabbit and rat models [44]. Specifically, a puncture‐induced IVDD was induced in rabbits to investigate its potential for IVD repair. However, since rabbits are known to mask many pain‐like behaviors, the degenerated and treated rabbit NP tissue was then transplanted onto L5 dorsal root ganglia of nude rats as a radiculopathy model for pain‐like behaviors [44]. Using this dual‐species experimental approach, GDF6 treatment partially rescued the IVDD in rabbits and also had less allodynia in the nude rats.

4.2. Rat

Rat models offer logistical advantages with respect to care and housing as a smaller vertebrate animal than rabbits yet remain sufficiently large to allow precise surgical injuries. Early rat puncture models targeted coccygeal IVDs because they are more accessible for surgical procedures and share anatomical and structural similarities with human lumbar IVDs (Table 1) [79, 115]. The size of rat IVDs requires precise microsurgical techniques to ensure reproducibility and minimize off‐target effects. Improvements in spine surgery methods in rats enabled puncture injuries in lumbar spines to induce IVDD (Table 1). Puncture injuries to the AF and NP, and vertebral EP cause distinct biomechanical changes, matrix breakdown, and progressive degeneration [67, 74, 83, 84, 115, 116, 117]. Rat IVD models are large enough for intradiscal injections, facilitating the study of cell therapies such as extracellular vesicles from NP cells [85].

Rat models have well‐established pain‐like behavioral assays that offer insights into pain mechanisms [15, 116]. Rat models also allow for the assessment of sensitization, neuroinflammation, and neuronal changes in both the peripheral and central nervous systems in addition to IVDD studies. Rat models therefore can be used to provide a comprehensive, whole‐animal understanding of the progression of IVDD and related pain sequelae. Such measurements have enabled the study of crosstalk between IVDD and the nervous systems, providing insights into pain mechanisms and neuroimmune interactions [12, 74, 75, 81, 84, 116, 118, 119].

Rat lumbar IVDs have a height of less than 1 mm, and 26‐gauge needles (approximately 0.5 mm in diameter) are commonly used to induce IVDD phenotypes (Table 1). However, both larger and smaller needles have been used to modulate the IVDD severity, ranging from slow, progressive mild/moderate to severe IVDD (acute) conditions with AF disorganization and loss of NP, alongside chronic IVDD, herniation, and Modic‐like changes (Table 1). Puncture injury methods have also evolved to include complete and incomplete AF puncture, injury at multiple lumbar levels, single versus multiple punctures, needle motion variations, and use of injectates (Figure 5).

FIGURE 5.

Evolution of puncture methods in rat IVDD models. Needle puncture injury and injection are commonly used in rat models of IVDD. (A) Controlled puncture depth to cause partial (incomplete) or full (complete) AF tears [74]. (B) Puncture to three adjacent lumbar levels via the anterior approach with injection of PBS or TNF‐alpha [13, 77]. The needle scrape method induces greater IVD disruption using a pin‐type needle without cutting bevel [81]. (D) EP microfracture in vivo model showing anterior transcorporeal approach [83, 120].

4.3. Mouse

Mouse models are essential tools for IVDD research due to their relatively short lifespan, which makes it technically feasible to study disease onset and progression across varying ages, something that is challenging to achieve in humans [121]. The availability of genetic mouse models offers an unparalleled genetic toolkit for dissecting molecular mechanisms underlying IVDD, making them indispensable for mechanistic studies. These models generally fall into categories such as aging, genetic, mechanically induced, and puncture injury models [15].

However, surgically induced puncture injuries are technically challenging because of their small IVD size, limiting precision and control. The limited space within the mouse abdomen complicates the surgical approach in the lumbar spine, making coccygeal IVD injuries the most used in puncture models (Table 1). Studies have reported altered biomechanical properties, along with reduced glycosaminoglycan and collagen contents [87, 88]. Puncture models in coccygeal mouse IVDs have applied different needle diameters, typically reported as a percent of the total IVD height, and controlled puncture depths to modulate the severity of degeneration (Figure 6) [88, 90]. This approach has even been applied to extremely small neonatal IVDs, with needle gauge adjusted to create injuries reaching up to 80% of IVD height in mice aged from postnatal day 1 (p1) through adulthood (Figure 6) [91, 96]. Mouse puncture injuries in the lumbar spine are less common and include needle puncture and scalpel incision [15, 89, 94]. Recently, a mouse lumbar IVDD puncture model was used to deliver extracellular vesicles, demonstrating the feasibility of using lumbar mouse IVD puncture models for studying IVDD and for screening injectable therapies, which is an important advance [93].

