Low back pain (LBP) is a leading cause of disability worldwide, affecting up to 80% of individuals at some point in their lives and imposing a significant socioeconomic burden (1). Among the primary causes of LBP, intervertebral disc degeneration (IDD) accounts for approximately 40% of cases (2, 3).

The intervertebral disc (IVD) is an avascular structure composed of three main components: (1) nucleus pulposus (NP), a central, gel-like region rich in proteoglycans (PGs), type II collagen (COLL II), elastin, and glycoproteins, responsible for load distribution and maintaining disc hydration; (2) annulus fibrosus (AF), a dense outer ring primarily made of type I collagen (COLL I), providing tensile strength and structural support; and (3) cartilage endplates (CEPs), a thin layers of hyaline cartilage that separate the disc from adjacent vertebral bodies and mediate nutrient and metabolite exchange (4–6).

The onset and progression of IDD are multifactorial, involving genetic predisposition, environmental factors, mechanical stress, aging, smoking, obesity, atherosclerosis, poor nutrition, and metabolic disorders (7, 8). A hallmark of IDD is the disruption of extracellular matrix (ECM) homeostasis, characterized by increased catabolic activity, driven by matrix metalloproteinases (MMPs), a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) enzymes, and pro-inflammatory cytokines, such as interleukins (IL)1β, IL6, IL8, prostaglandin E2 (PGE2), and nitric oxide (NO), coupled with reduced anabolic signaling from factors like tumor necrosis factor β (TGFβ), bone morphogenetic proteins (BMPs), and insulin-like growth factor (IGF). This imbalance leads to PG depletion and water loss in the NP, followed by structural disruption of the AF, fissure formation, neovascularization, neoinnervation, and progressive cell apoptosis (9, 10). Recent studies have highlighted also the critical role of CEP degeneration in IDD progression. In fact, it was reported that increased calcification impairs the CEP’s ability to regulate nutrient diffusion and metabolic waste removal, contributing to local hypoxia, lactic acid accumulation, and potentially initiating the degenerative cascade (11, 12).

Diagnosis of IDD typically involves clinical evaluation supported by imaging, with magnetic resonance imaging (MRI) being the most sensitive non-invasive method for assessing disc morphology and hydration status (13). Among the MRI-based grading systems, the Pfirrmann classification focuses on NP and AF characteristics (14), while the Modic classification assesses signal changes in the CEP and adjacent vertebral bone marrow (15). However, MRI offers limited resolution in detecting early or subtle histopathological change, highlighting the mandatory complementary role of histological analysis, which provides high-resolution insights into cellular and ECM alterations underlying IDD (16, 17). Despite important contributions from animal models, the limited availability of human histological data continues to challenge the accurate characterization of disc degeneration and the development of targeted therapies.

The aim of this systematic review is to provide a comprehensively analysis of histological findings in human IVDs reported over the past decade. Specifically, we include studies investigating histological and immunohistochemical features associated with IVDs degeneration in human specimens, encompassing different tissues and stages of disease. When available, comparisons with non-degenerated (healthy) discs were also considered. To our knowledge, this is the first systematic review to focus exclusively on human specimens spanning various stages of degeneration.

This systematic review was structured according to the PEO framework (Population, Exposure, Outcome). Eligible studies included those involving human IVDs of any degeneration severity (Population), with disc degeneration or degeneration grade as the exposure of interest, and histological and immunohistochemical findings as outcomes (Outcome). Studies were included regardless of the presence of a healthy control group (healthy IVDs), and both comparative and non-comparative study designs were considered.

On May 2025, a search was conducted in 3 databases (PubMed, Web of Science and Scopus) using the following string across all databases: (human intervertebral disc OR human intervertebral disc degeneration) AND (histology). The following filters were applied: a date ranges from 2015-05-01 to 2025-05-01, for PubMed and Web of Science, and 2015–2025 data range for Scopus.

The study selection process was illustrated using a flowchart, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, as shown in Figure 1.

