The first spinal surgery using autograft bone was done by Hibbs in 1911. Three types of autograft bone have been used since then: cancellous, cortical, and vascularized cortical [4]. Autograft bone has all properties for an ideal spine graft material: osteoinduction (bone morphogenic proteins (BMPs), particularly BMP2 and BMP7), osteoconduction (bone matrix/collagen), and most importantly osteogenic potential (stem cells and osteoblasts). Accordingly, autografts can provide immediate and long-term mechanical stability. Autografts used for spine fusion can be classified in two categories: local bone, which is harvested from the lamina, facets, or processes during decompression, and extraspinal material, most commonly an iliac crest bone graft (ICBG). With both grafts there are no issues with donor compatibility and the costs are low. Local bone is a cortical graft and provides immediate mechanical stability, but due to the pore size, cell migration and differentiation is impaired. This leads to lower rates of bone remodeling and long-term instability. The advantage of local bone over ICBG is that no extra procedure or harvest site is needed, causing fewer complications. In contrast, ICBG is the most commonly used graft and often referred to as a “gold standard” for spinal fusion. A purely cancellous structure is easily revascularized, and the large surface provides an ideal environment for bone formation. Although ICBG lacks compressive strength, the rapid bone formation leads to an increase in fusion mass that provides mechanical stability. If the fusion approach is posterior, ICBG can be harvested without an additional surgical incision. An important parameter when choosing the autologous bone graft is the patient’s age, as elderly patients will have less ICBG and a lower bone quality. A challenge with both types of autografts is the large amount of material needed for multilevel fusions. Sengupta and co-workers found that the ICBG was superior in overall fusion rates and that both grafts performed similarly for single-level fusions. In multilevel fusions ICBG outperformed the local bone graft (fusion rates of 66 % vs. 20 %) [5]. However, ICBG grafts carry more complications than local bone grafts. The most serious complications are related to the harvest and can lead to subsequent fractures, hernia, ureteral injury, instability, infection, and prolonged length of stay [6]. Even though postoperative pain at the donor site is one of the most common issues, Howard et al. found that the pain incidence was similar in patients with or without ICBG harvest [7]. A study done by Gruskay found that an increase in blood transfusion rates, surgical time, and length of stay were the only short-term complications associated with the use of ICBG as a graft material [8].

This bone graft is harvested from cadaver tissues and is usually used for anterior cervical and lumbar fusions. Allografts are depleted of cells and growth factors and therefore have osteoconductive and minor osteoinductive properties. Allograft bones can be fresh, fresh frozen, or freeze-dried. The mechanical properties of fresh-frozen allografts are superior to freeze-dried ones, but have higher immunogenicity. In the posterolateral approach, freeze-dried allografts failed to produce fusion, whereas ICBG led to an 80 % fusion rate [9]. Cortical allografts can provide immediate stabilization, but the remodeling process is slow and the resorption of the graft is increased. On the other hand, corticocancellous grafts do not have an immediate mechanical benefit, but due to the large contact area, they are easily integrated and can promote bone remodeling. One of the main health concerns with allografts is the transmission of diseases. Mroz reported that between 1994 and 2007, 96.5 % of musculoskeletal allografts were recalled due to contamination and recipient infection [10]. Furthermore, a study done by Jurgensmeier found frequent inconsistencies in tissue banks, with 39 % accepting samples from elderly donors (≤80 years of age) and only 50 % of banks excluding grafts from osteoporotic patients [11]. Those discrepancies in graft collection have a significant impact on the mechanical properties of the graft and can contribute to fusion failure. However, the use of allografts in surgeries avoids complications associated with ICBG harvest. Allografts are manufactured in various forms as strips, chips, or demineralized bone matrix.

Ceramics are osteoconductive grafts deficient in growth factors and cells. They are easily obtainable in large amounts, are disease-free, and are with a pore size that is suitable for cell and blood vessel ingrowth. Hydroxyapatite and β-tricalcium phosphate (β-TCP) are mainly used for spine surgeries due to their prolonged rate of resorption, up to a year for hydroxyapatite and several months for β-TCP. Furthermore, the structure and pore size of hydroxyapatite and β-TCP are very similar to cancellous bone. Once implanted, β-TCP is populated with a fibrovascular stroma that is soon replaced with osteogenic cells, leading to bone formation. Several clinical studies have found that both ceramic grafts induced similar fusion rates compared to patients who received ICBG [20, 21]. One of the most commonly used hydroxyapatites is Pro Osteon Coralline Hydroxyapatite (Interpore Cross International, Irvine, CA). It is derived from a sea coral and mainly consists of calcium carbonate. Another graft mixture is Collagraft (NeuColl, Inc., Palo Alto, CA), consisting of collagen and a mixture of hydroxyapatite and β-TCP closely resembling the natural bone structure. Silicate-substituted calcium phosphate and calcium sulfate have been used in the recent years as synthetic graft extenders, showing variable fusion rates. The general disadvantage of ceramic grafts is its brittleness and inability to sustain heavy loads. Due to the low mechanical stability, ceramics are usually combined with fixation instruments. Because ceramics lack osteogenic and osteoinductive properties, they are often combined with materials containing growth factors and cells.