Early Reherniation of Disk Material in Eleven Dogs with Surgically Treated Thoracolumbar Intervertebral Disk Extrusion


  • Bianca F. Hettlich Dr medvet, Diplomate ACVS,

    Corresponding author
    • Department of Small Animal Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, TX
    Search for more papers by this author
  • Sharon C. Kerwin DVM, Diplomate ACVS,

    1. Department of Small Animal Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, TX
    Search for more papers by this author
  • Jonathan M. Levine DVM, Diplomate ACVIM (Neurology)

    1. Department of Small Animal Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, TX
    Search for more papers by this author

  • Presented in part at the ESVOT Congress, Munich, Germany, September 2006.

Corresponding Author

Bianca Hettlich, Dr medvet, Diplomate ACVS, Department of Veterinary Clinical Sciences, 601 Vernon Tharp Street, Columbus, OH 43210 E-mail: bianca.hettlich@cvm.osu.edu



To report findings and outcomes of dogs with reherniation of nuclear material within 7 days of hemilaminectomy for acute thoracolumbar (TL) intervertebral disk extrusion.

Study Design

Retrospective case series.


Chondrodystrophic dogs (n = 11).


Dogs with acute neurologic decline within 1 week of surgical decompression for TL disk extrusion were identified. Advanced imaging was used to document extradural spinal cord compression at the previous surgery site. Ten dogs had a 2nd decompressive surgery to remove extruded nuclear material.


All dogs had acute neurologic deterioration (average, 2 neurologic grades) 2–7 days after initial hemilaminectomy. Computed tomography (CT; n = 10) or myelography (n = 1) documented extradural spinal cord compression compatible with extruded disk material at the previous hemilaminectomy site. Dogs that had a 2nd surgical decompression improved neurologically within 24 hours and were paraparetic at discharge. The single dog that did not have decompressive surgery did not regain deep nociception during 185-day follow-up.


Early reherniation at the site of previous hemilaminectomy can produce acute deterioration of neurologic function and should be investigated with diagnostic imaging. Repeat decompressive surgery can lead to functional recovery.

Surgical decompression is a common treatment for thoracolumbar (TL) intervertebral disk extrusion (IVDE) in dogs. Presurgical determinants of functional recovery include the rapidity of neurologic decline, presence of pelvic limb deep nociception, cerebrospinal fluid myelin basic protein concentration, and T2-weighted magnetic resonance imaging (MRI) signal of the injured spinal cord.[1-5] Affected dogs with intact deep nociception after surgery typically show neurologic improvement and will regain the ability to walk within 2–4 weeks.[2, 6, 7] In this population, neurologic decline in the perioperative and early postoperative period (<1 month) is believed to be uncommon. Factors postulated to be associated with early postoperative neurologic decline include ongoing secondary spinal cord injury, iatrogenic primary spinal cord injury, iatrogenic vertebral column instability, failure to adequately decompress the spinal cord, extrusion of additional nuclear material from the surgically addressed site of disk herniation, disk herniation at a previously unaffected intervertebral articulation, or infection.[8-12] Imaging studies have documented that residual disk material from initial decompressive surgery can lead to poor postoperative neurologic improvement.[10] It has also been shown that nonfenestrated disks can extrude additional material into the vertebral canal by 6 weeks after initial decompressive surgery[13]; however, there are few reports detailing outcomes for dogs with reherniation undergoing a 2nd surgery in the early postoperative period.[13, 14]

Our purpose is to report outcome in 11 dogs with acute neurologic decline within 7 days of decompressive surgery for TL IVDE with imaging and surgical (10 dogs) confirmation of reherniated nuclear material at the initial hemilaminectomy site.


Medical records (2004–2008) were searched for dogs that had hemilaminectomy for TL (affecting articulations T3-L7) IVDE. Dogs were included in the study if they had acute deterioration of neurologic signs within 7 days of initial hemilaminectomy and subsequently had advanced diagnostic imaging to confirm reherniation of nuclear material from the same disk space.