Duplicate removal was performed using Rayyan, after which the remaining articles were screened based on their titles and abstracts according to the PEOinclusion criteria.

The following types of studies were excluded: preclinical studies, reviews, and studies focusing on no histological outcomes, involving IVD after pharmacological or physical treatment, and/or with not degenerated IVD. Subsequently, the full texts of articles were reviewed, in cases where abstracts did not provide sufficient information, and they were assessed using the inclusion and exclusion criteria. The article selection process was independently conducted by two authors (FV, FS), with disagreement on study eligibility resolved by a third author (GG).

Relevant data were independently extracted by two authors (FV, FS) and recorded in a standardized extraction form. The data collected included MRI-based and histological grading systems (Table 1), methodological overview of histological and immunohistochemical assessments (Table 2), experimental groups, diagnosis, grading scores, histology, results, and reference (Table 3).

The methodological quality and risk of bias of the included studies were assessed using the QUADAS-2 tool (Quality Assessment of Diagnostic Accuracy Studies-2; Table 4) (18). This tool evaluates four key domains: patient selection, index test, reference standard, and flow and timing. Each domain is assessed for risk of bias. The assessment was performed independently by two authors (FV and FS), and any discrepancies were resolved by consensus or through consultation with a third author (GG).

A total of 515 studies were identified (55 from PubMed, 135 from Web of Science, and 325 from Scopus). After removing 80 duplicates, 435 studies remained. Following abstract screening, 50 studies were excluded, and an additional 340 were excluded after full-text review due to being reviews (n = 12) or having methodological limitations (in vitro/in vivo models, lack of histological outcomes, or non-degenerated IVDs, pharmacological or physical treatment interventions, n = 328).

Ultimately, 45 studies were included in the systematic review (Table 3; Figure 1) (11, 19–62).

This systematic review underscores the histological alterations observed in human IVDs across different stages of degeneration, providing a comprehensive reference framework that may support the identification of stage-specific biomarkers, guiding the development of targeted therapeutic strategies. While prior research has extensively utilized histological analyses on animal models to explore and analyze IDD degeneration (67, 68), this is the first systematic review that focus on human samples. Specifically, our review focused on three anatomical components, i.e., IVD, NP, and CEP, revealing specific trends in structural disruptions and molecular alterations.

Across the 45 included studies, a consistent pattern of histological changes was observed in degenerated IVDs. These included ECM disorganization, GAG and PG depletion, increased fibrosis, annular tears, cellular clustering, and signs of neovascularization and neoinnervation, particularly in the AF and CEP. These structural alterations reflect a progressive breakdown of the biomechanical and biochemical integrity of the IVD. Cell clustering, for instance, is considered a hallmark of degeneration and has been associated with attempts at cell proliferation in response to matrix degradation, although such attempts appear largely ineffective in halting the degenerative cascade (69). The demographic analysis revealed a clear age-related distinction between the non-degenerated and degenerated groups, with the former showing a significantly younger mean age compared to the latter (p < 0.001). This finding aligns with the well-established association between aging and IVD degeneration. Both groups exhibited a relatively balanced gender distribution. The origin of the samples also differed notably: non-degenerated discs were primarily obtained from cadavers or patients with conditions unrelated to degeneration, such as trauma, or fractures. Differently, degenerated samples were derived from patients with clinically diagnosed degenerative conditions, including disc herniation, chronic LBP, spondylolisthesis, or stenosis. These distinctions support the validity of sample classification and highlight the relevance of age and pathology in the progression of disc degeneration.

The molecular investigations performed on histological samples highlighted upregulation of catabolic enzymes (e.g., MMPs, ADAMTS4/5), inflammatory mediators (e.g., TNFα, IL-1β, IL-6), apoptotic and necroptotic regulators (e.g., Caspases and RIP3/MLKL), and neurovascular signaling factors (e.g., Sema3, ANG2, CXCR4). Conversely, anabolic, proliferation, and antioxidative mediators, such as SOX9, Ki67, PCNA, COL II, and NRF2, were markedly downregulated, especially with advanced degeneration. Severity-dependent associations suggested that molecular dysregulations intensify progressively as discs degenerate.

