Structural changes occurring at the surgical site long after hemilaminectomy in small-breed dogs with type I thoracolumbar intervertebral disk herniation Authors: Yusuke SAKAGUCHI, Hidetaka NISHIDA, Hiroshi TANAKA, Masahiko KITAMURA, Masanari NAKAYAMA Published: 2026 ABSTRACT Hemilaminectomy is the most common surgical procedure for type I thoracolumbar intervertebral disk herniation (TL-IVDH). Despite the clinical relevance of postoperative morphological assessment to predict postoperative complications, only a few studies have evaluated long-term changes at the surgical site, such as those affecting the spinal cord and vertebrae, after hemilaminectomy for type I TL-IVDH. This retrospective study describes postoperative changes at the surgical site, including disk height, in dogs that had undergone hemilaminectomy and subsequently experienced late recurrence. Nineteen dogs were eligible for inclusion in this study. The disk height index (DHI) at the initial affected site was significantly reduced compared to the time of initial onset. The maximal spinal cord compression rate at the initial affected site at the time of recurrence decreased significantly compared to that at the time of initial onset. Two dogs had bone regrowth at the site of first surgery at the time of recurrence. No recurrence of disk herniation or spinal cord injury caused by bone regrowth was observed at the surgical site. In conclusion, structural changes such as decreased DHI and bone regrowth were observed at the surgical site a long time after hemilaminectomy, but these changes were not the direct cause of the neurologic signs observed at recurrence. INTRODUCTION Thoracolumbar intervertebral disk herniation (TL-IVDH) is a common neurologic disorder in chondrodystrophic dogs including Dachshunds [2, 16]. It is classified into Hansen type I and type II [8]. Type I TL-IVDH results in spinal cord injury by extrusion of disk materials into the vertebral canal, causing a variety of neurologic signs ranging from mild discomfort with no neurologic deficits to paralysis with loss of pain perception [7, 10]. Surgical treatment for type I TL-IVDH involves removal of extruded disk materials in the vertebral canal [4, 15]. Several surgical procedures have been described, including hemilaminectomy, mini-hemilaminectomy, and partial lateral corpectomy [4, 15]. Of these, hemilaminectomy is the most common procedure. A previous study of dogs with TL-IVDH treated with hemilaminectomy or mini-hemilaminectomy, followed by assessment of the surgical site using MRI, found residual spinal cord compression at the site of initial surgery in 52% of dogs [17]. Also, vertebral new bone formation was observed in 19% [17]. Because these postoperative changes may contribute to clinical signs, it is important to recognize them. In human medicine, intervertebral disk degeneration is sometimes assessed after spinal surgery because its progression may contribute to the development of spinal instability [11, 18]. One method for evaluating disk degeneration is the disk height index (DHI) reported by Masuda et al. [13]. As disk degeneration progresses, the DHI decreases [13]. DHI has been widely used in experimental animal models, including dogs [13, 24], and has also been applied in human medicine [11, 25]. In veterinary medicine, DHI has also been used to assess the pathophysiology of type I TL-IVDH in dogs [19]. Despite the clinical relevance of postoperative morphological assessment to predict postoperative complications, only a few studies have evaluated long-term changes at the surgical site, such as those affecting the spinal cord and vertebrae, after hemilaminectomy for type I TL-IVDH. In addition, we found no studies that assessed the postoperative changes in the disk height. We describe postoperative changes at the surgical site, including intervertebral disk, vertebral body, and spinal cord, using CT in dogs that had undergone hemilaminectomy and subsequently suffered from late recurrence. MATERIALS AND METHODS Criteria for inclusion The medical records of dogs that were diagnosed with type I TL-IVDH by CT, CT-myelography and hemilaminectomy at Nakayama Veterinary Hospital between January 2015 and December 2023 were reviewed retrospectively. Dogs were included that exhibited improved neurologic signs postoperatively but deteriorated within the selected period and were diagnosed with recurrent type I TL-IVDH after diagnostic imaging and surgery. This choice was because these dogs let us investigate the structural changes that occurred at the surgical site following hemilaminectomy. Dogs with recurrence within 6 weeks postoperatively were excluded as early recurrence cases [12, 