A modified Frankel score (MFS) was used to rate the severity of spinal cord injury in all dogs.[15, 16] Dogs were classified as having paraplegia without deep nociception (grade 0), paraplegia without superficial nociception (grade 1), paraplegia with superficial nociception (grade 2), nonambulatory paraparesis (grade 3), ambulatory paraparesis and ataxia (grade 4), or spinal hyperesthesia only (grade 5).

Presurgical MFS, type of diagnostic imaging (myelography, computed tomography [CT], MRI), surgical findings during the 1st decompressive procedure, immediate postoperative MFS, MFS when neurologic decline was identified, type of subsequent diagnostic imaging (myelography, CT), findings during the 2nd surgical procedure (if applicable), and long-term functional outcome were recorded. Long-term outcome data were generated from neurologic examination at time of in-hospital recheck and subsequent telephone follow-up with owners or referring veterinarians. A previously described questionnaire[17] correlating owner answers with veterinary neurologic examination results was used for phone follow-up. In addition, the following intervals were recorded: time between onset of initial neurologic signs and 1st diagnostic imaging and surgery, time between 1st surgery and recurrence of signs, time to 2nd advanced imaging and surgery (if applicable), and time between 2nd surgery and discharge.


Five hundred and sixty TL decompressive surgeries were performed between 2004 and 2008. Early reherniation occurred in 11 dogs, a 2% incidence. The study population consisted exclusively of chondrodystrophic dogs: 8 Dachshunds, 2 Dachshund mixed breed dogs, and 1 Chihuahua (Table 1). Median age was 5 years (range, 4–8 years) and median weight was 7 kg (range, 3.9–8.9 kg). All dogs were admitted with an acute T3-L3 myelopathy with either nonambulatory paraparesis or paraplegia with intact superficial nociception (median, 2; range, 1–3; Table 2). Diagnostic workup was pursued for less than 48 hours from onset of neurologic deficits. Myelography by lumbar injection (9 dogs) confirmed extradural spinal cord compression compatible with extruded disk material. Computed Tomography (GE LightSpeed QXI 4 Slice H1 Gantry, Piscataway, NJ) and MRI (Siemens Magnetom Expert 1.0T, Malvern, PA) were performed in 1 dog each, which also confirmed an extradural compressive lesion compatible with extruded nuclear material. Localization of disk extrusion was: T11-T12 = 3; T12-T13 = 3; T13-L1 = 4; and L1-L2 = 1. Hemilaminectomy was performed in all dogs within 48 hours of onset of neurologic signs; 8 dogs had a single-level hemilaminectomy, whereas 2 had a 2-level and 1 a 3-level hemilaminectomy depending on the extent of extradural compression. Adequate decompression and removal of extruded nuclear material was assessed subjectively by the individual surgeon and was based on removal of all visible extruded material as well as lack of recovery of further material by probing from the ventral, dorsal, and ventrolateral aspect of the contralateral side of the vertebral canal. In 1 dog, camera assistance was used to evaluate the ventral aspect of the vertebral canal (1.9-mm arthroscope, Dyonics, Smith&Nephew, Andover, MA). One dog had postoperative CT as part of an unrelated study, which allowed for more objective evaluation of appropriate decompression.

Table 1. Summary Data for 11 dogs with Reherniation of Nuclear Material at a Previous Disk Extrusion Site
BreedAge (Years)Sex1st ImagingLocalization2nd Imaging2nd Surgery
  1. CT, computed tomography; MN, male neutered; M, male; FS, female spayed.

Dachshund5MNMyelogramT13/L1MyelogramSame side
Dachshund5MMyelogramT13/L1CTSame side
Dachshund4MNMyelogramT11/T12CTSame side
Dachshund6FSMyelogramT12-T13CTSame side
Dachshund6FSMyelogramT12-T13CTSame side
Dachshund4FSMyelogramT12-T13CTContralateral mini-hemilaminectomy
Dachshund7MCTT11/T12CTSame side
Dachshund mix5MNMyelogramT13/L1CTSame side
Dachshund mix4FSMRIT11-T12CTContralateral mini-hemilaminectomy
Chihuahua6FSMyelogramL1-L2CTContralateral mini-hemilaminectomy
Table 2. Summary of Modified Frankel Scores (MFS) for Each Dog at Different Evaluation Time Points
DogsPreoperativePostoperativeReherniationDischarge1st RecheckOwner Assessment
  1. Median value in parenthesis excludes dog #1 (nonsurgical case).