A central theme is the shift toward catabolism, driven by elevated protease activity (MMPs, ADAMTS) and inflammatory cytokines (ILs and TNFα). Impaired ECM integrity and loss of PG facilitate dehydration and compromise mechanical resilience. Downregulation of anabolic molecules such as SOX9, COL II, and ACAN further exacerbates structural decline. CEP calcification and diminished YAP1 likely restrict nutrient transport, aggravating hypoxic stress and undermining cellular viability. The interplay of inflammation and proteolysis underpins a self-reinforcing degenerative cycle that is evidently pronounced in advanced disease. Overexpression of catabolic and inflammatory factors is consistent with previous evidence indicating that inflammation is both a driver and a consequence of tissue breakdown (70, 71). Increased breakdown of ECM components is not countered by sufficient regenerative or anti-inflammatory signaling, leading to continued deterioration of disc structure and function (30).

Markers such as Caspase-3, −1, GSDMD, Cathepsin B, and RIP3/MLKL testify to simultaneous apoptosis, pyroptosis, and necroptosis within degenerated discs and elevated senescence-associated markers, including p16 and Defensin Beta 1, indicate accumulation of dysfunctional cells. These cell loss and senescence processes not only reduce functional cell pools but also contribute to pro-inflammatory secretomes and tissue dysfunction (72).

Upregulation of angiogenic mediators (ANG2, ANGPTL8, Sema3, NRP, CXCR4, SDF1, TIE2) and macrophage phenotypic markers (CCR7, CD163, CD206) reveals a marked gain in vascular and neural presence within degenerated disc layers. This aberrant innervation and neovascularization likely facilitate nociception and inflammation, linking structural breakdown with LBP symptoms (73).

Oxidative stress markers (e.g., NOX4) rose dramatically, in parallel with the diminishment of defense enzymes like PON1 and transcription factors such as NRF2. Depletion of mechanosensors (PIEZO1, AQP channels) and signaling regulators (WNT5a, VDR, EZH2, IRF2) suggests disruption of disc homeostasis, mechanosensitivity, and epigenetic equilibrium. These changes may impair cell physiology and response to mechanical load, accelerating degeneration (74).

Despite anatomical heterogeneity, IVD, NP, and CEP shared core degenerative patterns. However, CEP tissue showed a more pronounced inflammatory profile (e.g., COX2/PGE2) and mechanotransducive impairment (YAP1 reduction), likely linked to its unique nutrient-exchange role. CEP alterations may precipitate degeneration by compromising cellular nutrition, which subsequently amplifies NP/AF degradation. This layered evidence supports a model where CEP dysfunction initiates degeneration that cascades inward (75).

The PPI network analyses presented in this study provide a systems-level perspective on the molecular alterations occurring within distinct compartments of the degenerated IVD. Notably, our findings underscore the compartment-specific yet interrelated nature of the degenerative processes.

In the whole IVD, the emergence of a densely interconnected network, comprising MMPs, collagen isoforms, and signaling mediators such as SMADs and FOXO, reflects a coordinated activation of ECM remodeling pathways. The enrichment of biological processes related to collagen catabolism, cartilage development, and BMP/VEGF signaling aligns with the known hallmarks of IVD degeneration, including matrix degradation, neovascularization, and altered mechanotransduction.

Within the NP, the PPI network highlighted a dominant inflammatory signature, marked by strong interactions among cytokines (e.g., TNF, IL6) and their receptors (e.g., CXCR4), as well as ECM-modifying enzymes such as MMP3. GO enrichment further supported this proinflammatory milieu, pointing to sustained cytokine signaling, ECM disassembly, and angiogenic activity. These findings are consistent with prior studies indicating that chronic inflammation contributes to matrix breakdown and impairs regenerative potential in the NP.