FIGURE 6.

Puncture injury methods in mice applied to coccygeal and lumbar IVDs. (A) Coccygeal IVD puncture model in mice [90], with (a) schematic of the mouse coccygeal spine showing injured and internal control IVD levels. (b) Three needle sizes used for IVD injury to vary injury as a percentage of IVD height. (c) Unilateral and bilateral AF puncture injuries to modify severity. (B) Coccygeal IVD puncture to induce IVD herniation with 80% IVD height and 50% IVD width puncture depth in adult and neonatal mice [96]. Needle gauge was adjusted to retain similar relative injury size as neonatal mice grew from postnatal day 1 (p1) through p28. (C) Intradiscal injection into mice L4/5, L5/6, L6/S1 IVDs after performing a left unilateral incision in vivo to deliver extracellular vesicles engineered to deliver FOXF1 cargo [93].

Regarding pain and pain‐related behaviors, both humans and mice with IVDD experience difficulties such as altered gait, reduced physical activity, limited range of motion, heightened sensitivity to mechanical and thermal stimuli, and changes in nerve plasticity [15, 122, 123, 124, 125, 126]. Mice display pain behaviors comparable to those seen in humans, including lower tolerance to axial stretch, thermal sensitivity, and impaired movement [127]. In humans, thermal hyperalgesia linked to low back pain can present as sensations of cold, radiating discomfort, and cold‐induced pain, which aligns with results from cold plate and acetone sensitivity tests in mouse models [128]. Furthermore, mice with IVDD subjected to tail suspension show reduced immobility and increased efforts to alleviate discomfort, mirroring the human tendency to adjust movements to minimize pain [129, 130, 131, 132].

4.4. Large Animal Models

Sheep and goat models are highly relevant for puncture‐induced IVDD due to their IVD size, EP morphology, and biomechanical properties that closely resemble human pathology [30, 99, 115]. Their large size enables controlled, reproducible injuries and provides sufficient IVD volume for injecting biomaterials and regenerative therapies, as well as for evaluating surgical interventions. However, their high cost, long degenerative time courses, and increased regulatory oversight and need for specialized care limit their use in early‐phase research, making them more suitable for advanced preclinical testing.

Cow IVDs are often used in ex vivo puncture models due to their large size and availability from abattoirs (Table 1) [14, 16, 133, 134, 135]. Refined cow IVD organ culture systems provide high‐fidelity loading and culture conditions, enabling the study of degenerative changes and cellular responses without sacrificing live animals. Puncture injuries range from small needle punctures to larger defects created using biopsy punches or large syringe needles. These models have been used to assess cellular responses, biomechanical changes, and inflammatory processes. For instance, ex vivo studies have examined how needle punctures simulate human discography and demonstrated that injury severity varies by needle gauge [100, 136, 137]. Studies have additionally used cow IVDs to analyze inflammation in degeneration, finding that anti‐inflammatory treatments like diclofenac and MSC therapies could reduce inflammation and promote tissue regeneration [101]. Another study tested a gelatin‐based hydrogel for restoring IVD integrity post‐puncture, finding certain formulations could restore pressure levels comparable to intact IVDs [138]. Additionally, research has examined microscopic damage to the AF caused by needle punctures, highlighting collagen fiber structural changes that contribute to progressive degeneration [139]. The tough, fibrous AF of large animals can make it challenging to induce severe degeneration with biopsy punches and syringe needles, so scalpel injuries are common [29, 30, 97, 98].