20]. Dogs with an affected site extending more than 2 vertebral spaces at the time of initial onset were also excluded. The following data were recorded: breed, sex, age and body weight at the time of initial onset. We also logged the grade of neurologic severity at the time of initial onset, after 1st surgery, at the time of recurrence, and after 2nd surgery, and also the interval from 1st surgery to improvement of the neurologic signs, the interval from initial onset to recurrence, and the affected site at the time of initial onset and recurrence. CT images and CT-myelographic images were analyzed. In addition, CT images were used to assess the presence of spondylosis deformans at the initial affected site at the time of initial onset and recurrence, and to assess the presence of bone regrowth at the site of 1st surgery at the time of recurrence. Preoperative examination and surgery A neurologic examination was performed on all dogs. Based on a previous study, all dogs were classified into five categories: Grade 1, back pain only; Grade 2, ambulatory paraparesis; Grade 3, non-ambulatory paraparesis; Grade 4, paraplegia with intact deep pain sensation; Grade 5, paraplegia with no deep pain sensation [21]. Dogs with normal gait were assigned Grade 0. After the neurologic examination, CT (Activion 4; Toshiba Medical Systems Corp., Ohtawara, Japan) and myelography were performed in all dogs to assess the affected site. CT was performed with a scan thickness of 0.5 mm, 150 mA, 120 kV, and was reconstructed at a slice thickness of 1 mm using a bone algorithm with a window width of 2,000 HU and window level 600 HU [19]. Myelography was performed by lumbar puncture using iohexol 240 mgI/mL (0.45 mL/kg). After identifying the spinal cord compression site by CT-myelography, a hemilaminectomy was performed to remove the extruded disk materials from the vertebral canal. Postoperative follow-up According to a previous study, improvement of the neurologic signs was defined as follows: Grade 1, disappearance of hypersensitivity; Grade 2, improvement of ataxia; Grade 3 and 4, ambulatory; Grade 5, improvement of deep pain sensation and ambulatory [19]. When the neurologic signs improved but then deteriorated again, diagnostic imaging and surgery were performed and confirmed to be recurrent type I TL-IVDH. Image analysis Measurement of the disk height index: DHI is the ratio of disk height to adjacent vertebral body length [13]. In this context, the term “height” refers to the thickness of the intervertebral disk measured as the distance between adjacent vertebral endplates on the midsagittal plane. The DHI of the initial affected site was measured at the time of initial onset and the time of recurrence. For DHI measurement, 3-D Multiplanar reconstruction (MPR) images were first generated. 3-D MPR images were created using an open-source PACS workstation DICOM viewer (Horos version 3.3.6; Nimble Co LLC., Annapolis, MD, USA). On the dorsal plane, the image orientation was adjusted so that the spinous processes of the vertebrae adjacent to the target intervertebral disk were aligned in a straight line. Subsequently, on the transverse plane, the orientation was adjusted to correct for vertebral rotation by aligning the spinous processes along the midline, and a proper midsagittal image was reconstructed accordingly. The DHI was calculated using the midsagittal image of the disk by dividing twice the sum of the dorsal, mid, and ventral intervertebral disk heights by the sum of the dorsal, mid, and ventral vertebral lengths of the cranial and caudal vertebra (Fig. 1Fig. 1.A midsagittal CT-myelographic image of the herniated intervertebral disk at the time of initial onset. The disk height index is calculated using the formula: 2× (a+b+c)/(d+e+f+g+h+i).) [13]. These procedures were carried out as described previously [19]. Changes in the DHI of the initial affected site at the time of initial onset and recurrence were expressed as the %DHI and normalized to the measured preoperative disk height (%DHI=100 × postoperative DHI/preoperative DHI) [13]. A midsagittal CT-myelographic image of the herniated intervertebral disk at the time of initial onset. The disk height index is calculated using the formula: 2× (a+b+c)/(d+e+f+g+h+i). Measurement of maximal transverse spinal cord compression ratio: The spinal cord compression ratio in the affected site was analyzed using an open-source PACS workstation DICOM viewer (Horos version 3.3.6; Nimble Co LLC., Annapolis, MD, USA). Using CT-myelographic images, transverse images of the maximal compressed spinal cord and uncompressed (normal) spinal cord were obtained, and their areas were measured using Region