Conversely, the CEP displayed a unique shift toward an osteogenic and proinflammatory phenotype, with prominent interactions involving prostaglandin-related enzymes (e.g., PTGS2, EP4) and osteogenic transcription factors (e.g., OCN). GO terms associated with ossification, prostaglandin biosynthesis, and regulation of stem cell differentiation suggest that CEP degeneration may involve endochondral-like ossification processes, potentially impairing nutrient diffusion and exacerbating disc pathology.

Together, these data reinforce the concept that IVD degeneration is not a uniform process, but rather involves spatially defined molecular responses that converge toward a dysfunctional and inflammatory tissue state. Understanding these regional differences is critical for the development of targeted therapies aimed at restoring disc homeostasis and halting disease progression.

One of the challenges highlighted by this review is the methodological heterogeneity across studies. A variety of histological grading systems were employed, including the Sive et al. (35, 38, 43, 52) and Ritges et al. (36) systems, and, similarly imaging assessments relied predominantly on the Pfirrmann classification (11, 19–24, 29, 30, 33, 34, 39–42, 44–51, 53–57, 60–62) although the Thompson et al. (25, 26, 31, 54, 58) and Schneiderman et al. (37) systems were also used. This lack of standardization complicates direct comparison of findings and underscores the need for more uniform criteria in future research. Most studies used hematoxylin and eosin staining as a primary diagnostic staining (20, 28, 32, 34, 38, 41, 43, 44, 46, 50, 52, 53, 55, 57–60, 62), often complemented by Safranin O/Fast Green (22, 25, 26, 31, 32, 41, 55, 57–60, 62) or Alcian Blue (46, 53, 58–60) for matrix evaluation, and IHC for protein quantification in all studies.

Beyond the biological findings, this systematic review provides a detailed methodological mapping of human intervertebral disc histology studies, which represents a key element of its originality. In contrast to previous reviews that primarily focused on molecular pathways or experimental models, the present work systematically integrates information on tissue source, anatomical localization, histological processing, immunohistochemical methodologies, and clinical characteristics of the studied populations (74–76).

Considerable heterogeneity emerged in antibody selection, staining protocols, and quantification approaches across studies. While most investigations relied on qualitative or semi-quantitative immunohistochemical assessments, only a limited number employed standardized scoring systems or digital image analysis for signal quantification. Differences were also observed in data representation, with results variably reported as percentage of positive cells, staining intensity, or descriptive localization, limiting direct cross-study comparisons. Anatomical localization represented an additional source of variability. Most studies focused on lumbar intervertebral discs, whereas cervical specimens were less frequently analyzed, often in distinct clinical contexts. Moreover, biopsy origin within the disc (whole IVD, nucleus pulposus, annulus fibrosus, or cartilage endplate) differed substantially among studies, influencing both histological appearance and molecular readouts.

Clinical heterogeneity further contributed to variability, as degenerated samples were obtained from patients with diverse diagnoses, including disc herniation, chronic low back pain, spinal stenosis, and spondylolisthesis, while non-degenerated controls were frequently derived from cadaveric donors or patients undergoing surgery for trauma or deformity. Importantly, the timing of tissue collection relative to imaging assessment or symptom duration was inconsistently reported, representing a critical limitation for the interpretation of degeneration-stage–specific findings.

By systematically capturing and synthesizing these methodological dimensions, the present review provides a structured framework that enables more informed interpretation of histological data and highlights key parameters that should be standardized in future human disc studies to improve reproducibility and translational relevance.

Despite variability in methodological approaches, the consistency in reported histological features across studies and tissue types strengthens the reliability of the findings. Additionally, the risk of bias assessment revealed overall moderate risk of bias across several domains. Most studies presented a low risk in patient selection, especially those using surgical samples with well-defined diagnostic criteria. Nevertheless, a number of studies lacked transparency regarding blinding procedures during histological and IHC analysis, introducing a potential risk of bias in the “index test” domain. Moreover, the absence of standardized histological grading systems in some studies, along with variability in sample preservation and processing protocols, may have affected internal validity and reproducibility. Additionally, flow and timing concerns emerged when the interval between MRI-based grading and histological analysis was not clearly reported. Despite these limitations, the consistency of the reported degenerative markers across independent cohorts reinforces the validity of the main conclusions.