Porcine models can be valuable for puncture‐induced IVDD, as their lumbar IVDs share some similarities with humans in biochemical composition and mechanical loading environment [140, 141]. However, their IVD height is generally smaller than in humans, and notochordal cell retention can persist in certain breeds, particularly miniature pigs, potentially affecting healing and IVDD progression [16]. Despite these differences, puncture studies in porcine models have demonstrated progressive degeneration over 3–6 months, characterized by IVD height loss, proteoglycan depletion, and collagen remodeling [102, 104, 105, 142]. Porcine IVDs may retain a notochordal nucleus pulposus into more advanced ages, although this appears to be breed‐dependent, with notochordal cell retention observed in breeds such as Yucatan pigs [102]. While their rapid growth rate allows for longitudinal assessments, the high costs make porcine models most relevant to large animal therapeutic validation stages. Puncture injuries in porcine models commonly describe needle diameter rather than needle gauge, and scalpel injuries and trephines are also applied to induce IVDD. A 16‐gauge needle inducing IVDD with annular puncture and partial nucleotomy has been shown to induce severe IVDD [104].

Dogs, particularly chondrodystrophic breeds, serve as naturally occurring models predisposed to IVDD. Cervical and lumbar IVD puncture models replicate annular tears, nucleus pulposus extrusion, and biomechanical instability, similar to human degeneration [107, 108, 109, 143, 144]. However, the natural occurrence of IVDD in dogs, especially in chondrodystrophic breeds, makes induced puncture models less applicable for studying disease pathophysiology. Instead, puncture injuries in dogs are primarily used for intradiscal delivery of therapeutics (Table 1) [145, 146]. Treatment of dog “patients” makes for an excellent late‐stage model for validating therapeutics for IVDD. Nonetheless, the large genetic heterogeneity across dog breeds requires relatively large sample sizes to carefully control research studies. Although spontaneous IVDD is more frequent and severe in chondrodystrophic breeds, the histological features of degeneration, such as chondroid metaplasia, are similar in both chondrodystrophic and non‐chondrodystrophic breeds once IVDD develops [145].

5. Variability in Puncture Needle Injuries and Considerations Across Small and Large Animal Models

Puncture methodologies vary widely within and across species, influencing IVDD phenotype and translational relevance. Factors like spinal level, needle type, and insertion dynamics, among others, can profoundly alter the IVDD conditions induced. Nevertheless, these models are often broadly referred to as “puncture” injuries, and the level of methodological detail provided across studies is highly variable (Table 1). This section, therefore, describes the many variations in surgical puncture method details and suggests a set of reporting puncture details to augment our understanding of the injury induced and improve reproducibility and comparability across labs.

5.1. Spinal Levels and IVD Compartment Injured

IVD anatomy and biomechanical properties vary substantially with spinal level. Lumbar IVDs are frequently targeted for their weight‐bearing function and human relevance, while coccygeal and cervical IVDs are chosen for their accessibility and ability to address model‐specific considerations, such as applying mechanical loading with external fixators. Rat and mouse models commonly apply puncture injuries to lumbar and coccygeal IVDs [13, 67, 68, 70, 75, 80, 84, 115, 117]. IVD punctures can be superficial to injure only the AF or deep enough to injure both AF and NP [74]. The EP defect can be incorporated into the injury model through the IVD or via a transcorporeal approach to injure the EP [83, 117]. EP injury from puncture disrupts axial biomechanical properties more severely, while AF injury from puncture has a greater impact on torsional biomechanical properties [117].

5.2. Needle Type, Gauge, and Diameter as a % of IVD Height

IVD puncture injuries most commonly use syringe needles with sharp, beveled tips. However, injuries can also be induced by trephines, scalpels, biopsy punches, or pins [74, 81, 98, 147], with each puncture tool causing distinct injuries and contributing to different severities of structural damage and degeneration. Sharp, beveled needles produce clean, uniform cuts through AF fibers to simulate annular tears. However, in small animal models, such as rats and mice, the relatively long bevel length compared to IVD size may result in a more cut‐like injury than intended when varying needle gauge, necessitating careful consideration in study design [98, 147]. Biopsy punches and/or blunt needles induce broader, biopsy‐type injury, causing more extensive fiber disruption, mimicking severe IVD injury or herniation. Sharp pins, though cutting the fewest number of annular fibers initially, can induce IVD disruption if manipulated post‐insertion, thereby cutting more fibers through rotational or translational motion [81]. While no studies to our knowledge have deliberately examined the effects of sharp versus blunt needles on the IVD or cartilage, one dog study revealed differences in hemorrhage incidence upon internal structure needle penetration [148]. The different puncture‐type injuries with needles, punches, and pins induce distinct injury patterns that can simulate various degenerative conditions. Punches, due to their larger diameter, typically cause more severe injuries, while pins induce less morbidity for the same type of injury.