of Interest (ROI). The normal spinal cord was defined as the area of the spinal cord that was not compressed and was cranial to and closest to the spinal cord compression site. The contours of the ROI were defined above the contrast medium. The maximal transverse spinal cord compression ratio was calculated as the area of the uncompressed spinal cord minus the area of the maximal transverse compressed spinal cord, divided by the area of the uncompressed spinal cord and multiplied by 100 [16]. These procedures were carried out as described previously [19]. Statistical analysis All analyses were performed using statistical analysis software (Prism version 10.4.1; GraphPad Software Inc., Boston, MA, USA). The Shapiro-Wilk normality test and paired t-test was used to compare the DHI of the initial affected site at the time of initial onset and recurrence. The Shapiro-Wilk normality test and paired t-test was also used to compare the spinal cord compression ratio at the initial affected site at the times of initial onset and recurrence. The Shapiro-Wilk normality test and Spearman’s rank correlation coefficient was used to evaluate the correlation between the %DHI and the time from initial onset to recurrence. P<0.05 was taken as statistically significant. All data were expressed as the mean ± SD. Correlations were classified as follows: <0.19, very weak correlation; 0.2 to 0.39, weak correlation; 0.4 to 0.59, moderate correlation; 0.6 to 0.79, strong correlation; and >0.8, very strong correlation. RESULTS Animals A total of 504 dogs were diagnosed with type I TL-IVDH by CT, CT-myelography and hemilaminectomy during the interval of this study. Of these, 19 dogs (3.8%) were eligible for inclusion. These were 16 Miniature Dachshunds, 1 Toy Poodle, 1 French Bulldog, and 1 Mixed breed dog. There were 8 intact males, 3 males neutered, 4 intact females, and 4 females spayed. Their median age was 7 years (range, 2–15 years) and median body weight was 6.7 kg (range, 4.0–11.9 kg) at the time of initial onset. Clinical findings for dogs at the time of initial onset and after hemilaminectomy The grade of neurologic severity at the time of initial onset was grade 2 in 1 dog, grade 3 in 7 dogs, grade 4 in 9 dogs and grade 5 in 2 dogs. Disk herniations at the time of initial onset were located at T12-13 in 2 dogs, T13-L1 in 4 dogs, L1-2 in 3 dogs, L2-3 in 8 dogs, L3-4 in 1 dog, and L4-5 in 1 dog. After 1st surgery, all dogs with grade 2 and grade 3 at the time of initial onset became Grade 0. Seven dogs with grade 4 became grade 0, and 2 became grade 2. One dog with grade 5 became grade 0, and one became grade 2. The improvement rate in neurologic signs after 1st surgery was 100%. The median interval from 1st surgery to improvement of the neurologic signs was 5 days (range, 1–29 days). Clinical findings for dogs at the time of recurrence and after surgery The median interval from initial onset to recurrence was 454 days (range, 52–2,119 days). The grade of neurologic severity at the time of recurrence was grade 2 in 2 dogs, grade 3 in 1 dog, grade 4 in 12 dogs and grade 5 in 4 dogs. Disk herniations at the time of recurrence were located at T10-11 in 1 dog, T11-12 in 2 dogs, T12-13 in 2 dogs, T13-L1 in 7 dogs, L1-2 in 2 dogs, L2-3 in 3 dogs, and L3-4 in 3 dogs. In one dog, the affected site involved two vertebrae between T10-T12. No dog was found in which the affected site at the time of recurrence was the same as that at the time of initial onset. The affected site at the time of recurrence was the cranial intervertebral disk adjacent to the initial affected site in 8 dogs, the caudal intervertebral disk in 3 dogs, and other sites in 8 dogs. After 2nd surgery, one dog with grade 2 at the time of recurrence became grade 0, and 1 dog remained grade 2. One dog with grade 3 became grade 0. Six dogs with grade 4 became grade 0, and 6 became grade 2. Two dogs with grade 5 became grade 2, 1 became grade 3, and 1 became grade 4. The improvement rate of neurologic signs after 2nd surgery was 89.5%. The DHI before and long after hemilaminectomy The DHI at the initial affected site at the time of initial onset was 0.0946 ± 0.0167 (Table 1Table 1.Comparison of the disk height index and the spinal cord compression rate at the initial affected site between at the time of initial onset and recurrenceInitial onsetRecurrenceDisk height index0.0946 ± 0.01670.0843 ± 0.0141Spinal cord compression rate (%)50.8 ± 17.015.7 ± 11.2). The DHI at the initial affected site at the time of recurrence was 0.0843 ± 0.0141. The DHI at the initial affected site at the time of recurrence was significantly less than that at the time of initial onset (P=0.0009). The %DHI was 89.6 ± 9.59%. There was no correlation between the %DHI and the time from initial onset to recurrence (Fig. 2Fig. 2.Scatterplots showing the % disk height index and the time from initial onset to recurrence. There was no correlation between the % disk height index and the time from initial onset to recurrence.). Table 1.Comparison of the disk height index and the spinal cord compression rate at the initial affected site between at the time of initial onset and recurrenceInitial onsetRecurrenceDisk height index0.0946 ± 0.01670.0843 ± 0.0141Spinal cord compression rate (%)50.8 ± 17.015.7 ± 11.2 Comparison of the disk height index and the spinal cord compression rate at the initial affected site between at the time of initial onset and recurrence Scatterplots showing the % disk height index and the time from initial onset to recurrence. There was no correlation between the % disk height index and the time from initial onset to recurrence. Maximal spinal cord compression rate before and long after hemilaminectomy The maximal spinal cord compression rate at the time of initial onset was 50.8 ± 17.0% (Table 1). The maximal spinal cord compression rate at the initial affected site at the time of recurrence was 15.7 ± 11.2%. The maximal spinal cord compression rate at the initial affected site at the time of recurrence was significantly less than that at the time of initial onset (P<0.0001). In all dogs, spinal cord compression was confirmed at the site of 1st surgery at the time of recurrence. Changes in the vertebral body before and long after hemilaminectomy One dog had spondylosis deformans at the initial affected site at the time of initial onset. Four dogs had spondylosis deformans at the initial affected site at the time of recurrence. Two dogs had bone regrowth at the hemilaminectomy site of 1st surgery at the time of recurrence (Fig. 3Fig. 3.CT-myelographic images at the level of the T12-13 in a dog. (A) Preoperative transverse CT-myelographic image. (B) Transverse CT-myelographic image at the time of recurrence. Bone regrowth was observed around the surgically resected vertebral arch.). One of these 2 dogs did not have spondylosis deformans at the initial affected site at the time of initial onset, but did have it at the time of recurrence. CT-myelographic images at the level of the T12-13 in a dog. (A) Preoperative transverse CT-myelographic image. (B) Transverse CT-myelographic image at the time of recurrence. Bone regrowth was observed around the surgically resected vertebral arch. DISCUSSION In this study, the DHI at the surgical site decreased after hemilaminectomy. Spinal cord compression at the initial affected site was observed in all dogs at the time of recurrence. Furthermore, no correlation was observed between the %DHI and the time to recurrence. These findings suggest that the decrease in the DHI at the surgical site may not necessarily reflect only the progression of intervertebral disk degeneration over time. In this study, spinal cord compression at the site of the initial surgery was identified at the time of recurrence in all dogs, with a compression rate of approximately 15%. In all of these dogs, neurologic signs improved immediately after 1st surgery and there was no apparent deterioration of neurologic signs until recurrence. In type I TL-IVDH, extruded disk materials cause compression and contusion of the spinal cord, giving rise to neurologic signs [10]. It follows that spinal cord compression is only one of the factors causing neurologic signs, and neurologic signs may improve even if spinal cord compression remains after surgery. In a previous study, MRI was performed immediately after surgery for type I TL-IVDH, and showed spinal cord compression in all dogs, with a compression rate of about 35% [3]. The rate of spinal cord compression immediately after surgery was not associated with improvement in neurologic signs [3]. In the present study, it is difficult to determine whether the spinal cord compression at the initial surgical site was caused by residual disk materials left after surgery or by newly extruded disk materials after surgery, because a long time had passed between the surgery and recurrence. Recovery of motor function was achieved after 2nd surgery in approximately 90% of dogs, suggesting that these spinal cord compressions were just sufficient not to cause neurologic signs. Bone regrowth was observed around surgically resected vertebral arches in 2 of 19 dogs. In a previous veterinary case report, a case of spinal cord compression due to bone regrowth at the surgical site after hemilaminectomy was reported, requiring reoperation [23]. The two dogs in which bone regrowth was observed