  2. a

    *Dog with presumed spinal walking.

Median23133 (3.5)4

In the postoperative period until deterioration, neurologic status remained unchanged in 6 dogs and improved in 5 dogs. Superficial nociception was intact in all dogs with a median MFS of 3 (range, 2–4) before neurologic worsening. Nine were hospitalized for postoperative care and rehabilitation. The main evaluating clinician who was also the surgeon remained the same throughout the entire hospitalization period for all 9 dogs. Two dogs were discharged from the hospital before neurologic worsening and were subsequently evaluated and operated on by a different primary clinician.

All dogs had acute neurologic decline 2–7 days postoperatively (average, 3.9 days) with a change of MFS of ≥2 grades in 9 dogs. Only 2 dogs maintained superficial nociception, whereas the other dogs lost either superficial or deep nociception (median MFS, 1; range, 0–3) after acute worsening. All dogs had repeat diagnostic imaging (CT, 10 dogs; myelography, 1 dog) to determine the cause of neurologic decline. Three CT scans were obtained under sedation (medetomidine hydrochloride, 15 μg/kg intravenously). Imaging documented extradural spinal cord compression at the previous hemilaminectomy site compatible with the presence of disk material within the vertebral canal in all dogs (Fig 1).

Figure 1.

Transverse bone weighed computed tomography (CT) image at the level of the T12 vertebra in a 6-year-old female spayed Dachshund. A left-sided hemilaminectomy has been performed. A large mineralized opacity is occupying approximately 75% of the cross-sectional area of the vertebral canal (black arrow).

One dog was treated conservatively because of financial constraints and 10 dogs had a 2nd decompressive surgery after diagnostic imaging. Seven dogs were decompressed through the previous hemilaminectomy site and 3 dogs were decompressed through a contralateral mini-hemilaminectomy because of lateralization of the reherniated material to the opposite side. Subjectively, moderate-to-large amounts of nuclear material were recovered from all operated dogs, which were similar to amounts recovered during the initial decompressive surgery (in 8 repeat surgeries, the same surgeon also performed the 1st procedure allowing for more consistent assessment). Dogs that had a 2nd surgery improved neurologically within 24 hours of surgery and had a median MFS of 3 (range, 2–3) at time of discharge (mean, 6.9 days). Eight dogs returned for neurologic evaluation between 12 and 60 days postoperative (mean, 36.9 days). Median MFS at recheck had improved to 3.5. Three of 4 dogs classified as nonambulatory had improved motor function compared with previous examinations but were still categorized as MFS 3.

Telephone follow-up was available for 7 surgical dogs, 4–30 months after the 2nd surgery (mean, 18.4 months). All owners reported their dogs to be significantly improved with occasional weakness (MFS = 4). The single, conservatively treated dog remained paraplegic with absent deep nociception at a neurologic recheck 60 days after the 2nd imaging examination. During a 2nd recheck at 185 days, the dog still had absent deep nociception but was ambulatory paraparetic. It was determined that the dog had developed spinal walking.


Although there are many potential causes of neurologic decline after decompressive spinal surgery for acute disk extrusion, herniation of another disk appears to be most common.[11, 18, 19] The incidence of disk extrusions at other disk spaces has been reported to range from 2.6% to 26.5%.[19] Recurrence of neurologic deficits less than 1 month after surgery are because of the development of disk herniation at a new site, whereas recurrence of signs within the 1st month was because of extrusion of more nuclear material from the previously injured and operated disk.[11] Confirming the exact recurrence rate because of reherniation or extrusion of a new disk is difficult to determine from published reports. Not all recurrences reported were confirmed by imaging or surgery but were suspected based on clinical signs, the evaluation periods were greatly variable, and the reported outcome was often a mixture between previously operated and new disk spaces.