Needle gauge is equally consequential, as larger needles (26‐gauge, ~90% of mouse IVD height) induce greater annular fissures and matrix extrusion, accelerating degenerative changes. Conversely, smaller gauges (29‐gauge, ~65% of mouse IVD height) result in subtle, progressive alterations [88]. Standardizing needle diameter relative to IVD height is crucial for cross‐species comparisons, as the same needle gauge may have very different effects in small versus large animals. A 21‐gauge needle in a murine model may induce catastrophic failure, while the same gauge in a porcine IVD would result in minor perturbations. A study employing a 27‐gauge needle injury impacted 52% of rat IVD height, causing severe biomechanical changes but only 10% of ovine IVD height, causing minimal biomechanical changes [115]. Consequently, it is important to report the needle gauge (or diameter) relative to the IVD height.

5.3. Puncture Depth, Number of Punctures, and Puncture Approach

The depth of needle puncture, causing a complete or incomplete AF rupture, influences the severity of IVDD. A superficial, incomplete AF injury is typically designed to disrupt AF fibers while preserving NP integrity and avoiding herniation. This leads to smaller biomechanical changes, particularly in torsional biomechanics. In contrast, a complete AF injury leads to NP depressurization, biomechanical instability, and proteoglycan loss, accelerating IVDD progression (Figure 5) [74]. Nevertheless, Osti and colleagues showed that AF defects alone were sufficient to induce degeneration in the intervertebral joint complex [29]. The number of punctures per IVD also impacts IVDD progression, and the seminal papers on this topic showed repeated injuries compound structural and inflammatory deterioration [21]. A triple puncture model with injuries at mid‐sagittal as well as anterolateral at both sides has also been reported in the literature (Figure 5) [13, 77, 118]. It has been demonstrated that multiple annular punctures in rat lumbar IVDs resulted in significantly higher degeneration grades and IVD height loss compared to single punctures. This indicates that the severity of degeneration is positively correlated with the number of punctures [77]. Ulrich et al. additionally showed that repeated injury during active healing prolongs inflammation and exacerbates IVDD, emphasizing that damage accumulation and its associated inflammation may underlie pathologic, rather than physiologic aging [141].

Needle insertion trajectory (i.e., posterior, lateral, or anterolateral approach) is typically determined to optimize the surgical approach and limit the involvement of surrounding nerve structures. The choice of approach often depends on the surgical feasibility, which can vary based on the species and the specific injury model used. For rat species, anterior or anterolateral approaches are common, as posterior approaches are typically avoided to reduce the risk of injury to spinal muscles. The anterior approach can induce clinically relevant IVDD‐like conditions without causing posterior IVD herniation and associated neuropathic radiculopathy. This is of importance for small animal models, where limited surgical access requires careful consideration to avoid injury to surrounding neural structures [13, 120, 149]. In contrast, lateral approaches are more commonly employed in larger animal models, where surgical access is more extensive and anatomically similar to human conditions. These approaches provide a broader field of view and help minimize damage to adjacent ligamentous and neural structures. Moreover, larger punctures that induce herniation‐type injuries can be made anteriorly or laterally to avoid direct nerve damage [66, 98, 120, 149].

5.4. Needle Motion: Scraping, Rotation, and More

Needle motion within the IVD, such as rotation or scraping, further influences tissue disruption. Rotational motion intensifies annular fiber disorganization and extensive fibrocartilage remodeling, while scraping along the annular plane induces significant structural defects [75, 80, 81, 117]. Often, rotational motion is employed not to exacerbate degeneration, but to create a uniform injury when puncturing with an angled bevel [22]. The reporting of these variations enables modeling of distinct clinical IVDD phenotypes.

