Standard anterior cervical decompression and fusion (ACDF) is regarded as the gold standard for the treatment of cervical radiculopathy and myelopathy [6,10,17]. One of the important factors affecting the overall outcome is the restoration of the intervertebral disc height, which is achieved via anterior cervical distraction, and placement of an appropriately sized intervertebral graft [2,3]. However, the decision of the ideal graft height for successful anterior reconstruction of the cervical spine remains controversial, and the selection of an improperly sized graft might adversely affect the clinical outcome.

While a large graft may distract the anterior column, thereby indirectly decompressing the neural elements through indirect increase of the foraminal height and unbuckling of the ligamentum flavum [1], excessive anterior distraction may decrease the load transmission in the posterior column, thereby subjecting the anterior graft, vertebral bodies, and attached muscular structures (e.g., longus colli) to excessive loads [15]. Over-distraction could be associated with facet joint injury or subsequent postoperative neck pain [8,9], although few studies have reported this problem [11].

In the current report, the authors report on two patients who presented with prolonged cervicalgia after initial ACDF or artificial disc replacement (ADR) using over-sized implants. In both cases, the pain was disregarded as ‘neuropathic pain of unknown origin’ because the original surgeon considered achievement of radiographic solid fusion and a well-maintained ADR construct as indicators that the source of pain had been eliminated.

Although ACDF has become a well-established surgical technique, it is associated with some commonly reported postoperative complications such as graft collapse, non-union, and adjacent segment degeneration [7,12,13]. However, once rigid fusion through ACDF has been achieved, as corroborated by various radiological examinations, and fusion-related complications are absent, there would be not many debates among the spine surgeons upon the idea that the original pain sources have been eliminated and the residual symptom might be related to the elongated nerve root or distracted facet joint due to the restoration of the disc height by the cage insertion.

In contrast to the commonly held view based mainly on these theoretical considerations, in the cases described here, the patients experienced worsening of cervicalgia in spite of achieving solid fusion and a well-maintained ADR construct. Moreover, these symptoms were not relieved even upon elimination of the other possible sources of pain, such as elongated nerve root or distracted facet joint. Upon observing the patients carefully over the course of several months before arriving at the decision of performing revision surgery, the authors noted that the patients were unable to tolerate the limitation on neck movement due to pain provocation, especially during flexion. Therefore, the main hypothesis and rationale underlying the decision of revision surgery was that the ‘patient cannot tolerate the elongated state itself’ with excessive disc height distraction.

Typically, grafts during the ACDF procedures are sized by interference fit at the time of surgery or based on preoperative radiographic templating, and any excessive anterior distraction may proportionally decrease posterior column load transmission, thereby subjecting the anterior graft and vertebral bodies to excessive loads [15,16]. Accordingly, with larger grafts, cervical muscular tone would likely increase in both distractive and compressive forces, and this increased postoperative muscular compressive load would be increasingly transmitted anteriorly through the graft [15]. However, it is common for surgeons to be more interested in attaining solid fusion with a firmly fitted graft, neglecting the dynamic factors such as the effect on the cervical musculature and neck motion of the excessive distractive forces required for the insertion of an enlarged cage or its maintenance after fusion.

It is difficult to obviously delineate a specific boundary for an optimized disc height restoration using an ideally sized graft during ACDF or ADR. Actually, there is a paucity of literature addressing this issue since most of these procedures are routinely performed by an arbitrary selection of graft height based on the preoperative radiographic data or on the intraoperative redundancy created after decompression. Chang et al. [5] have reported about the relationship between the increased intervertebral disc height and development of postoperative axial neck pain after anterior cervical fusion based on the outcomes from their 155 patient series. Through their average increase of disc height by 2.62 mm, 55 patients (35.4%) of their series experienced initial posterior neck pain after ACDF. Although they have concluded that there is no significant relationship was observed between the radiological evaluation results regarding the increase in the intervertebral space and clinical findings since most of these neck pain improved after operation and were relieved during the follow-up period, still 21 patients have suffered from persistent axial neck pain refractory to various conservative treatments even up to a year after initial surgery. Another report from Ha et al. [9] has attributed the exaggerated intraoperative vertebral distraction to the development of transient postoperative axial neck pain by measuring the maximum torque applied on the inserted retractor. Among their 24 consecutive series treated with single level ACDF, either applied with high- (>6 kgf·cm) or low-torque (<6 kgf·cm) distractive force before insertion of graft, they reported that only the high-torque group have featured with the immediate postoperative neck pain development. However, all of these experiences of pain were transient not lasting more than 5 days and they could not draw any conclusion regarding the difference of disc height increase in number (average increase of 2.1 mm) between the low torque group and the high torque group nor significant relationships among the intervertebral disc height, visual analogue scales, neck disability index scores and torque.

Additionally, several studies have been conducted to evaluate the appropriate graft height in anterior cervical surgery. An et al. [1] performed a cadaver radiographic study and determined that using discs with a preoperative height between 3.5 and 6 mm, the graft height would be elongated by 2 mm postoperatively. Brower et al. [4] retrospectively reviewed ACDF data for 59 patients and found a trend toward nonunion when the preoperative disc height was distracted by more than 4 mm. One study used a finite-element model for evaluating the optimal graft size [14]. In contrast, there are other reports describing that the preoperative disc height was not an accurate predictor of graft loads in ex vivo model experiments [16].

Most of surgeons would simply regard that an overdistraction by inserting an excessive graft material is generally considered to yield postoperative neck pain by inducing a concomitant distraction on the posterior facet joint or spasm of the posterior neck muscle. However, as there has been no previous study defining the relationship between graft size and postoperative axial neck pain conducted, the underlying mechanism of posterior neck pain has not been fully elucidated. Simply other than muscle spasm or the excessive distraction of the facet joint, the instability of the vertebral segment can also lead to posterior neck pain. Olsewski et al. [15] have assumed a hypothesis that the appropriate amount of cervical distraction that might affect clinical outcome has not been well defined in the literature and performed a biomechanical analysis using a Smith-Robinson type cervical spine model. Forces across the posterior elements and graft site were measured, during the flexion loading, and compared as the disc space was distracted. Their curious finding was that the ratio of posterior element load to the graft load with increasing disc space distraction significantly decreased due to the significant decrease of posterior element load after the same distraction (from 46.1±22.0 to 18.7±9.7 N/Nm). These findings were more profoundly significant during the spondylotic specimens (original disc height of 4-5 mm) distractions in excess of 3.0 mm from preoperative height, which might help explaining a limit of effective disc space distraction. Therefore, simply regarding the posterior element including facet joint as a source for postoperative axial neck pain after over distraction (like more than 3 mm) would not be proper and prudentially consider the unimaginable painful load that has been transferred over the inserted, but still unstable, graft.

To sum up, the main learning from these two cases that the authors want to reinforce is not that neck pain can be caused by distracted anatomical structures, but the fact that a solidly fused, improper sized cage can place excessive pressure on the cervical musculature, resulting in restriction of neck motion. This indicates that retaining the original physique as well as disc height is sometimes more beneficial for the patient, instead of reconstruction or reduction to achieve normalcy.

The ideal graft height for successful anterior reconstruction of the cervical spine remains unknown and measurements of the preoperative disc height may not be useful predictors of subsequent graft forces. The increased distraction force required to insert and maintain these improper sized grafts results in increased stress and pressure both on the graft and on the cervical neck musculature, leading to intractable cervicalgia even after solid fusion. Surgeons should pay careful attention to these dynamic factors and should not disregard the pain as neuropathic or idiopathic by relying on the verified solid fusion.