Figure 2.

Transverse soft tissue weighted CT images at level of the T11-T12 vertebral articulation from a 7-year-old male Dachshund. (A) Preoperative CT image showing a mineralized opacity (black arrow) within the vertebral canal at the T11-T12 intervertebral disk space. (B) Postoperative CT image documenting removal of the majority of extruded mineralized material via hemilaminectomy. (C) CT image after acute neurologic decline showing the presence of a large amount of mineralized material within the vertebral canal (black arrow head).

Recently, an MRI study on chondrodystrophic dogs after TL hemilaminectomy for disk extrusion found residual (rather than reherniated) nuclear material to be the cause of continued spinal cord compression and poor postoperative recovery in 8 of 10 dogs.[10] The residual disk was thought to have been caused by incomplete removal at initial surgery (3 dogs), approach to the incorrect intervertebral disk space (3 dogs), and inadvertent shift of disk to the contralateral side of the vertebral canal (1 dog). Another report describes early recurrence of disk extrusion after hemilaminectomy in 3 dogs.[14] These dogs either did not improve or worsened within the 1st 10 days of surgery, and neurologic decline was assumed to be because of extrusion of more nuclear material. Interestingly, 1 of these 3 dogs had undergone disk fenestration of the affected site at time of 1st decompressive surgery but still extruded more nuclear material postoperatively. Although this occurred only in 1 dog, it still highlights the difficultly to ensure complete removal of the nucleus pulposus by fenestration.

Postoperative imaging could document adequate decompression after hemilaminectomy and adequate removal of extruded nuclear material. Repeat spinal imaging after decompressive surgery is not routinely performed in veterinary medicine and most surgeons rely on inspection and probing of the vertebral canal to ensure adequate removal of disk material. In this study, we made the assumption that recurrence of neurologic deficits after hemilaminectomy was because of extrusion of additional nuclear material rather than the presence of residual material. This was supported by several facts. The vertebral canal was at least subjectively adequately decompressed based on findings during the initial surgery; however, it is possible that some nuclear material remained within the spinal canal. In 2 dogs, proper initial decompression was documented by the use of intraoperative endoscopy of the vertebral canal and postoperative CT (Fig 2).

All dogs either remained neurologically stable (6 dogs) or improved (5 dogs) after the 1st surgery before showing an acute deterioration. This decline was usually severe with a decrease of 2 neurologic grades on average, leading to loss of superficial or deep nociception in 9 dogs. Nine dogs were still hospitalized and under the care of the same clinician when the neurologic deterioration occurred, which allowed for more consistent evaluation. Imaging after acute neurologic worsening documented the presence of moderate-to-large amounts of nuclear material (either in form of an extradural compression on myelography or the presence of mineralized material within the vertebral canal on CT), which according to the individual surgeons was not present at the conclusion of the initial surgery. Extradural compression was located over the previously extruded disk making it less likely that the nuclear material originated from an adjacent intervertebral disk space.

Indirect assessment of adequate decompression during the 1st surgery by comparing initial and postdecline images was difficult as the repeat imaging modality was different in most dogs and involved a less-precise modality initially (myelography). Because of an ability to produce transverse images, CT and MRI are better able to determine lateralization of compressive nuclear material and may be useful for the assessment of dogs with suspect reherniation. In our study, pre- and postdecline images of 3 dogs were considered of value in supporting reherniation. One dog had myelography in both instances, which suggested a left ventral compression initially followed by a right lateral compression. One dog had CT imaging immediately after the 1st surgery, which documented adequate decompression, with a subsequent CT showing renewed mineralized extradural compression. One dog had MRI initially, followed by CT; both modalities showed compression on opposite sides of the vertebral canal.