5.5. Dwell Time: Duration of Needle Retention in the IVD

The duration the needle remains within the IVD during puncture (“dwell time”) influences the type and severity of injury produced. Immediate needle withdrawal creates a controlled defect with mechanical impairments such as reduced compressive and torsional stiffness, but these changes may not progress over time [88]. In contrast, longer dwell times can induce more extensive tissue responses. Past injury models have implemented a range of dwell times, from immediate withdrawal up to 5 min. Studies aiming to mimic severe, acute IVDD typically employ dwell times of 30 s or longer, up to 5 min, to induce more pronounced and rapid degeneration. Conversely, shorter dwell times of 5–15 s are generally used to model slow, progressive mild‐to‐moderate IVDD. Based on these findings, we recommend that future studies carefully select and explicitly report dwell time according to the desired injury phenotype, as it significantly affects the severity and progression of degeneration (Table 1) [13, 75, 80, 95].

5.6. Injectates and Chemonucleolysis

One of the advantages of puncture models is their compatibility with intradiscal delivery of exogenous agents, allowing modulation of the degeneration process, either accelerating or altering the course of IVDD. Injectates include phosphate‐buffered saline, pro‐inflammatory cytokines, and proteolytic enzymes to induce matrix degradation and accelerate degeneration [13, 115, 150]. Lai et al. found that intradiscal injection of TNFα, NGF, and VEGF in rats accelerated IVDD and pain‐like behaviors [74]. Since young animals heal well, the addition of pro‐inflammatory injectates can therefore enhance a nociceptive response. On the other hand, Matta et al. demonstrated that a single injection of recombinant TGF‐β1 and CTGF proteins (NTG‐101) into injured IVDs in preclinical rodent and dog species maintained IVD height, improved ECM retention, and reduced inflammatory markers as a preclinical screening tool [108]. Negative and positive pressure may also be employed to disrupt and potentially extrude the NP [57, 59, 60, 62]. Removal of NP tissue volume via aspiration better enables delivery of cells or other biomaterials for repairs, although IVD tissue volume can also be removed by enzymatic degradation.

Chemonucleolysis is a commonly used method for inducing IVDD by enzymatically degrading the NP. Chemonucleolysis is best known for its clinical application to reduce pain in lumbar disc herniation by chemically depressurizing the NP [151]. Injection of enzymes such as chymopapain or chondroitinase ABC directly into the IVD in preclinical models can also mimic the loss of proteoglycans and hydration observed in human IVDD [152, 153]. In both of these conditions, enzyme delivery typically occurs through a small‐gauge syringe needle so that IVD disruption is predominantly due to enzymatic protein degradation with minimal mechanical disruption to the annulus from the needle puncture. In contrast, puncture injury models simulate IVDD conditions with greater structural disruption by the needle insertion, motion, and so forth. However, these conditions can be combined in preclinical IVDD models.

Chemonucleolysis to induce biochemical matrix degradation has been applied to multiple species, including rats, mice, dogs, goats, and sheep. IVDD models with chemonucleolysis depend on the concentration of enzyme delivered. For example, IVDD induced in rabbits by intradiscal enzyme injection was dependent on control injection volume and concentration [46]. Both large and small species have shown levels of IVDD from chemonucleolysis, although spontaneous regeneration has been observed in smaller animals such as rats [154, 155]. Chemonucleolysis with agents like ChABC in goats, sheep, and dogs has not shown regeneration [156, 157, 158]. Methodologically, intradiscal injection requires greater precision and high enzyme concentration due to limited IVD space and NP volume. The larger size of IVDs in larger animals requires greater enzyme volumes to achieve effective matrix degeneration, exacerbating degeneration [99]. Puncture and enzymatic models offer complementary approaches for studying IVDD, simulating mechanical and biochemical degeneration, respectively, with their usage guided by the targeted pathology, species‐specific anatomy, and study objectives.

6. Limitations on the Use of the Puncture Model

Puncture models are widely used to induce IVDD in animal models due to their simplicity, reproducibility, and flexibility across species. They are especially valuable for testing hypotheses on IVDD pathogenesis and for early‐stage therapeutic screening. However, they have notable limitations.

These models create acute, externally applied (“outside‐in”) injuries that differ from the chronic, spontaneous (“inside‐out”) degeneration seen in herniation conditions in humans. They typically focus on disrupting the NP and AF, with limited involvement of vertebral marrow, EPs, or posterior elements, which are important contributors to IVDD and pain in humans [83]. Injury models that inject catabolic agents (e.g., cytokines or enzymes) through a needle to accelerate degeneration are sometimes grouped with puncture models but are mechanistically distinct. In these cases, degeneration is driven by the injected substance rather than the mechanical injury itself [13, 74, 108, 115]. These models often minimize physical disruption by using small needles specifically to avoid puncture‐like injury. As such, they should be categorized separately when reporting or comparing outcomes.