in this study may also require revision hemilaminectomy in the future. In this study, the spondylosis deformans was evaluated to investigate structural changes at the surgical site following hemilaminectomy. Spondylosis deformans tends to be more prominent in middle-aged to older dogs. Newly developed spondylosis deformans at the surgical site at the time of recurrence was observed in 3 dogs. Based on these findings, it remains unclear whether the development of spondylosis deformans is attributable to age-related changes, the effect of the surgical intervention, or other factors. The late recurrence rate of type I TL-IVDH in this study was 3.8%. Previous studies have found late recurrence rates of 2.3–36% after surgery [1, 5, 12, 14]. Some of these studies define recurrence based on neurologic signs, whereas others define it based on surgery. The lower recurrence rate in this study compared to previous studies may be due to surgical confirmation being an inclusion criterion. In veterinary medicine, recurrence at the surgical site is uncommon in late recurrence of type I TL-IVDH [5, 6, 14]. When recurrence of clinical signs was observed, mild IVDH was detected at the same site; however, no cases were identified in which this site was considered the responsible lesion in our study. One possible explanation is that, once disk materials have extruded, the remaining material available for subsequent extrusion is reduced, making re-extrusion at the same site less likely to be sufficient to result in neurological signs. Minor re-extrusion may occur and contribute to the decrease in DHI. This may explain the apparent discrepancy between the low rate of clinical recurrence at the same site and postoperative changes in DHI. At our institution, we perform mini-hemilaminectomy or partial lateral corpectomy in dogs when adhesions between the herniated disk materials and spinal cord are suspected or when type II TL-IVDH is suspected. This study focused on small-breed dogs with type I TL-IVDH, and thus did not evaluate changes at the surgical site resulting from other surgical procedures. Further studies comparing hemilaminectomy with other procedures, such as mini-laminectomy, are expected to provide a more detailed assessment of how different surgical approaches affect the surgical site. There were some limitations in this study including the retrospective single-center design, the small number of dogs included, bias by breed, lack of biomechanical assessment of stability at the site of hemilaminectomy, and lack of assessment of disk degeneration by MRI. Dogs with early recurrence were excluded from this study in order to evaluate changes at the surgical site over a long period of time after surgery. It is known that early recurrence often occurs at the same site as the surgical site [9, 20, 22]. In addition, dogs in which a hemilaminectomy had been performed and a subsequent abdominal CT was available were not included. The CT acquisition parameters used for abdominal imaging differ substantially from those used for spinal CT examinations. These differences may affect the accuracy and consistency of intervertebral disk space measurements. Therefore, to maintain methodological consistency and ensure reliable morphometric evaluation, only cases examined using the standardized spinal CT protocol were included in this study. Furthermore, intra- and inter-observer variability and measurement reproducibility of DHI were not evaluated. Although DHI is a well-established measurement method widely used in dogs, the lack of reproducibility assessment represents a methodological limitation. In addition, blinding during measurement of DHI and spinal cord compression rate was not performed. In most cases, the surgical site was readily identifiable on imaging due to partial removal of the vertebral arch at the time of initial surgery, making observer blinding impractical. These factors may have introduced observer bias and should be considered when interpreting the results. Because this study was descriptive study, we were not able to evaluate the associations between changes in DHI and postoperative spinal cord compression rate, bone regrowth at the surgical site, spondylosis deformans, or the late recurrence rate. Further studies are needed to clarify whether and how these factors are related to postoperative changes in DHI. In conclusion, structural changes, including decreased DHI and bone regrowth, were observed at the surgical site long after hemilaminectomy, but these changes were not the direct cause of the neurologic signs observed at the time of recurrence. Conflicts of Interest The authors declare no conflict of interest to this report.