Possible causes for early reherniation have not been identified. Uncontrolled postoperative activity may exert undue pressure on the already injured intervertebral disk and allow further nuclear material to extrude through the compromised annulus fibrosus. Strict cage confinement for several weeks is generally recommended after hemilaminectomy to allow scar tissue development and fibrosing of annular tears. Conversely in people with lumbar disk prolapse, lack of postoperative restrictions did not correlate with recurrence of signs or the ultimate outcome.[20] In our study, the degree of postoperative activity was unlikely the cause of reherniation as 9 dogs were still hospitalized and strict cage rest was enforced. All dogs had standard postoperative exercises such as passive range of motion and balancing exercises starting 24–48 hours after surgery. Aquatic therapy was initiated 3–4 days postoperatively. It is unknown whether postoperative rehabilitation has an influence on the recurrence of disk extrusion. In a review of randomized, controlled human trials assessing the effects of rehabilitation after 1st-time lumbar disk surgery, no evidence was found that rehabilitation increased the reoperation rate[21]; however, no such review has been performed for dogs undergoing rehabilitation after disk surgery.

Dachshunds have a significantly higher risk for late reoperation and the overall prognosis for dogs with reoperation was identical to dogs only having had 1 surgery.[11] Even with a 2nd surgery, a functional ambulatory outcome was still achieved in 91% of reoperated dogs.[11] Regarding early reherniation, that study only had 5 dogs with recurrence of neurologic signs between 1 and 21 days after initial surgery. Outcome data were available for 4 of these 5 dogs, with 2 dogs recovering voluntary ambulation. In our study, all 10 dogs that had a 2nd surgery regained voluntary motor function before discharge from the hospital. Considering that the average MFS was 0.8 after acute neurologic deterioration, this was an excellent improvement in a short period of time. The single conservatively treated dog continued to lack deep nociception, but did develop spinal walking 6 months after proposed reherniation.

Fenestration of a herniated disk has been recommended to reduce the incidence of future disk extrusion; however, it is still unknown to what degree fenestration really prevents extrusion of additional nuclear material.[14, 18, 22, 23] In a prospective MRI study assessing the effect of fenestration on recurrence of extrusion, the fenestrated group showed no evidence of reherniation, whereas 6 of 10 dogs had evidence of further disk extrusion into the vertebral canal at 6 weeks post surgery.[13] Although the number of dogs was small (n = 19), it still documents that a certain degree of further disk extrusion is possible, but may not lead to clinical signs. Only 1 dog deteriorated significantly, warranting repeat decompression to remove reherniated disk material. This study also found that no fenestration was complete and all fenestrated dogs had residual nuclear material within the disk, allowing for potential of reherniation. Different spinal approaches and techniques for disk fenestration have been evaluated to assess complete removal of nuclear material.[24, 25] Unfortunately, no currently reported fenestration procedure can completely assure that the nucleus pulposus is entirely removed. Further research is required to assess the most effective method of disk fenestration as well as prospective studies on its effectiveness in preventing further extrusion of nuclear material.

Rates of early postoperative recurrence of neurologic signs have been reported from 1 to 7.7%.[10, 11, 26] The incidence of early reherniation in our case series was 2% over a 4-year period (2004–2008). Data regarding outcome after confirmed early reherniation are limited to a small number of cases but most dogs appear to show good functional recovery.[8, 12, 15, 20] Reoperation, whether it be shortly after previous decompressive surgery or later, also does not seem to influence time to or degree of recovery when compared with dogs undergoing their 1st spinal surgery.[7] Considering the consistently good functional recovery in operated dogs in our study, we strongly recommend advanced imaging in dogs with acute neurologic decline in the early postoperative period to allow for timely reoperation to remove renewed spinal cord compression.

Neurologic decline in dogs with surgically treated TL IVDE can occur in the early postoperative period because of reherniation of disk material. Advanced imaging modalities (CT, MRI), which allow transverse imaging of the vertebral canal, may aid in recognizing reherniation by better determining the exact location of compressive material within the canal. In dogs with early reherniation, surgical removal of extruded nuclear material resulted in functional ambulation in all cases. A proactive approach that includes advanced vertebral column imaging should be taken in dogs with surgically treated TL IVDE that exhibit early postoperative neurologic decline.