The findings from this review have several clinical implications. The identification of key molecular actors offers potential biomarkers and therapeutic targets. Elevated MMPs, pro-inflammatory cytokines, and angiogenic factors could serve as stage-specific degeneration indicators, potentially refine diagnostic accuracy beyond MRI. Therapeutically, strategies aimed at restoring anabolic pathways (e.g., SOX9, COL II), enhancing antioxidative capacity (e.g., NRF2 modulation), and inhibiting catabolic/inflammatory signaling (e.g., targeting ADAMTS, MMPs, IL-1β/TNFα or Sema3) warrant exploration. Preserving CEP integrity and transport function (e.g., reducing calcification, upregulating YAP1) may offer early intervention avenues.

Several are the strength of the present systematic review. Analyzing the histology of the human IVD provides direct insight into the cellular and structural changes associated with IVD degeneration in its physiological biological context and offers distinct advantages over relying exclusively on animal models, particularly in translating findings to clinical practice. Human histological studies capture the authentic tissue microenvironment, including native inflammatory responses, ECM remodeling, and cellular phenotypes within the context of age-related degeneration. In contrast, animal models, though useful for mechanistic and interventional experiments, present significant species differences in spinal anatomy, disc composition, mechanical loading, and immune response, which can limit the extrapolation of results to humans (76). Additionally, quadrupedal biomechanics in most animal models differ markedly from the bipedal forces experienced by the human lumbar spine, making it challenging to replicate physiological loading conditions accurately (77). Human tissue analysis also circumvents ethical concerns, lifespan limitations, and standardization issues inherent in animal research. Ultimately, direct histological evaluation of human IVDs not only complements preclinical data but also ensures greater clinical relevance in understanding degenerative disc disease.

However, some limitations must be acknowledged. First, most of the included studies were cross-sectional in nature, limiting causal inferences about the sequence of degenerative changes. Second, variations in tissue procurement, processing, and staining techniques could introduce bias or artifacts. Moreover, healthy control samples were often derived from cadaveric donors or patients undergoing surgery for unrelated conditions (e.g., trauma), which may not perfectly represent truly healthy discs. Additionally, many studies did not report complete clinical information about the donors, such as comorbidities, smoking status, or physical activity levels, factors known to influence disc health. The heterogeneity in patient age across groups may also confound interpretation, as age-related changes may overlap with pathological degeneration.

This systematic review consolidates current histological and molecular evidence from human IVD samples, highlighting the progressive and compartment-specific nature of degenerative changes. By integrating structural, cellular, and protein-level alterations, our findings reveal a consistent pattern of matrix disorganization, inflammation, vascularization, and cell death across degenerated discs, with region-specific molecular signatures in the NP, and CEP. The observed upregulation of catabolic enzymes, inflammatory cytokines, and pro-angiogenic factors, alongside the downregulation of anabolic, antioxidative, and mechanosensitive pathways, underscores a multifactorial degenerative cascade that worsens with disease severity.

Despite methodological heterogeneity, the convergence of histological markers across independent studies strengthens the validity of these findings and supports their translational relevance. Importantly, this review emphasizes the unique value of human tissue analysis over animal models, offering clinically meaningful insights into disc biology in the context of aging and pathology.

The identification of key degenerative mediators offers promising opportunities for biomarker discovery and targeted therapeutic development. Interventions aimed at modulating inflammation, enhancing matrix synthesis, preserving CEP function, and counteracting oxidative stress may represent viable strategies for delaying or reversing disc degeneration. Future research should prioritize longitudinal human studies, standardize histological grading protocols, incorporate comprehensive clinical metadata, and integrate histological findings with functional and mechanistic experiments to validate the biological relevance of the identified pathways and support their translational application in degenerative disc disease.