Another challenge is methodological inconsistency. Studies vary widely in puncture parameters like needle gauge, depth, dwell time, and number of punctures, yet these details are often underreported. For example, dwell time impacts degeneration severity; short durations (5–15 s) often mimic slow, progressive IVDD, while longer dwell times (30 s to 5 min) produce more severe, acute degeneration [13, 75, 80, 95]. Standardized reporting of these variables is critical for reproducibility and interpretation. Species differences in IVD biology, such as the presence of notochordal cells and variable healing responses, also affect translational relevance [85]. Moreover, while these injury models can induce IVDD, they do not always replicate the complex, multifactorial nature of human IVDD or reliably model back pain [12, 159, 160].

Given these limitations, puncture models should be carefully selected and reported with sufficient detail. In some cases, alternative models, such as genetic, aging, or mechanical loading models [15, 16], may be more suitable for studying chronic disease processes or multi‐compartment IVD pathology. Injury model selection should align with the specific research question and be clearly justified to enhance translational relevance and scientific rigor.

7. Conclusions and Recommendations

The papers selected in this review highlight the wide variety of injury models collectively described as “puncture” models of IVDD. While grouping these diverse models under the umbrella of puncture injuries helps clarify the concept of surgically induced IVD injury, these models represent many different conditions and injury types that mimic distinct IVDD phenotypes. The broad range of injuries and applications enabled by puncture models underscores their significant impact on advancing IVD research. These models have simplified the induction of IVD injuries and accelerated investigations into pathophysiology, mechanisms, and therapeutic screening. However, many studies lack detailed methodological reporting on puncture injuries, and these methodological details are now known to influence IVDD severity and progression.

We recommend detailed descriptions of puncture injury methods be included in papers to enhance clarity on how injuries are induced. Standardized reporting of key parameters improves methodological transparency and reproducibility and includes spinal levels, IVDD phenotype, needle gauge, puncture depth, number of punctures and needle motion, puncture approach, injectate, and dwell time (Table 2). Such methodological details augment basic animal characteristics, including age, weight, and sex.

TABLE 2.

Recommendations on reporting of IVDD puncture model details.

Puncture model characteristics	Details
Spinal levels	Specify injured levels and controls (e.g., L3‐L4 injury, L5‐L6 control)
IVDD phenotype	Slow, progressive IVDD; severe IVDD (acute); chronic IVDD; herniation, Modic‐like, others
Compartment	Annulus fibrosus (AF), nucleus pulposus (NP), endplate (EP), others
Needle G, (puncture % IVD height)	Syringe needle, sharp needle, trephine, scalpel, gauge size and/or diameter (e.g., 21G, 18G, 0.8 mm), needle gauge as % of IVD height
Puncture depth, % of IVD diameter	Specify in mm and % of IVD diameter
# of punctures, needle motion	Single, triple, or custom pattern, rotating, twisting, scraping, or other motion applied
Puncture approach	Anterolateral, posterolateral, retroperitoneal, or others
Injectate, dwell time	Saline, cytokines, enzymes, or other agents used, duration needle remains in place (e.g., seconds, minutes)

Variations in puncture techniques can markedly affect the severity and type of IVDD induced and include slow, progressive IVDD; severe IVDD; chronic IVDD; herniation; and Modic‐like changes (Table 3). Severe acute IVDD models often use dwell times exceeding 30 s, puncturing both the NP and AF, frequently at lower lumbar levels. Larger needles are typical, with puncture depths of 3–5 mm (roughly 50%–100% of IVD diameter), and needle rotation and multiple punctures are common to increase injury severity. Chronic IVDD models tend to use smaller gauge needles (around 26G) without needle motion, often combined with pro‐inflammatory agent injections and longer recovery periods. Slow, progressive mild‐to‐moderate IVDD models vary in needle size but typically use dwell times under 30 s and involve full IVD puncture at various spinal levels. In this review, we focus on key phenotype characterizations, including histology, radiography, MRI (which parallels human clinical imaging and Modic changes), pain‐like behaviors, biomechanical function, and compositional assessments (Table 4).

TABLE 3.

Puncture injuries strategies and IVDD phenotypes induced.

IVDD phenotype	Puncture % IVD height	Puncture depth, % of diameter	# of punctures, needle motion
Slow, progressive mild/moderate IVDD	25%–50% IVD height	20%–50% of IVD diameter	Single puncture, static or minor rotation, 180° rotation creates a more uniform injury due to the angled bevel
Severe IVDD (acute)	70%–90% of IVD height	60%–80% of IVD diameter	Single or multiple punctures, aggressive rotation or scraping to induce greater disruption
Chronic IVDD	Small‐gauge needle repeated punctures + cytokine injection	30%–50% of IVD diameter	Single puncture, inject pro‐inflammatory agents (TNFα, LPS, etc.) and/or wait 4 weeks post‐injury
Herniation	Large‐gauge needle (16G or larger), scalpel incision, or punch (2 mm for large animal)	80%–100% of IVD diameter	Single puncture, ensure NP extrusion/aspiration and AF fiber separation
Modic‐like changes	Endplate‐involving injury	3–5 mm into endplate	Single or multiple punctures, combine with inflammatory agent injection

TABLE 4.

Outcome measures for characterizing IVDD in puncture models.

IVDD characterization	Details
Histology	IVDD grading and other structural changes
Radiography	IVD height loss, disc height index (DHI), vertebral and spinal chape changes. Osteophyte formation
Magnetic resonance imaging (MRI)	Signal intensity, NP area, and other human clinical correlates
Pain‐like behaviors	Mechanical and thermal hypersensitivity, grip strength, spontaneous and stress‐related behaviors, and other relevant measures
Functional changes (biomechanics)	Biomechanical properties of motion segment and IVD components
Compositional measures	Immunohistochemistry (IHC), gene and protein measurements

Puncture models of IVDD have significantly advanced spine research by reproducing a spectrum of IVDD phenotypes, from chronic degeneration to acute herniation and inflammation‐induced pain. Marking 20 years since the introduction of modern puncture models, their use has become the predominant method for inducing IVDD, with an exponential growth in related studies. These injury models are applied across many species and involve various injury types mimicking distinct phenotypes, including slow progressive IVDD, chronic degeneration, severe acute IVDD, herniation, and Modic‐like changes. They also allow investigation of the interactions between the degenerating IVD, surrounding tissues, and the nervous system. To better interpret the broad literature spanning many injury types grouped as “puncture” models, we recommend more detailed and standardized reporting of key methodological factors and model objectives as described in this paper.

In conclusion, puncture models of IVDD remain powerful tools for modeling distinct IVDD phenotypes. However, the primary limitations of puncture models include their acute, surgically induced inflammatory nature that differs from the gradual degeneration often seen clinically/careful selection and transparent reporting of puncture parameters aligned with specific research questions will enhance reproducibility and translational relevance. As the IVD puncture model literature expands, we believe that adopting standardized injury method reporting parameters and IVDD phenotype being modeled will maximize their utility contextualizing studies in a broader literature understanding IVDD pathogenesis and therapeutics.

Author Contributions

Conceptualization: C.J. and J.C.I. Resource acquisition: C.J., J.J.H., and J.C.I. Literature reviews: C.J., J.J.H., and J.C.I. Results and figures: C.J., J.J.H., and J.C.I. Interpretations and edits: C.J., J.J.H., Y.L., S.C., A.C.H., A.L., K.M., J.K., and J.C.I. Funding acquisition: J.C.I. Project administration: J.C.I. Supervision: J.C.I. Writing – original draft: C.J., J.J.H., and J.C.I. Writing – review, editing, and approval: C.J., J.J.H., Y.L., S.C., A.C.H., A.L., K.M., J.K., and J.C.I.

Conflicts of Interest

The authors declare no conflicts of interest.

Jain C., Huang J. J., Lee Y., et al., “Animal Models of Disc Degeneration Using Puncture Injury: A 20 Year Perspective,” JOR Spine 8, no. 3 (2025): e70093, 10.1002/jsp2.70093.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.